repository_name
stringclasses
316 values
func_path_in_repository
stringlengths
6
223
func_name
stringlengths
1
134
language
stringclasses
1 value
func_code_string
stringlengths
57
65.5k
func_documentation_string
stringlengths
1
46.3k
split_name
stringclasses
1 value
func_code_url
stringlengths
91
315
called_functions
listlengths
1
156
enclosing_scope
stringlengths
2
1.48M
mugurbil/gnm
gnm/utils.py
update_params
python
def update_params(state, t): mu = (1.-t)*state['x'] + t*state['mu'] L = state['L'] / np.sqrt(2.*t - t**2) return mu, L
Update parameters updates mean and precision to the step size Inputs: state : mu : mean L : cholesky factor of the precision matrix t : step size Outputs: mu : updated mean L : updated cholesky factor of the precision matrix
train
https://github.com/mugurbil/gnm/blob/4f9711fb9d78cc02820c25234bc3ab9615014f11/gnm/utils.py#L22-L42
null
# -*- coding: utf-8 -*- """ Functions needed by the gnm sampler 1. test 2. update_params 3. log_K 4. multi_normal 5. det 6. optimize 7. function """ __all__ = ['test','update_params','log_K','multi_normal', 'det','optimize','function'] import numpy as np la = np.linalg def test(): import quickstart def log_K(Z, X, t): """ Log K Log of the proposal probability density function for gnm Inputs : Z : proposed to x : proposed from Outputs : log of the probability density function """ m, L = update_params(X, t) z = Z['x'] return np.log(det(L))-la.norm(np.dot(L.T,z-m))**2/2. def multi_normal(X, t): """ Multivariate normal sampler: Generates normal samples with mean m, precision matrix LL' Inputs: x : propose from Outputs: normal with mean m and precision LL' """ m, L = update_params(X, t) z = np.random.standard_normal(np.shape(m)) # generate i.i.d N(0,1) return la.solve(L.T,z)+m def det(L): """ Determinant Compute the determinant given a lower triangular matrix Inputs: L : lower triangular matrix Outputs: det_L : determinant of L """ size_L = L.shape if np.size(L) == 1: return np.array(L) else: try: assert np.all(np.tril(L)==L) except AssertionError: print 'Error: Input is not a lower triangular matrix.' return 0 try: assert size_L[0] == size_L[1] except AssertionError: print 'Error: Not a square matrix.' return 0 det_L = 1. for i in xrange(size_L[1]): det_L = det_L*L[i,i] return det_L def optimize(t, f_0, d_f_0, f_t, d_f_t, t1=0.05, t2=0.5): """ Third order approximation to find the minimum of the function f : function to be optimized over Inputs : t : previous step size f_0 : f(0), function value at 0 d_f_0: f'(0), the derivative of the function at 0 f_t : f(t), function value at t d_f_t : f'(t), the derivative of the function at t t1 : step size reduction if minimum is at 0 or it can't be found t2 : step size reduction if minimum is at 1 Outputs : alpha_new : the new step size that minimizes the function """ if t <= 0 : print("Error: please enter non-negative t") return t a = (t*d_f_t-2*f_t+2*f_0+t*d_f_0)/(t**3) b = (f_t-f_0-t*d_f_0)/(t**2) c = d_f_0 A = 3*a B = b-t*a C = B**2-A*c if C == 0. : if c>0 or d_f_t>0 : t_new = t1*t else : t_new = t2*t elif A == 0 : t_new = -c/2./B elif C > 0 : t_new = (-B+np.sqrt(C))/A else : t_new = t1*t # check the bounds on new step size if t_new < t1*t : t_new = t1*t elif t_new > t2*t : t_new = t2*t return t_new class function(object): def __init__(self, f, args): """ Init Initialize the developer function class Inputs : f : user defined function --- Inputs of f : x : input value args : the arguments that the function takes Outputs of f : chi_x : boolean flag indicating whether the function is defined at x or not f_x : f(x), function value at x J_x : f'(x), the jacobian of the function evaluated at x Demo : chi_x, f_x, J_x = f(x,args) --- args : the arguments that the user defined function takes """ self._f = f self._args = args self._count = 0 def __call__(self, x): """ Call Calls the user defined function Inputs: x : input value Outputs: chi_x, f_x, J_x = f(x,args) """ self._count += 1 x = np.reshape(np.array(x), (-1)) chi_x, f_x, J_x = self._f(x, self.args) f_x = np.reshape(np.array(f_x), (-1,1)) return chi_x, f_x, np.array(J_x) def Jtest(self, x_min, x_max, dx=0.0002, N=1000, eps_max=0.0001, p=2, l_max=50, r=0.5): """ Gradient Checker Test the function's jacobian against the numerical jacobian Inputs : x_min : lower bound on the domain x_max : upper bound on the domain ** Warning ** x_min and x_max must be arrays of the same dimension Optional inputs : dx : (2*10^-4) the ratio of dx to the size of the domain N : (1000) number of test points eps_max : (10^-4) the maximum value error is allowed to be to confirm convergence p : (2) to specify the norm of the error (p-norm) l_max : (40) maximum number of tries to reduce dx r : (0.5) dx will be multiplied by this constant each step the error exceeds error_bound until l_max is reached ** Warning ** keep in mind machine precision when changing l_max and r Outputs : error : * 1 if did it did not pass the checks, * 0 if converged, * eps, the error of the numerical gradient point if no convergence """ x_min = np.reshape(np.array(x_min), (-1)) x_max = np.reshape(np.array(x_max), (-1)) # begin checks try: for i in xrange(np.size(x_min)): assert x_min[i] < x_max[i] except: print("Error: All values of x_min should be less than the " +"corresponding values for x_max.") exit(0) # end checks # begin test k = 0 while k < N : # reset dx each time D_x = (x_max - x_min) * dx # Sample random points in the state space x = np.random.uniform(x_min + D_x, x_max - D_x) # Compute jacobian at x l = 0 test_convergence = 1 while test_convergence: chi_x, f_x, J_x = self.__call__(x) D_f_x = J_x * 0. for j in xrange(np.size(x)): # calculate the derivative of each component of f d_x = D_x * 0. d_x[j] = D_x[j] chi_x_r, f_x_r, J_x_r = self.__call__(x + d_x) chi_x_l, f_x_l, J_x_l = self.__call__(x - d_x) # check if the function is defined on these points if( not(chi_x and chi_x_r and chi_x_l)): # discard this trial if one of the values is not defined test_convergence = 0 # break outer loop break d_f = (f_x_r - f_x_l) / (2. * d_x[j]) D_f_x[:,j] = d_f[:,0] eps = la.norm(D_f_x - J_x, p) / np.size(J_x) if (eps < eps_max): test_convergence = 0 # break outer loop k += 1 else: D_x = D_x * r if (l > l_max): # numerical gradient did not converge return eps l += 1 return 0 # end test """ Properties: 1. f 2. args 3. count """ @property def f(self): return self._f @property def args(self): return self._args @property def count(self): return self._count ## end function ##
mugurbil/gnm
gnm/utils.py
log_K
python
def log_K(Z, X, t): m, L = update_params(X, t) z = Z['x'] return np.log(det(L))-la.norm(np.dot(L.T,z-m))**2/2.
Log K Log of the proposal probability density function for gnm Inputs : Z : proposed to x : proposed from Outputs : log of the probability density function
train
https://github.com/mugurbil/gnm/blob/4f9711fb9d78cc02820c25234bc3ab9615014f11/gnm/utils.py#L44-L58
[ "def det(L):\n \"\"\"\n Determinant\n Compute the determinant given a lower triangular matrix\n Inputs: \n L :\n lower triangular matrix\n Outputs: \n det_L : \n determinant of L\n \"\"\"\n size_L = L.shape\n if np.size(L) == 1:\n return np.array(L)\n else: \n try: \n assert np.all(np.tril(L)==L)\n except AssertionError:\n print 'Error: Input is not a lower triangular matrix.'\n return 0\n try:\n assert size_L[0] == size_L[1]\n except AssertionError:\n print 'Error: Not a square matrix.'\n return 0\n det_L = 1.\n for i in xrange(size_L[1]):\n det_L = det_L*L[i,i]\n return det_L\n", "def update_params(state, t):\n \"\"\"\n Update parameters\n updates mean and precision to the step size\n Inputs:\n state :\n mu :\n mean\n L :\n cholesky factor of the precision matrix\n t : \n step size\n Outputs:\n mu :\n updated mean\n L :\n updated cholesky factor of the precision matrix \n \"\"\"\n mu = (1.-t)*state['x'] + t*state['mu']\n L = state['L'] / np.sqrt(2.*t - t**2)\n return mu, L\n" ]
# -*- coding: utf-8 -*- """ Functions needed by the gnm sampler 1. test 2. update_params 3. log_K 4. multi_normal 5. det 6. optimize 7. function """ __all__ = ['test','update_params','log_K','multi_normal', 'det','optimize','function'] import numpy as np la = np.linalg def test(): import quickstart def update_params(state, t): """ Update parameters updates mean and precision to the step size Inputs: state : mu : mean L : cholesky factor of the precision matrix t : step size Outputs: mu : updated mean L : updated cholesky factor of the precision matrix """ mu = (1.-t)*state['x'] + t*state['mu'] L = state['L'] / np.sqrt(2.*t - t**2) return mu, L def multi_normal(X, t): """ Multivariate normal sampler: Generates normal samples with mean m, precision matrix LL' Inputs: x : propose from Outputs: normal with mean m and precision LL' """ m, L = update_params(X, t) z = np.random.standard_normal(np.shape(m)) # generate i.i.d N(0,1) return la.solve(L.T,z)+m def det(L): """ Determinant Compute the determinant given a lower triangular matrix Inputs: L : lower triangular matrix Outputs: det_L : determinant of L """ size_L = L.shape if np.size(L) == 1: return np.array(L) else: try: assert np.all(np.tril(L)==L) except AssertionError: print 'Error: Input is not a lower triangular matrix.' return 0 try: assert size_L[0] == size_L[1] except AssertionError: print 'Error: Not a square matrix.' return 0 det_L = 1. for i in xrange(size_L[1]): det_L = det_L*L[i,i] return det_L def optimize(t, f_0, d_f_0, f_t, d_f_t, t1=0.05, t2=0.5): """ Third order approximation to find the minimum of the function f : function to be optimized over Inputs : t : previous step size f_0 : f(0), function value at 0 d_f_0: f'(0), the derivative of the function at 0 f_t : f(t), function value at t d_f_t : f'(t), the derivative of the function at t t1 : step size reduction if minimum is at 0 or it can't be found t2 : step size reduction if minimum is at 1 Outputs : alpha_new : the new step size that minimizes the function """ if t <= 0 : print("Error: please enter non-negative t") return t a = (t*d_f_t-2*f_t+2*f_0+t*d_f_0)/(t**3) b = (f_t-f_0-t*d_f_0)/(t**2) c = d_f_0 A = 3*a B = b-t*a C = B**2-A*c if C == 0. : if c>0 or d_f_t>0 : t_new = t1*t else : t_new = t2*t elif A == 0 : t_new = -c/2./B elif C > 0 : t_new = (-B+np.sqrt(C))/A else : t_new = t1*t # check the bounds on new step size if t_new < t1*t : t_new = t1*t elif t_new > t2*t : t_new = t2*t return t_new class function(object): def __init__(self, f, args): """ Init Initialize the developer function class Inputs : f : user defined function --- Inputs of f : x : input value args : the arguments that the function takes Outputs of f : chi_x : boolean flag indicating whether the function is defined at x or not f_x : f(x), function value at x J_x : f'(x), the jacobian of the function evaluated at x Demo : chi_x, f_x, J_x = f(x,args) --- args : the arguments that the user defined function takes """ self._f = f self._args = args self._count = 0 def __call__(self, x): """ Call Calls the user defined function Inputs: x : input value Outputs: chi_x, f_x, J_x = f(x,args) """ self._count += 1 x = np.reshape(np.array(x), (-1)) chi_x, f_x, J_x = self._f(x, self.args) f_x = np.reshape(np.array(f_x), (-1,1)) return chi_x, f_x, np.array(J_x) def Jtest(self, x_min, x_max, dx=0.0002, N=1000, eps_max=0.0001, p=2, l_max=50, r=0.5): """ Gradient Checker Test the function's jacobian against the numerical jacobian Inputs : x_min : lower bound on the domain x_max : upper bound on the domain ** Warning ** x_min and x_max must be arrays of the same dimension Optional inputs : dx : (2*10^-4) the ratio of dx to the size of the domain N : (1000) number of test points eps_max : (10^-4) the maximum value error is allowed to be to confirm convergence p : (2) to specify the norm of the error (p-norm) l_max : (40) maximum number of tries to reduce dx r : (0.5) dx will be multiplied by this constant each step the error exceeds error_bound until l_max is reached ** Warning ** keep in mind machine precision when changing l_max and r Outputs : error : * 1 if did it did not pass the checks, * 0 if converged, * eps, the error of the numerical gradient point if no convergence """ x_min = np.reshape(np.array(x_min), (-1)) x_max = np.reshape(np.array(x_max), (-1)) # begin checks try: for i in xrange(np.size(x_min)): assert x_min[i] < x_max[i] except: print("Error: All values of x_min should be less than the " +"corresponding values for x_max.") exit(0) # end checks # begin test k = 0 while k < N : # reset dx each time D_x = (x_max - x_min) * dx # Sample random points in the state space x = np.random.uniform(x_min + D_x, x_max - D_x) # Compute jacobian at x l = 0 test_convergence = 1 while test_convergence: chi_x, f_x, J_x = self.__call__(x) D_f_x = J_x * 0. for j in xrange(np.size(x)): # calculate the derivative of each component of f d_x = D_x * 0. d_x[j] = D_x[j] chi_x_r, f_x_r, J_x_r = self.__call__(x + d_x) chi_x_l, f_x_l, J_x_l = self.__call__(x - d_x) # check if the function is defined on these points if( not(chi_x and chi_x_r and chi_x_l)): # discard this trial if one of the values is not defined test_convergence = 0 # break outer loop break d_f = (f_x_r - f_x_l) / (2. * d_x[j]) D_f_x[:,j] = d_f[:,0] eps = la.norm(D_f_x - J_x, p) / np.size(J_x) if (eps < eps_max): test_convergence = 0 # break outer loop k += 1 else: D_x = D_x * r if (l > l_max): # numerical gradient did not converge return eps l += 1 return 0 # end test """ Properties: 1. f 2. args 3. count """ @property def f(self): return self._f @property def args(self): return self._args @property def count(self): return self._count ## end function ##
mugurbil/gnm
gnm/utils.py
multi_normal
python
def multi_normal(X, t): m, L = update_params(X, t) z = np.random.standard_normal(np.shape(m)) # generate i.i.d N(0,1) return la.solve(L.T,z)+m
Multivariate normal sampler: Generates normal samples with mean m, precision matrix LL' Inputs: x : propose from Outputs: normal with mean m and precision LL'
train
https://github.com/mugurbil/gnm/blob/4f9711fb9d78cc02820c25234bc3ab9615014f11/gnm/utils.py#L60-L72
[ "def update_params(state, t):\n \"\"\"\n Update parameters\n updates mean and precision to the step size\n Inputs:\n state :\n mu :\n mean\n L :\n cholesky factor of the precision matrix\n t : \n step size\n Outputs:\n mu :\n updated mean\n L :\n updated cholesky factor of the precision matrix \n \"\"\"\n mu = (1.-t)*state['x'] + t*state['mu']\n L = state['L'] / np.sqrt(2.*t - t**2)\n return mu, L\n" ]
# -*- coding: utf-8 -*- """ Functions needed by the gnm sampler 1. test 2. update_params 3. log_K 4. multi_normal 5. det 6. optimize 7. function """ __all__ = ['test','update_params','log_K','multi_normal', 'det','optimize','function'] import numpy as np la = np.linalg def test(): import quickstart def update_params(state, t): """ Update parameters updates mean and precision to the step size Inputs: state : mu : mean L : cholesky factor of the precision matrix t : step size Outputs: mu : updated mean L : updated cholesky factor of the precision matrix """ mu = (1.-t)*state['x'] + t*state['mu'] L = state['L'] / np.sqrt(2.*t - t**2) return mu, L def log_K(Z, X, t): """ Log K Log of the proposal probability density function for gnm Inputs : Z : proposed to x : proposed from Outputs : log of the probability density function """ m, L = update_params(X, t) z = Z['x'] return np.log(det(L))-la.norm(np.dot(L.T,z-m))**2/2. def det(L): """ Determinant Compute the determinant given a lower triangular matrix Inputs: L : lower triangular matrix Outputs: det_L : determinant of L """ size_L = L.shape if np.size(L) == 1: return np.array(L) else: try: assert np.all(np.tril(L)==L) except AssertionError: print 'Error: Input is not a lower triangular matrix.' return 0 try: assert size_L[0] == size_L[1] except AssertionError: print 'Error: Not a square matrix.' return 0 det_L = 1. for i in xrange(size_L[1]): det_L = det_L*L[i,i] return det_L def optimize(t, f_0, d_f_0, f_t, d_f_t, t1=0.05, t2=0.5): """ Third order approximation to find the minimum of the function f : function to be optimized over Inputs : t : previous step size f_0 : f(0), function value at 0 d_f_0: f'(0), the derivative of the function at 0 f_t : f(t), function value at t d_f_t : f'(t), the derivative of the function at t t1 : step size reduction if minimum is at 0 or it can't be found t2 : step size reduction if minimum is at 1 Outputs : alpha_new : the new step size that minimizes the function """ if t <= 0 : print("Error: please enter non-negative t") return t a = (t*d_f_t-2*f_t+2*f_0+t*d_f_0)/(t**3) b = (f_t-f_0-t*d_f_0)/(t**2) c = d_f_0 A = 3*a B = b-t*a C = B**2-A*c if C == 0. : if c>0 or d_f_t>0 : t_new = t1*t else : t_new = t2*t elif A == 0 : t_new = -c/2./B elif C > 0 : t_new = (-B+np.sqrt(C))/A else : t_new = t1*t # check the bounds on new step size if t_new < t1*t : t_new = t1*t elif t_new > t2*t : t_new = t2*t return t_new class function(object): def __init__(self, f, args): """ Init Initialize the developer function class Inputs : f : user defined function --- Inputs of f : x : input value args : the arguments that the function takes Outputs of f : chi_x : boolean flag indicating whether the function is defined at x or not f_x : f(x), function value at x J_x : f'(x), the jacobian of the function evaluated at x Demo : chi_x, f_x, J_x = f(x,args) --- args : the arguments that the user defined function takes """ self._f = f self._args = args self._count = 0 def __call__(self, x): """ Call Calls the user defined function Inputs: x : input value Outputs: chi_x, f_x, J_x = f(x,args) """ self._count += 1 x = np.reshape(np.array(x), (-1)) chi_x, f_x, J_x = self._f(x, self.args) f_x = np.reshape(np.array(f_x), (-1,1)) return chi_x, f_x, np.array(J_x) def Jtest(self, x_min, x_max, dx=0.0002, N=1000, eps_max=0.0001, p=2, l_max=50, r=0.5): """ Gradient Checker Test the function's jacobian against the numerical jacobian Inputs : x_min : lower bound on the domain x_max : upper bound on the domain ** Warning ** x_min and x_max must be arrays of the same dimension Optional inputs : dx : (2*10^-4) the ratio of dx to the size of the domain N : (1000) number of test points eps_max : (10^-4) the maximum value error is allowed to be to confirm convergence p : (2) to specify the norm of the error (p-norm) l_max : (40) maximum number of tries to reduce dx r : (0.5) dx will be multiplied by this constant each step the error exceeds error_bound until l_max is reached ** Warning ** keep in mind machine precision when changing l_max and r Outputs : error : * 1 if did it did not pass the checks, * 0 if converged, * eps, the error of the numerical gradient point if no convergence """ x_min = np.reshape(np.array(x_min), (-1)) x_max = np.reshape(np.array(x_max), (-1)) # begin checks try: for i in xrange(np.size(x_min)): assert x_min[i] < x_max[i] except: print("Error: All values of x_min should be less than the " +"corresponding values for x_max.") exit(0) # end checks # begin test k = 0 while k < N : # reset dx each time D_x = (x_max - x_min) * dx # Sample random points in the state space x = np.random.uniform(x_min + D_x, x_max - D_x) # Compute jacobian at x l = 0 test_convergence = 1 while test_convergence: chi_x, f_x, J_x = self.__call__(x) D_f_x = J_x * 0. for j in xrange(np.size(x)): # calculate the derivative of each component of f d_x = D_x * 0. d_x[j] = D_x[j] chi_x_r, f_x_r, J_x_r = self.__call__(x + d_x) chi_x_l, f_x_l, J_x_l = self.__call__(x - d_x) # check if the function is defined on these points if( not(chi_x and chi_x_r and chi_x_l)): # discard this trial if one of the values is not defined test_convergence = 0 # break outer loop break d_f = (f_x_r - f_x_l) / (2. * d_x[j]) D_f_x[:,j] = d_f[:,0] eps = la.norm(D_f_x - J_x, p) / np.size(J_x) if (eps < eps_max): test_convergence = 0 # break outer loop k += 1 else: D_x = D_x * r if (l > l_max): # numerical gradient did not converge return eps l += 1 return 0 # end test """ Properties: 1. f 2. args 3. count """ @property def f(self): return self._f @property def args(self): return self._args @property def count(self): return self._count ## end function ##
mugurbil/gnm
gnm/utils.py
det
python
def det(L): size_L = L.shape if np.size(L) == 1: return np.array(L) else: try: assert np.all(np.tril(L)==L) except AssertionError: print 'Error: Input is not a lower triangular matrix.' return 0 try: assert size_L[0] == size_L[1] except AssertionError: print 'Error: Not a square matrix.' return 0 det_L = 1. for i in xrange(size_L[1]): det_L = det_L*L[i,i] return det_L
Determinant Compute the determinant given a lower triangular matrix Inputs: L : lower triangular matrix Outputs: det_L : determinant of L
train
https://github.com/mugurbil/gnm/blob/4f9711fb9d78cc02820c25234bc3ab9615014f11/gnm/utils.py#L74-L102
null
# -*- coding: utf-8 -*- """ Functions needed by the gnm sampler 1. test 2. update_params 3. log_K 4. multi_normal 5. det 6. optimize 7. function """ __all__ = ['test','update_params','log_K','multi_normal', 'det','optimize','function'] import numpy as np la = np.linalg def test(): import quickstart def update_params(state, t): """ Update parameters updates mean and precision to the step size Inputs: state : mu : mean L : cholesky factor of the precision matrix t : step size Outputs: mu : updated mean L : updated cholesky factor of the precision matrix """ mu = (1.-t)*state['x'] + t*state['mu'] L = state['L'] / np.sqrt(2.*t - t**2) return mu, L def log_K(Z, X, t): """ Log K Log of the proposal probability density function for gnm Inputs : Z : proposed to x : proposed from Outputs : log of the probability density function """ m, L = update_params(X, t) z = Z['x'] return np.log(det(L))-la.norm(np.dot(L.T,z-m))**2/2. def multi_normal(X, t): """ Multivariate normal sampler: Generates normal samples with mean m, precision matrix LL' Inputs: x : propose from Outputs: normal with mean m and precision LL' """ m, L = update_params(X, t) z = np.random.standard_normal(np.shape(m)) # generate i.i.d N(0,1) return la.solve(L.T,z)+m def optimize(t, f_0, d_f_0, f_t, d_f_t, t1=0.05, t2=0.5): """ Third order approximation to find the minimum of the function f : function to be optimized over Inputs : t : previous step size f_0 : f(0), function value at 0 d_f_0: f'(0), the derivative of the function at 0 f_t : f(t), function value at t d_f_t : f'(t), the derivative of the function at t t1 : step size reduction if minimum is at 0 or it can't be found t2 : step size reduction if minimum is at 1 Outputs : alpha_new : the new step size that minimizes the function """ if t <= 0 : print("Error: please enter non-negative t") return t a = (t*d_f_t-2*f_t+2*f_0+t*d_f_0)/(t**3) b = (f_t-f_0-t*d_f_0)/(t**2) c = d_f_0 A = 3*a B = b-t*a C = B**2-A*c if C == 0. : if c>0 or d_f_t>0 : t_new = t1*t else : t_new = t2*t elif A == 0 : t_new = -c/2./B elif C > 0 : t_new = (-B+np.sqrt(C))/A else : t_new = t1*t # check the bounds on new step size if t_new < t1*t : t_new = t1*t elif t_new > t2*t : t_new = t2*t return t_new class function(object): def __init__(self, f, args): """ Init Initialize the developer function class Inputs : f : user defined function --- Inputs of f : x : input value args : the arguments that the function takes Outputs of f : chi_x : boolean flag indicating whether the function is defined at x or not f_x : f(x), function value at x J_x : f'(x), the jacobian of the function evaluated at x Demo : chi_x, f_x, J_x = f(x,args) --- args : the arguments that the user defined function takes """ self._f = f self._args = args self._count = 0 def __call__(self, x): """ Call Calls the user defined function Inputs: x : input value Outputs: chi_x, f_x, J_x = f(x,args) """ self._count += 1 x = np.reshape(np.array(x), (-1)) chi_x, f_x, J_x = self._f(x, self.args) f_x = np.reshape(np.array(f_x), (-1,1)) return chi_x, f_x, np.array(J_x) def Jtest(self, x_min, x_max, dx=0.0002, N=1000, eps_max=0.0001, p=2, l_max=50, r=0.5): """ Gradient Checker Test the function's jacobian against the numerical jacobian Inputs : x_min : lower bound on the domain x_max : upper bound on the domain ** Warning ** x_min and x_max must be arrays of the same dimension Optional inputs : dx : (2*10^-4) the ratio of dx to the size of the domain N : (1000) number of test points eps_max : (10^-4) the maximum value error is allowed to be to confirm convergence p : (2) to specify the norm of the error (p-norm) l_max : (40) maximum number of tries to reduce dx r : (0.5) dx will be multiplied by this constant each step the error exceeds error_bound until l_max is reached ** Warning ** keep in mind machine precision when changing l_max and r Outputs : error : * 1 if did it did not pass the checks, * 0 if converged, * eps, the error of the numerical gradient point if no convergence """ x_min = np.reshape(np.array(x_min), (-1)) x_max = np.reshape(np.array(x_max), (-1)) # begin checks try: for i in xrange(np.size(x_min)): assert x_min[i] < x_max[i] except: print("Error: All values of x_min should be less than the " +"corresponding values for x_max.") exit(0) # end checks # begin test k = 0 while k < N : # reset dx each time D_x = (x_max - x_min) * dx # Sample random points in the state space x = np.random.uniform(x_min + D_x, x_max - D_x) # Compute jacobian at x l = 0 test_convergence = 1 while test_convergence: chi_x, f_x, J_x = self.__call__(x) D_f_x = J_x * 0. for j in xrange(np.size(x)): # calculate the derivative of each component of f d_x = D_x * 0. d_x[j] = D_x[j] chi_x_r, f_x_r, J_x_r = self.__call__(x + d_x) chi_x_l, f_x_l, J_x_l = self.__call__(x - d_x) # check if the function is defined on these points if( not(chi_x and chi_x_r and chi_x_l)): # discard this trial if one of the values is not defined test_convergence = 0 # break outer loop break d_f = (f_x_r - f_x_l) / (2. * d_x[j]) D_f_x[:,j] = d_f[:,0] eps = la.norm(D_f_x - J_x, p) / np.size(J_x) if (eps < eps_max): test_convergence = 0 # break outer loop k += 1 else: D_x = D_x * r if (l > l_max): # numerical gradient did not converge return eps l += 1 return 0 # end test """ Properties: 1. f 2. args 3. count """ @property def f(self): return self._f @property def args(self): return self._args @property def count(self): return self._count ## end function ##
mugurbil/gnm
gnm/utils.py
optimize
python
def optimize(t, f_0, d_f_0, f_t, d_f_t, t1=0.05, t2=0.5): if t <= 0 : print("Error: please enter non-negative t") return t a = (t*d_f_t-2*f_t+2*f_0+t*d_f_0)/(t**3) b = (f_t-f_0-t*d_f_0)/(t**2) c = d_f_0 A = 3*a B = b-t*a C = B**2-A*c if C == 0. : if c>0 or d_f_t>0 : t_new = t1*t else : t_new = t2*t elif A == 0 : t_new = -c/2./B elif C > 0 : t_new = (-B+np.sqrt(C))/A else : t_new = t1*t # check the bounds on new step size if t_new < t1*t : t_new = t1*t elif t_new > t2*t : t_new = t2*t return t_new
Third order approximation to find the minimum of the function f : function to be optimized over Inputs : t : previous step size f_0 : f(0), function value at 0 d_f_0: f'(0), the derivative of the function at 0 f_t : f(t), function value at t d_f_t : f'(t), the derivative of the function at t t1 : step size reduction if minimum is at 0 or it can't be found t2 : step size reduction if minimum is at 1 Outputs : alpha_new : the new step size that minimizes the function
train
https://github.com/mugurbil/gnm/blob/4f9711fb9d78cc02820c25234bc3ab9615014f11/gnm/utils.py#L104-L156
null
# -*- coding: utf-8 -*- """ Functions needed by the gnm sampler 1. test 2. update_params 3. log_K 4. multi_normal 5. det 6. optimize 7. function """ __all__ = ['test','update_params','log_K','multi_normal', 'det','optimize','function'] import numpy as np la = np.linalg def test(): import quickstart def update_params(state, t): """ Update parameters updates mean and precision to the step size Inputs: state : mu : mean L : cholesky factor of the precision matrix t : step size Outputs: mu : updated mean L : updated cholesky factor of the precision matrix """ mu = (1.-t)*state['x'] + t*state['mu'] L = state['L'] / np.sqrt(2.*t - t**2) return mu, L def log_K(Z, X, t): """ Log K Log of the proposal probability density function for gnm Inputs : Z : proposed to x : proposed from Outputs : log of the probability density function """ m, L = update_params(X, t) z = Z['x'] return np.log(det(L))-la.norm(np.dot(L.T,z-m))**2/2. def multi_normal(X, t): """ Multivariate normal sampler: Generates normal samples with mean m, precision matrix LL' Inputs: x : propose from Outputs: normal with mean m and precision LL' """ m, L = update_params(X, t) z = np.random.standard_normal(np.shape(m)) # generate i.i.d N(0,1) return la.solve(L.T,z)+m def det(L): """ Determinant Compute the determinant given a lower triangular matrix Inputs: L : lower triangular matrix Outputs: det_L : determinant of L """ size_L = L.shape if np.size(L) == 1: return np.array(L) else: try: assert np.all(np.tril(L)==L) except AssertionError: print 'Error: Input is not a lower triangular matrix.' return 0 try: assert size_L[0] == size_L[1] except AssertionError: print 'Error: Not a square matrix.' return 0 det_L = 1. for i in xrange(size_L[1]): det_L = det_L*L[i,i] return det_L class function(object): def __init__(self, f, args): """ Init Initialize the developer function class Inputs : f : user defined function --- Inputs of f : x : input value args : the arguments that the function takes Outputs of f : chi_x : boolean flag indicating whether the function is defined at x or not f_x : f(x), function value at x J_x : f'(x), the jacobian of the function evaluated at x Demo : chi_x, f_x, J_x = f(x,args) --- args : the arguments that the user defined function takes """ self._f = f self._args = args self._count = 0 def __call__(self, x): """ Call Calls the user defined function Inputs: x : input value Outputs: chi_x, f_x, J_x = f(x,args) """ self._count += 1 x = np.reshape(np.array(x), (-1)) chi_x, f_x, J_x = self._f(x, self.args) f_x = np.reshape(np.array(f_x), (-1,1)) return chi_x, f_x, np.array(J_x) def Jtest(self, x_min, x_max, dx=0.0002, N=1000, eps_max=0.0001, p=2, l_max=50, r=0.5): """ Gradient Checker Test the function's jacobian against the numerical jacobian Inputs : x_min : lower bound on the domain x_max : upper bound on the domain ** Warning ** x_min and x_max must be arrays of the same dimension Optional inputs : dx : (2*10^-4) the ratio of dx to the size of the domain N : (1000) number of test points eps_max : (10^-4) the maximum value error is allowed to be to confirm convergence p : (2) to specify the norm of the error (p-norm) l_max : (40) maximum number of tries to reduce dx r : (0.5) dx will be multiplied by this constant each step the error exceeds error_bound until l_max is reached ** Warning ** keep in mind machine precision when changing l_max and r Outputs : error : * 1 if did it did not pass the checks, * 0 if converged, * eps, the error of the numerical gradient point if no convergence """ x_min = np.reshape(np.array(x_min), (-1)) x_max = np.reshape(np.array(x_max), (-1)) # begin checks try: for i in xrange(np.size(x_min)): assert x_min[i] < x_max[i] except: print("Error: All values of x_min should be less than the " +"corresponding values for x_max.") exit(0) # end checks # begin test k = 0 while k < N : # reset dx each time D_x = (x_max - x_min) * dx # Sample random points in the state space x = np.random.uniform(x_min + D_x, x_max - D_x) # Compute jacobian at x l = 0 test_convergence = 1 while test_convergence: chi_x, f_x, J_x = self.__call__(x) D_f_x = J_x * 0. for j in xrange(np.size(x)): # calculate the derivative of each component of f d_x = D_x * 0. d_x[j] = D_x[j] chi_x_r, f_x_r, J_x_r = self.__call__(x + d_x) chi_x_l, f_x_l, J_x_l = self.__call__(x - d_x) # check if the function is defined on these points if( not(chi_x and chi_x_r and chi_x_l)): # discard this trial if one of the values is not defined test_convergence = 0 # break outer loop break d_f = (f_x_r - f_x_l) / (2. * d_x[j]) D_f_x[:,j] = d_f[:,0] eps = la.norm(D_f_x - J_x, p) / np.size(J_x) if (eps < eps_max): test_convergence = 0 # break outer loop k += 1 else: D_x = D_x * r if (l > l_max): # numerical gradient did not converge return eps l += 1 return 0 # end test """ Properties: 1. f 2. args 3. count """ @property def f(self): return self._f @property def args(self): return self._args @property def count(self): return self._count ## end function ##
mugurbil/gnm
gnm/gnm.py
sampler.prior
python
def prior(self, m, H): if self._prior == True: raise Warning("prior information is already set") else: self._prior = True # mean self._m = np.reshape(np.array(m), (-1,1)) try : assert np.size(self._m) == self._n except : raise TypeError("mean has to be an array of size n") # precision self._H = np.array(H) try : assert np.shape(self._H) == (self._n, self._n) except : raise TypeError("precision has to be a matrix of shape n by n") # precalculations self._ln_H_ = np.log(la.det(self._H))/2. self._Hm = np.dot(self._H, self._m)
Set prior Set prior values Inputs : m : mean of the prior H : precision matrix of the prior Hiddens : ln_H_ : log(det(H))/2 calculate this once to use everytime log prior is called Hm : < H, m > calculate this once to use everytime proposal is called
train
https://github.com/mugurbil/gnm/blob/4f9711fb9d78cc02820c25234bc3ab9615014f11/gnm/gnm.py#L93-L129
null
class sampler(object): def __init__(self, x, model, args): """ Init Initialize the GNM sampler class Inputs : x : initial guess model : user defined data model function args : arguments for the model """ self._args = args self._f = function(model, args) x = np.reshape(np.array(x), (-1,1)) self._n = np.size(x) # size of input space try: x = np.reshape(np.array(x), (-1)) out_x = model(x, self._args) except TypeError as e: raise TypeError(str(e)[:-7]+" needed)") """ except IndexError as e: raise IndexError("initial guess size does not fit model()\n " +str(e)) except Exception as e: print("Error: Model function could not be evaluated.") print(" - Check size of intial guess.") print(" - Check definition of the model.") print(str(e)) print(type(e)) raise RuntimeError("model() could not be evaluated\n ") """ try: chi_x, f_x, J_x = out_x f_x = np.reshape(np.array(f_x), (-1,1)) except: raise TypeError("model() needs to have 3 outputs: chi_x, f_x, J_x") try: assert chi_x == True except AssertionError: raise ValueError("initial guess out of range") try: self._mn = np.size(f_x) J_x = np.array(J_x) assert np.shape(J_x) == (self._mn, self._n) except: raise TypeError("Shape of Jacobian, " + str(np.shape(J_x)) + ", is not correct, (%d, %d)." % (self._mn, self._n)) x = np.reshape(np.array(x), (-1,1)) self._X = {'x':x,'f':f_x,'J':J_x} # state of x # prior parameters self._prior = False # back-off parameters self._max_steps = 1 self._step_size = 0.1 self._dynamic = False self._opts = {} # sampler outputs self._chain = None self._n_samples = 0 self._n_accepted = 0 def Jtest(self, x_min, x_max, dx=0.0002, N=1000, eps_max=0.0001, p=2, l_max=50, r=0.5): """ Gradient Checker Test the function's jacobian against the numerical jacobian """ # check inputs x_min and x_max try : assert np.size(x_min) == self._n except : raise TypeError("dimension of x_min, %d, does not match the " "dimension of input, %d" % (np.size(x_min), self._n)) try : assert np.size(x_max) == self._n except : raise TypeError("dimension of x_max, %d, does not match the " "dimension of input, %d." % (np.size(x_max), self._n)) # end checks and call developer function return self._f.Jtest(x_min, x_max, dx=dx, N=N, eps_max=eps_max, p=p, l_max=l_max, r=r) def static(self, max_steps, step_size): """ Set Back-off to Static Set the sampler parameters for static back off Inputs : max_steps : maximum optimization steps to be taken step_size : the step size of the back-off """ self._dynamic = False # begin checks try : self._max_steps = int(max_steps) except : print("Error: Input 1 (max_steps) has to be an int.") return 0 try : assert self._max_steps >= 0 except : print("Warning: Input 1 (max_steps) has to be non-negative.") print("Setting max_steps to 0.") self._max_steps = 0 if max_steps > 0 : try : assert step_size == float(step_size) except AssertionError : print("Warning: Input 2 (step_size) is not a float. Converted.") step_size = float(step_size) except : print("Error: Input 2 (step_size) has to be a float.") return 0 try : assert 0. < step_size < 1. except : print("Warning: Input 2 (step_size) has to be between 0 and 1.") print("Setting step_size to 0.2.") step_size = 0.2 self._step_size = step_size if step_size**max_steps < 10**(-15): print("Warning: Back-off gets dangerously small.") # end checks def dynamic(self, max_steps, opts={}): """ Dynamic Switch Set the sampler parameters for dynamic back off Inputs : max_steps : maximum back-off steps to be taken Optional Inputs: opts : ({}) dictionary containing fancy options """ self._dynamic = True # begin checks try : self._max_steps = int(max_steps) except : raise TypeError("input 1 (max_steps) has to be an integer") return 0 try : assert self._max_steps >= 0 except : raise Warning("input 1 (max_steps) has to be non-negative. Setting (max_steps) to 0.") self._max_steps = 0 self._opts = opts # end checks def sample(self, n_samples, divs=1, visual=False, safe=False): """ Sample Sampling Inputs : n_samples : number of samples to generate Optional Inputs : divs : (1) number of divisions visual : show progress safe : save the chain at every division """ if visual: print("Sampling: 0%") for i in xrange(divs): self._sample(int(n_samples/divs)) if visual: sys.stdout.write("\033[F") # curser up print("Sampling: "+str(int(i*100./divs)+1)+'%') if safe: self.save(path="chain_{:}.dat".format(i)) if n_samples % divs != 0: self._sample(n_samples % divs) if safe: self.save(path="chain_{:}.dat".format(divs)) def save(self, path="chain.dat"): """ Save Save data to file Inputs : path : specifies the path name of the file to be loaded to """ # create dictionary for data dic = {} dic['chain'] = self._chain.tolist() dic['step_count'] = self._step_count.tolist() dic['n_samples'] = self._n_samples dic['n_accepted'] = self._n_accepted dic['x'] = self._X['x'].tolist() dic['f'] = self._X['f'].tolist() dic['J'] = self._X['J'].tolist() # write data to file file = open(path, 'w') json.dump(dic, file) file.close() def load(self, path="chain.dat"): """ Load Load data from file Inputs : path : specifies the path name of the file to be loaded from """ # read data from file file = open(path, 'r') dic = json.load(file) file.close() # get data from dictionary self._chain = np.array(dic['chain']) self._step_count = np.array(dic['step_count']) self._n_samples = dic['n_samples'] self._n_accepted = dic['n_accepted'] self._X = {} self._X['x'] = dic['x'] self._X['f'] = dic['f'] self._X['J'] = dic['J'] def burn(self, n_burned): """ Burn Burn the inital samples to adjust for convergence of the chain cut the first (n_burned) burn-in samples Inputs : chain : the full Markov chain n_burned : number of samples to cut Hidden Outputs : chain : chain with the firt n_burned samples cut """ self._chain = self._chain[n_burned:] def acor(self, k = 5): """ Autocorrelation time of the chain return the autocorrelation time for each parameters Inputs : k : parameter in self-consistent window Outputs : t : autocorrelation time of the chain """ try: import acor except ImportError: print("Can't import acor, please download it.") return 0 n = np.shape(self._chain)[1] t = np.zeros(n) for i in xrange(n): t[i] = acor.acor(self._chain[:,i],k)[0] return t def posterior(self, x): """ Posterior density ** not normalized ** This is used to plot the theoretical curve for tests. Inputs : x : input value Outputs : p : p(x)=pi(x)*exp{-||f(x)||^2/(2)} posterior probability of x """ x = np.reshape(np.array(x), (-1,1)) chi_x, f_x, J_x = self._f(x) if chi_x : p = np.exp(-la.norm(f_x)**2/2.) if self._prior: m = self._m H = self._H p = p * np.exp(-np.dot((x-m).T,np.dot(H,x-m))/2.) return p else : return 0 def error_bars(self, n_bins, d_min, d_max): """ Error Bars create bars and error bars to plot Inputs : n_bins : number of bins plot_range : (shape) = (number of dimensions, 2) matrix which contain the min and max for each dimension as rows Outputs : x : domain p_x : estimated posterior using the chain on the domain error : estimated error for p_x """ # fetch data chain = self._chain len_chain = len(chain) try: n_dims = np.shape(chain)[1] except: n_dims = 1 # begin checks try: assert n_bins == int(n_bins) except: raise TypeError("number of bins has to be an integer") d_min = np.reshape(np.array(d_min), (-1,1)) d_max = np.reshape(np.array(d_max), (-1,1)) try: assert np.size(d_min) == n_dims except: raise TypeError("domain minimum has wrong size") try: assert np.size(d_max) == n_dims except: raise TypeError("domain maximum has wrong size") # end checks # initialize outputs p_x = np.zeros(n_bins) # esitmate of posterior error = np.zeros(n_bins) # error bars x = np.zeros((n_dims, n_bins)) # centers of bins # set dx v = d_max-d_min v_2 = np.dot(v.T, v)[0][0] # bin count for i in xrange(len_chain): bin_no = int(np.floor(np.dot(chain[i].T-d_min,v)/v_2*n_bins)[0]) if n_bins > bin_no > -1: p_x[bin_no] += 1. # end count dx = np.sqrt(v_2)/n_bins p_x = p_x/(len_chain*dx) # find error for i in xrange(n_bins): p = p_x[i] error[i] = np.sqrt(p*(1./dx-p)/(len_chain)) x[:,i] = (d_min+v*(0.5+i)/n_bins)[0] # end find return x, p_x, error # end error_bars # internal methods def _sample(self, n_samples): """ Sample Generate samples for posterior distribution using Gauss-Newton proposal parameters Inputs : n_samples : number of samples to generate Hidden Outputs : chain : chain of samples n_samples : length of chain n_accepted : number of proposals accepted step_count : count of the steps accepted """ try : n_samples = int(n_samples) except : raise TypeError("number of samples has to be an integer") exit() # fetch info X = self._proposal_params(self._X) k_max = self._max_steps # initialize chain = np.zeros((n_samples, self._n)) n_accepted = 0 step_count = np.zeros(k_max+2) # begin outer loop for i in xrange(n_samples): accepted = False # check if sample is accepted r_ = [1] # list of step sizes Z_ = [X] # initialize list of Z s self._r_ = r_ log_P_z_x = 0. + X['log_p'] k = 0 # back-off steps taken so far while k <= k_max: # get proposal chi_z = False while not chi_z: z = multi_normal(X, r_[-1]) chi_z, f_z, J_z = self._f(z) Z = self._proposal_params({'x':z,'f':f_z,'J':J_z}) Z_.append(Z) self._Z_ = Z_ log_P_z_x += log_K(Z, X, r_[-1]) # N is the Numerator of the acceptance, N = P_x_z self._N_is_0 = False # check to see if N = 0, to use in _log_P log_N = self._log_P(X, Z, k) # calculating acceptance probability if self._N_is_0 == True : A_z_x = 0. elif log_N >= log_P_z_x : A_z_x = 1. else : A_z_x = np.exp(log_N - log_P_z_x) # acceptance rejection if np.random.rand() <= A_z_x: accepted = True break else : log_P_z_x += np.log(1. - A_z_x) self._back_off() k += 1 # end of steps for loop if accepted == True : chain[i,:] = z[:,0] X = Z # for statistics n_accepted += 1 step_count[k+1] += 1 else : chain[i,:] = X['x'][:,0] # for statistics step_count[0] += 1 # end outer loop # update stored info self._X = X # outputs if self._n_samples == 0 : self._chain = chain self._step_count = step_count else : self._chain = np.append(self._chain, chain, axis=0) self._step_count = np.add(self._step_count, step_count) self._n_samples += n_samples self._n_accepted += n_accepted # end sample def _proposal_params(self, state): """ Proposal parameters Calculate parameters needed for the proposal. Inputs : state : x : the present sample, the place to linearize around f : f(x), function value at x J : f'(x), the jacobian of the function evaluated at x Outputs : state : mu : the mean vector L : the lower triangular cholesky factor of P log_p : log(p(x)) log of the posterior density """ x = state['x'] f = state['f'] J = state['J'] JJ = np.dot(J.T,J) if self._prior: m = self._m H = self._H Hm = self._Hm # LL' = P = H+J'J L = la.cholesky(H+JJ) # mu = (P^-1)(Hm-J'f+J'Jx) mu = la.solve(L.T,la.solve(L,Hm-np.dot(J.T,f)+np.dot(JJ,x))) else: # P = J'J L = la.cholesky(JJ) # mu = x-(P^-1)J'f mu = x-la.solve(L.T,la.solve(L,np.dot(J.T,f))) state['L'] = L state['mu'] = mu state['log_p'] = self._log_post(x,f) return state def _log_P(self, X , Z, k): """ Log of the probability of transition from z to x with k steps log ( P_k (x, z) ) Inputs : X : state to be proposed to Z : state to be proposed from k : number of recursions, depth """ r_ = self._r_ Z_ = self._Z_ # zero case if k == 0 : log_P = Z['log_p'] + log_K(X, Z, r_[k]) # recursice case else : P_zk_z = np.exp( self._log_P(Z_[k], Z, k-1) ) P_z_zk = np.exp( self._log_P(Z, Z_[k], k-1) ) # flag if P_zk_z <= P_z_zk : self._N_is_0 = True log_P = -np.inf else : log_P = np.log( P_zk_z - P_z_zk ) + log_K(X, Z, r_[k]) return log_P def _back_off(self): """ Back off Calculate the back off step size Inputs : Z_ : list of states in current proposal r_ : list of back offs in current proposal q : step size reduction dynamic : set to True if you want to use the dynamic back-off Outputs : """ q = self._step_size r = self._r_[-1] Z_ = self._Z_ if self._dynamic: p_0 = la.norm(Z_[0]['f']) dp_0 = p_0*2*la.norm(Z_[0]['J']) p_r = la.norm(Z_[-1]['f']) dp_r = p_0*2*la.norm(Z_[-1]['J']) r_new = optimize(r, p_0**2, dp_0, p_r**2, dp_r) else : r_new = r * q self._r_.append(r_new) def _log_post(self,x,f_x): """ Log of the posterior density This is used to calculete acceptance probability for sampling. Inputs : x : input value f_x : f(x), function value at x Outputs : log(p_x) : log[pi(x)]-||f(x)||^2/(2) log of the posterior probability """ # least squares part -||f(x)||^2/2 log_likelihood = (-la.norm(f_x)**2)/(2.) # prior part -(x-m)'H(x-m)/2 if self._prior: m = self._m H = self._H log_prior = self._ln_H_-np.dot((x-m).T,np.dot(H,x-m))/2. return log_prior+log_likelihood else: return log_likelihood """ Properties: 1. chain 2. n_samples 3. n_accepted 4. accept_rate 5. step_count 6. call_count 7. max_steps 8. step_size """ @property def chain(self): return self._chain @property def n_samples(self): return self._n_samples @property def n_accepted(self): return self._n_accepted @property def accept_rate(self): return float(self._n_accepted)/self._n_samples @property def step_count(self): return self._step_count @property def call_count(self): return self._f.count @property def max_steps(self): return self._max_steps @property def step_size(self): return self._step_size
mugurbil/gnm
gnm/gnm.py
sampler.static
python
def static(self, max_steps, step_size): self._dynamic = False # begin checks try : self._max_steps = int(max_steps) except : print("Error: Input 1 (max_steps) has to be an int.") return 0 try : assert self._max_steps >= 0 except : print("Warning: Input 1 (max_steps) has to be non-negative.") print("Setting max_steps to 0.") self._max_steps = 0 if max_steps > 0 : try : assert step_size == float(step_size) except AssertionError : print("Warning: Input 2 (step_size) is not a float. Converted.") step_size = float(step_size) except : print("Error: Input 2 (step_size) has to be a float.") return 0 try : assert 0. < step_size < 1. except : print("Warning: Input 2 (step_size) has to be between 0 and 1.") print("Setting step_size to 0.2.") step_size = 0.2 self._step_size = step_size if step_size**max_steps < 10**(-15): print("Warning: Back-off gets dangerously small.")
Set Back-off to Static Set the sampler parameters for static back off Inputs : max_steps : maximum optimization steps to be taken step_size : the step size of the back-off
train
https://github.com/mugurbil/gnm/blob/4f9711fb9d78cc02820c25234bc3ab9615014f11/gnm/gnm.py#L152-L194
null
class sampler(object): def __init__(self, x, model, args): """ Init Initialize the GNM sampler class Inputs : x : initial guess model : user defined data model function args : arguments for the model """ self._args = args self._f = function(model, args) x = np.reshape(np.array(x), (-1,1)) self._n = np.size(x) # size of input space try: x = np.reshape(np.array(x), (-1)) out_x = model(x, self._args) except TypeError as e: raise TypeError(str(e)[:-7]+" needed)") """ except IndexError as e: raise IndexError("initial guess size does not fit model()\n " +str(e)) except Exception as e: print("Error: Model function could not be evaluated.") print(" - Check size of intial guess.") print(" - Check definition of the model.") print(str(e)) print(type(e)) raise RuntimeError("model() could not be evaluated\n ") """ try: chi_x, f_x, J_x = out_x f_x = np.reshape(np.array(f_x), (-1,1)) except: raise TypeError("model() needs to have 3 outputs: chi_x, f_x, J_x") try: assert chi_x == True except AssertionError: raise ValueError("initial guess out of range") try: self._mn = np.size(f_x) J_x = np.array(J_x) assert np.shape(J_x) == (self._mn, self._n) except: raise TypeError("Shape of Jacobian, " + str(np.shape(J_x)) + ", is not correct, (%d, %d)." % (self._mn, self._n)) x = np.reshape(np.array(x), (-1,1)) self._X = {'x':x,'f':f_x,'J':J_x} # state of x # prior parameters self._prior = False # back-off parameters self._max_steps = 1 self._step_size = 0.1 self._dynamic = False self._opts = {} # sampler outputs self._chain = None self._n_samples = 0 self._n_accepted = 0 def prior(self, m, H): """ Set prior Set prior values Inputs : m : mean of the prior H : precision matrix of the prior Hiddens : ln_H_ : log(det(H))/2 calculate this once to use everytime log prior is called Hm : < H, m > calculate this once to use everytime proposal is called """ if self._prior == True: raise Warning("prior information is already set") else: self._prior = True # mean self._m = np.reshape(np.array(m), (-1,1)) try : assert np.size(self._m) == self._n except : raise TypeError("mean has to be an array of size n") # precision self._H = np.array(H) try : assert np.shape(self._H) == (self._n, self._n) except : raise TypeError("precision has to be a matrix of shape n by n") # precalculations self._ln_H_ = np.log(la.det(self._H))/2. self._Hm = np.dot(self._H, self._m) def Jtest(self, x_min, x_max, dx=0.0002, N=1000, eps_max=0.0001, p=2, l_max=50, r=0.5): """ Gradient Checker Test the function's jacobian against the numerical jacobian """ # check inputs x_min and x_max try : assert np.size(x_min) == self._n except : raise TypeError("dimension of x_min, %d, does not match the " "dimension of input, %d" % (np.size(x_min), self._n)) try : assert np.size(x_max) == self._n except : raise TypeError("dimension of x_max, %d, does not match the " "dimension of input, %d." % (np.size(x_max), self._n)) # end checks and call developer function return self._f.Jtest(x_min, x_max, dx=dx, N=N, eps_max=eps_max, p=p, l_max=l_max, r=r) # end checks def dynamic(self, max_steps, opts={}): """ Dynamic Switch Set the sampler parameters for dynamic back off Inputs : max_steps : maximum back-off steps to be taken Optional Inputs: opts : ({}) dictionary containing fancy options """ self._dynamic = True # begin checks try : self._max_steps = int(max_steps) except : raise TypeError("input 1 (max_steps) has to be an integer") return 0 try : assert self._max_steps >= 0 except : raise Warning("input 1 (max_steps) has to be non-negative. Setting (max_steps) to 0.") self._max_steps = 0 self._opts = opts # end checks def sample(self, n_samples, divs=1, visual=False, safe=False): """ Sample Sampling Inputs : n_samples : number of samples to generate Optional Inputs : divs : (1) number of divisions visual : show progress safe : save the chain at every division """ if visual: print("Sampling: 0%") for i in xrange(divs): self._sample(int(n_samples/divs)) if visual: sys.stdout.write("\033[F") # curser up print("Sampling: "+str(int(i*100./divs)+1)+'%') if safe: self.save(path="chain_{:}.dat".format(i)) if n_samples % divs != 0: self._sample(n_samples % divs) if safe: self.save(path="chain_{:}.dat".format(divs)) def save(self, path="chain.dat"): """ Save Save data to file Inputs : path : specifies the path name of the file to be loaded to """ # create dictionary for data dic = {} dic['chain'] = self._chain.tolist() dic['step_count'] = self._step_count.tolist() dic['n_samples'] = self._n_samples dic['n_accepted'] = self._n_accepted dic['x'] = self._X['x'].tolist() dic['f'] = self._X['f'].tolist() dic['J'] = self._X['J'].tolist() # write data to file file = open(path, 'w') json.dump(dic, file) file.close() def load(self, path="chain.dat"): """ Load Load data from file Inputs : path : specifies the path name of the file to be loaded from """ # read data from file file = open(path, 'r') dic = json.load(file) file.close() # get data from dictionary self._chain = np.array(dic['chain']) self._step_count = np.array(dic['step_count']) self._n_samples = dic['n_samples'] self._n_accepted = dic['n_accepted'] self._X = {} self._X['x'] = dic['x'] self._X['f'] = dic['f'] self._X['J'] = dic['J'] def burn(self, n_burned): """ Burn Burn the inital samples to adjust for convergence of the chain cut the first (n_burned) burn-in samples Inputs : chain : the full Markov chain n_burned : number of samples to cut Hidden Outputs : chain : chain with the firt n_burned samples cut """ self._chain = self._chain[n_burned:] def acor(self, k = 5): """ Autocorrelation time of the chain return the autocorrelation time for each parameters Inputs : k : parameter in self-consistent window Outputs : t : autocorrelation time of the chain """ try: import acor except ImportError: print("Can't import acor, please download it.") return 0 n = np.shape(self._chain)[1] t = np.zeros(n) for i in xrange(n): t[i] = acor.acor(self._chain[:,i],k)[0] return t def posterior(self, x): """ Posterior density ** not normalized ** This is used to plot the theoretical curve for tests. Inputs : x : input value Outputs : p : p(x)=pi(x)*exp{-||f(x)||^2/(2)} posterior probability of x """ x = np.reshape(np.array(x), (-1,1)) chi_x, f_x, J_x = self._f(x) if chi_x : p = np.exp(-la.norm(f_x)**2/2.) if self._prior: m = self._m H = self._H p = p * np.exp(-np.dot((x-m).T,np.dot(H,x-m))/2.) return p else : return 0 def error_bars(self, n_bins, d_min, d_max): """ Error Bars create bars and error bars to plot Inputs : n_bins : number of bins plot_range : (shape) = (number of dimensions, 2) matrix which contain the min and max for each dimension as rows Outputs : x : domain p_x : estimated posterior using the chain on the domain error : estimated error for p_x """ # fetch data chain = self._chain len_chain = len(chain) try: n_dims = np.shape(chain)[1] except: n_dims = 1 # begin checks try: assert n_bins == int(n_bins) except: raise TypeError("number of bins has to be an integer") d_min = np.reshape(np.array(d_min), (-1,1)) d_max = np.reshape(np.array(d_max), (-1,1)) try: assert np.size(d_min) == n_dims except: raise TypeError("domain minimum has wrong size") try: assert np.size(d_max) == n_dims except: raise TypeError("domain maximum has wrong size") # end checks # initialize outputs p_x = np.zeros(n_bins) # esitmate of posterior error = np.zeros(n_bins) # error bars x = np.zeros((n_dims, n_bins)) # centers of bins # set dx v = d_max-d_min v_2 = np.dot(v.T, v)[0][0] # bin count for i in xrange(len_chain): bin_no = int(np.floor(np.dot(chain[i].T-d_min,v)/v_2*n_bins)[0]) if n_bins > bin_no > -1: p_x[bin_no] += 1. # end count dx = np.sqrt(v_2)/n_bins p_x = p_x/(len_chain*dx) # find error for i in xrange(n_bins): p = p_x[i] error[i] = np.sqrt(p*(1./dx-p)/(len_chain)) x[:,i] = (d_min+v*(0.5+i)/n_bins)[0] # end find return x, p_x, error # end error_bars # internal methods def _sample(self, n_samples): """ Sample Generate samples for posterior distribution using Gauss-Newton proposal parameters Inputs : n_samples : number of samples to generate Hidden Outputs : chain : chain of samples n_samples : length of chain n_accepted : number of proposals accepted step_count : count of the steps accepted """ try : n_samples = int(n_samples) except : raise TypeError("number of samples has to be an integer") exit() # fetch info X = self._proposal_params(self._X) k_max = self._max_steps # initialize chain = np.zeros((n_samples, self._n)) n_accepted = 0 step_count = np.zeros(k_max+2) # begin outer loop for i in xrange(n_samples): accepted = False # check if sample is accepted r_ = [1] # list of step sizes Z_ = [X] # initialize list of Z s self._r_ = r_ log_P_z_x = 0. + X['log_p'] k = 0 # back-off steps taken so far while k <= k_max: # get proposal chi_z = False while not chi_z: z = multi_normal(X, r_[-1]) chi_z, f_z, J_z = self._f(z) Z = self._proposal_params({'x':z,'f':f_z,'J':J_z}) Z_.append(Z) self._Z_ = Z_ log_P_z_x += log_K(Z, X, r_[-1]) # N is the Numerator of the acceptance, N = P_x_z self._N_is_0 = False # check to see if N = 0, to use in _log_P log_N = self._log_P(X, Z, k) # calculating acceptance probability if self._N_is_0 == True : A_z_x = 0. elif log_N >= log_P_z_x : A_z_x = 1. else : A_z_x = np.exp(log_N - log_P_z_x) # acceptance rejection if np.random.rand() <= A_z_x: accepted = True break else : log_P_z_x += np.log(1. - A_z_x) self._back_off() k += 1 # end of steps for loop if accepted == True : chain[i,:] = z[:,0] X = Z # for statistics n_accepted += 1 step_count[k+1] += 1 else : chain[i,:] = X['x'][:,0] # for statistics step_count[0] += 1 # end outer loop # update stored info self._X = X # outputs if self._n_samples == 0 : self._chain = chain self._step_count = step_count else : self._chain = np.append(self._chain, chain, axis=0) self._step_count = np.add(self._step_count, step_count) self._n_samples += n_samples self._n_accepted += n_accepted # end sample def _proposal_params(self, state): """ Proposal parameters Calculate parameters needed for the proposal. Inputs : state : x : the present sample, the place to linearize around f : f(x), function value at x J : f'(x), the jacobian of the function evaluated at x Outputs : state : mu : the mean vector L : the lower triangular cholesky factor of P log_p : log(p(x)) log of the posterior density """ x = state['x'] f = state['f'] J = state['J'] JJ = np.dot(J.T,J) if self._prior: m = self._m H = self._H Hm = self._Hm # LL' = P = H+J'J L = la.cholesky(H+JJ) # mu = (P^-1)(Hm-J'f+J'Jx) mu = la.solve(L.T,la.solve(L,Hm-np.dot(J.T,f)+np.dot(JJ,x))) else: # P = J'J L = la.cholesky(JJ) # mu = x-(P^-1)J'f mu = x-la.solve(L.T,la.solve(L,np.dot(J.T,f))) state['L'] = L state['mu'] = mu state['log_p'] = self._log_post(x,f) return state def _log_P(self, X , Z, k): """ Log of the probability of transition from z to x with k steps log ( P_k (x, z) ) Inputs : X : state to be proposed to Z : state to be proposed from k : number of recursions, depth """ r_ = self._r_ Z_ = self._Z_ # zero case if k == 0 : log_P = Z['log_p'] + log_K(X, Z, r_[k]) # recursice case else : P_zk_z = np.exp( self._log_P(Z_[k], Z, k-1) ) P_z_zk = np.exp( self._log_P(Z, Z_[k], k-1) ) # flag if P_zk_z <= P_z_zk : self._N_is_0 = True log_P = -np.inf else : log_P = np.log( P_zk_z - P_z_zk ) + log_K(X, Z, r_[k]) return log_P def _back_off(self): """ Back off Calculate the back off step size Inputs : Z_ : list of states in current proposal r_ : list of back offs in current proposal q : step size reduction dynamic : set to True if you want to use the dynamic back-off Outputs : """ q = self._step_size r = self._r_[-1] Z_ = self._Z_ if self._dynamic: p_0 = la.norm(Z_[0]['f']) dp_0 = p_0*2*la.norm(Z_[0]['J']) p_r = la.norm(Z_[-1]['f']) dp_r = p_0*2*la.norm(Z_[-1]['J']) r_new = optimize(r, p_0**2, dp_0, p_r**2, dp_r) else : r_new = r * q self._r_.append(r_new) def _log_post(self,x,f_x): """ Log of the posterior density This is used to calculete acceptance probability for sampling. Inputs : x : input value f_x : f(x), function value at x Outputs : log(p_x) : log[pi(x)]-||f(x)||^2/(2) log of the posterior probability """ # least squares part -||f(x)||^2/2 log_likelihood = (-la.norm(f_x)**2)/(2.) # prior part -(x-m)'H(x-m)/2 if self._prior: m = self._m H = self._H log_prior = self._ln_H_-np.dot((x-m).T,np.dot(H,x-m))/2. return log_prior+log_likelihood else: return log_likelihood """ Properties: 1. chain 2. n_samples 3. n_accepted 4. accept_rate 5. step_count 6. call_count 7. max_steps 8. step_size """ @property def chain(self): return self._chain @property def n_samples(self): return self._n_samples @property def n_accepted(self): return self._n_accepted @property def accept_rate(self): return float(self._n_accepted)/self._n_samples @property def step_count(self): return self._step_count @property def call_count(self): return self._f.count @property def max_steps(self): return self._max_steps @property def step_size(self): return self._step_size
mugurbil/gnm
gnm/gnm.py
sampler.dynamic
python
def dynamic(self, max_steps, opts={}): self._dynamic = True # begin checks try : self._max_steps = int(max_steps) except : raise TypeError("input 1 (max_steps) has to be an integer") return 0 try : assert self._max_steps >= 0 except : raise Warning("input 1 (max_steps) has to be non-negative. Setting (max_steps) to 0.") self._max_steps = 0 self._opts = opts
Dynamic Switch Set the sampler parameters for dynamic back off Inputs : max_steps : maximum back-off steps to be taken Optional Inputs: opts : ({}) dictionary containing fancy options
train
https://github.com/mugurbil/gnm/blob/4f9711fb9d78cc02820c25234bc3ab9615014f11/gnm/gnm.py#L197-L222
null
class sampler(object): def __init__(self, x, model, args): """ Init Initialize the GNM sampler class Inputs : x : initial guess model : user defined data model function args : arguments for the model """ self._args = args self._f = function(model, args) x = np.reshape(np.array(x), (-1,1)) self._n = np.size(x) # size of input space try: x = np.reshape(np.array(x), (-1)) out_x = model(x, self._args) except TypeError as e: raise TypeError(str(e)[:-7]+" needed)") """ except IndexError as e: raise IndexError("initial guess size does not fit model()\n " +str(e)) except Exception as e: print("Error: Model function could not be evaluated.") print(" - Check size of intial guess.") print(" - Check definition of the model.") print(str(e)) print(type(e)) raise RuntimeError("model() could not be evaluated\n ") """ try: chi_x, f_x, J_x = out_x f_x = np.reshape(np.array(f_x), (-1,1)) except: raise TypeError("model() needs to have 3 outputs: chi_x, f_x, J_x") try: assert chi_x == True except AssertionError: raise ValueError("initial guess out of range") try: self._mn = np.size(f_x) J_x = np.array(J_x) assert np.shape(J_x) == (self._mn, self._n) except: raise TypeError("Shape of Jacobian, " + str(np.shape(J_x)) + ", is not correct, (%d, %d)." % (self._mn, self._n)) x = np.reshape(np.array(x), (-1,1)) self._X = {'x':x,'f':f_x,'J':J_x} # state of x # prior parameters self._prior = False # back-off parameters self._max_steps = 1 self._step_size = 0.1 self._dynamic = False self._opts = {} # sampler outputs self._chain = None self._n_samples = 0 self._n_accepted = 0 def prior(self, m, H): """ Set prior Set prior values Inputs : m : mean of the prior H : precision matrix of the prior Hiddens : ln_H_ : log(det(H))/2 calculate this once to use everytime log prior is called Hm : < H, m > calculate this once to use everytime proposal is called """ if self._prior == True: raise Warning("prior information is already set") else: self._prior = True # mean self._m = np.reshape(np.array(m), (-1,1)) try : assert np.size(self._m) == self._n except : raise TypeError("mean has to be an array of size n") # precision self._H = np.array(H) try : assert np.shape(self._H) == (self._n, self._n) except : raise TypeError("precision has to be a matrix of shape n by n") # precalculations self._ln_H_ = np.log(la.det(self._H))/2. self._Hm = np.dot(self._H, self._m) def Jtest(self, x_min, x_max, dx=0.0002, N=1000, eps_max=0.0001, p=2, l_max=50, r=0.5): """ Gradient Checker Test the function's jacobian against the numerical jacobian """ # check inputs x_min and x_max try : assert np.size(x_min) == self._n except : raise TypeError("dimension of x_min, %d, does not match the " "dimension of input, %d" % (np.size(x_min), self._n)) try : assert np.size(x_max) == self._n except : raise TypeError("dimension of x_max, %d, does not match the " "dimension of input, %d." % (np.size(x_max), self._n)) # end checks and call developer function return self._f.Jtest(x_min, x_max, dx=dx, N=N, eps_max=eps_max, p=p, l_max=l_max, r=r) def static(self, max_steps, step_size): """ Set Back-off to Static Set the sampler parameters for static back off Inputs : max_steps : maximum optimization steps to be taken step_size : the step size of the back-off """ self._dynamic = False # begin checks try : self._max_steps = int(max_steps) except : print("Error: Input 1 (max_steps) has to be an int.") return 0 try : assert self._max_steps >= 0 except : print("Warning: Input 1 (max_steps) has to be non-negative.") print("Setting max_steps to 0.") self._max_steps = 0 if max_steps > 0 : try : assert step_size == float(step_size) except AssertionError : print("Warning: Input 2 (step_size) is not a float. Converted.") step_size = float(step_size) except : print("Error: Input 2 (step_size) has to be a float.") return 0 try : assert 0. < step_size < 1. except : print("Warning: Input 2 (step_size) has to be between 0 and 1.") print("Setting step_size to 0.2.") step_size = 0.2 self._step_size = step_size if step_size**max_steps < 10**(-15): print("Warning: Back-off gets dangerously small.") # end checks # end checks def sample(self, n_samples, divs=1, visual=False, safe=False): """ Sample Sampling Inputs : n_samples : number of samples to generate Optional Inputs : divs : (1) number of divisions visual : show progress safe : save the chain at every division """ if visual: print("Sampling: 0%") for i in xrange(divs): self._sample(int(n_samples/divs)) if visual: sys.stdout.write("\033[F") # curser up print("Sampling: "+str(int(i*100./divs)+1)+'%') if safe: self.save(path="chain_{:}.dat".format(i)) if n_samples % divs != 0: self._sample(n_samples % divs) if safe: self.save(path="chain_{:}.dat".format(divs)) def save(self, path="chain.dat"): """ Save Save data to file Inputs : path : specifies the path name of the file to be loaded to """ # create dictionary for data dic = {} dic['chain'] = self._chain.tolist() dic['step_count'] = self._step_count.tolist() dic['n_samples'] = self._n_samples dic['n_accepted'] = self._n_accepted dic['x'] = self._X['x'].tolist() dic['f'] = self._X['f'].tolist() dic['J'] = self._X['J'].tolist() # write data to file file = open(path, 'w') json.dump(dic, file) file.close() def load(self, path="chain.dat"): """ Load Load data from file Inputs : path : specifies the path name of the file to be loaded from """ # read data from file file = open(path, 'r') dic = json.load(file) file.close() # get data from dictionary self._chain = np.array(dic['chain']) self._step_count = np.array(dic['step_count']) self._n_samples = dic['n_samples'] self._n_accepted = dic['n_accepted'] self._X = {} self._X['x'] = dic['x'] self._X['f'] = dic['f'] self._X['J'] = dic['J'] def burn(self, n_burned): """ Burn Burn the inital samples to adjust for convergence of the chain cut the first (n_burned) burn-in samples Inputs : chain : the full Markov chain n_burned : number of samples to cut Hidden Outputs : chain : chain with the firt n_burned samples cut """ self._chain = self._chain[n_burned:] def acor(self, k = 5): """ Autocorrelation time of the chain return the autocorrelation time for each parameters Inputs : k : parameter in self-consistent window Outputs : t : autocorrelation time of the chain """ try: import acor except ImportError: print("Can't import acor, please download it.") return 0 n = np.shape(self._chain)[1] t = np.zeros(n) for i in xrange(n): t[i] = acor.acor(self._chain[:,i],k)[0] return t def posterior(self, x): """ Posterior density ** not normalized ** This is used to plot the theoretical curve for tests. Inputs : x : input value Outputs : p : p(x)=pi(x)*exp{-||f(x)||^2/(2)} posterior probability of x """ x = np.reshape(np.array(x), (-1,1)) chi_x, f_x, J_x = self._f(x) if chi_x : p = np.exp(-la.norm(f_x)**2/2.) if self._prior: m = self._m H = self._H p = p * np.exp(-np.dot((x-m).T,np.dot(H,x-m))/2.) return p else : return 0 def error_bars(self, n_bins, d_min, d_max): """ Error Bars create bars and error bars to plot Inputs : n_bins : number of bins plot_range : (shape) = (number of dimensions, 2) matrix which contain the min and max for each dimension as rows Outputs : x : domain p_x : estimated posterior using the chain on the domain error : estimated error for p_x """ # fetch data chain = self._chain len_chain = len(chain) try: n_dims = np.shape(chain)[1] except: n_dims = 1 # begin checks try: assert n_bins == int(n_bins) except: raise TypeError("number of bins has to be an integer") d_min = np.reshape(np.array(d_min), (-1,1)) d_max = np.reshape(np.array(d_max), (-1,1)) try: assert np.size(d_min) == n_dims except: raise TypeError("domain minimum has wrong size") try: assert np.size(d_max) == n_dims except: raise TypeError("domain maximum has wrong size") # end checks # initialize outputs p_x = np.zeros(n_bins) # esitmate of posterior error = np.zeros(n_bins) # error bars x = np.zeros((n_dims, n_bins)) # centers of bins # set dx v = d_max-d_min v_2 = np.dot(v.T, v)[0][0] # bin count for i in xrange(len_chain): bin_no = int(np.floor(np.dot(chain[i].T-d_min,v)/v_2*n_bins)[0]) if n_bins > bin_no > -1: p_x[bin_no] += 1. # end count dx = np.sqrt(v_2)/n_bins p_x = p_x/(len_chain*dx) # find error for i in xrange(n_bins): p = p_x[i] error[i] = np.sqrt(p*(1./dx-p)/(len_chain)) x[:,i] = (d_min+v*(0.5+i)/n_bins)[0] # end find return x, p_x, error # end error_bars # internal methods def _sample(self, n_samples): """ Sample Generate samples for posterior distribution using Gauss-Newton proposal parameters Inputs : n_samples : number of samples to generate Hidden Outputs : chain : chain of samples n_samples : length of chain n_accepted : number of proposals accepted step_count : count of the steps accepted """ try : n_samples = int(n_samples) except : raise TypeError("number of samples has to be an integer") exit() # fetch info X = self._proposal_params(self._X) k_max = self._max_steps # initialize chain = np.zeros((n_samples, self._n)) n_accepted = 0 step_count = np.zeros(k_max+2) # begin outer loop for i in xrange(n_samples): accepted = False # check if sample is accepted r_ = [1] # list of step sizes Z_ = [X] # initialize list of Z s self._r_ = r_ log_P_z_x = 0. + X['log_p'] k = 0 # back-off steps taken so far while k <= k_max: # get proposal chi_z = False while not chi_z: z = multi_normal(X, r_[-1]) chi_z, f_z, J_z = self._f(z) Z = self._proposal_params({'x':z,'f':f_z,'J':J_z}) Z_.append(Z) self._Z_ = Z_ log_P_z_x += log_K(Z, X, r_[-1]) # N is the Numerator of the acceptance, N = P_x_z self._N_is_0 = False # check to see if N = 0, to use in _log_P log_N = self._log_P(X, Z, k) # calculating acceptance probability if self._N_is_0 == True : A_z_x = 0. elif log_N >= log_P_z_x : A_z_x = 1. else : A_z_x = np.exp(log_N - log_P_z_x) # acceptance rejection if np.random.rand() <= A_z_x: accepted = True break else : log_P_z_x += np.log(1. - A_z_x) self._back_off() k += 1 # end of steps for loop if accepted == True : chain[i,:] = z[:,0] X = Z # for statistics n_accepted += 1 step_count[k+1] += 1 else : chain[i,:] = X['x'][:,0] # for statistics step_count[0] += 1 # end outer loop # update stored info self._X = X # outputs if self._n_samples == 0 : self._chain = chain self._step_count = step_count else : self._chain = np.append(self._chain, chain, axis=0) self._step_count = np.add(self._step_count, step_count) self._n_samples += n_samples self._n_accepted += n_accepted # end sample def _proposal_params(self, state): """ Proposal parameters Calculate parameters needed for the proposal. Inputs : state : x : the present sample, the place to linearize around f : f(x), function value at x J : f'(x), the jacobian of the function evaluated at x Outputs : state : mu : the mean vector L : the lower triangular cholesky factor of P log_p : log(p(x)) log of the posterior density """ x = state['x'] f = state['f'] J = state['J'] JJ = np.dot(J.T,J) if self._prior: m = self._m H = self._H Hm = self._Hm # LL' = P = H+J'J L = la.cholesky(H+JJ) # mu = (P^-1)(Hm-J'f+J'Jx) mu = la.solve(L.T,la.solve(L,Hm-np.dot(J.T,f)+np.dot(JJ,x))) else: # P = J'J L = la.cholesky(JJ) # mu = x-(P^-1)J'f mu = x-la.solve(L.T,la.solve(L,np.dot(J.T,f))) state['L'] = L state['mu'] = mu state['log_p'] = self._log_post(x,f) return state def _log_P(self, X , Z, k): """ Log of the probability of transition from z to x with k steps log ( P_k (x, z) ) Inputs : X : state to be proposed to Z : state to be proposed from k : number of recursions, depth """ r_ = self._r_ Z_ = self._Z_ # zero case if k == 0 : log_P = Z['log_p'] + log_K(X, Z, r_[k]) # recursice case else : P_zk_z = np.exp( self._log_P(Z_[k], Z, k-1) ) P_z_zk = np.exp( self._log_P(Z, Z_[k], k-1) ) # flag if P_zk_z <= P_z_zk : self._N_is_0 = True log_P = -np.inf else : log_P = np.log( P_zk_z - P_z_zk ) + log_K(X, Z, r_[k]) return log_P def _back_off(self): """ Back off Calculate the back off step size Inputs : Z_ : list of states in current proposal r_ : list of back offs in current proposal q : step size reduction dynamic : set to True if you want to use the dynamic back-off Outputs : """ q = self._step_size r = self._r_[-1] Z_ = self._Z_ if self._dynamic: p_0 = la.norm(Z_[0]['f']) dp_0 = p_0*2*la.norm(Z_[0]['J']) p_r = la.norm(Z_[-1]['f']) dp_r = p_0*2*la.norm(Z_[-1]['J']) r_new = optimize(r, p_0**2, dp_0, p_r**2, dp_r) else : r_new = r * q self._r_.append(r_new) def _log_post(self,x,f_x): """ Log of the posterior density This is used to calculete acceptance probability for sampling. Inputs : x : input value f_x : f(x), function value at x Outputs : log(p_x) : log[pi(x)]-||f(x)||^2/(2) log of the posterior probability """ # least squares part -||f(x)||^2/2 log_likelihood = (-la.norm(f_x)**2)/(2.) # prior part -(x-m)'H(x-m)/2 if self._prior: m = self._m H = self._H log_prior = self._ln_H_-np.dot((x-m).T,np.dot(H,x-m))/2. return log_prior+log_likelihood else: return log_likelihood """ Properties: 1. chain 2. n_samples 3. n_accepted 4. accept_rate 5. step_count 6. call_count 7. max_steps 8. step_size """ @property def chain(self): return self._chain @property def n_samples(self): return self._n_samples @property def n_accepted(self): return self._n_accepted @property def accept_rate(self): return float(self._n_accepted)/self._n_samples @property def step_count(self): return self._step_count @property def call_count(self): return self._f.count @property def max_steps(self): return self._max_steps @property def step_size(self): return self._step_size
mugurbil/gnm
gnm/gnm.py
sampler.sample
python
def sample(self, n_samples, divs=1, visual=False, safe=False): if visual: print("Sampling: 0%") for i in xrange(divs): self._sample(int(n_samples/divs)) if visual: sys.stdout.write("\033[F") # curser up print("Sampling: "+str(int(i*100./divs)+1)+'%') if safe: self.save(path="chain_{:}.dat".format(i)) if n_samples % divs != 0: self._sample(n_samples % divs) if safe: self.save(path="chain_{:}.dat".format(divs))
Sample Sampling Inputs : n_samples : number of samples to generate Optional Inputs : divs : (1) number of divisions visual : show progress safe : save the chain at every division
train
https://github.com/mugurbil/gnm/blob/4f9711fb9d78cc02820c25234bc3ab9615014f11/gnm/gnm.py#L225-L252
[ "def save(self, path=\"chain.dat\"):\n \"\"\"\nSave\n Save data to file\nInputs : \n path :\n specifies the path name of the file to be loaded to\n \"\"\"\n # create dictionary for data\n dic = {}\n dic['chain'] = self._chain.tolist()\n dic['step_count'] = self._step_count.tolist()\n dic['n_samples'] = self._n_samples\n dic['n_accepted'] = self._n_accepted\n dic['x'] = self._X['x'].tolist()\n dic['f'] = self._X['f'].tolist()\n dic['J'] = self._X['J'].tolist()\n\n # write data to file\n file = open(path, 'w')\n json.dump(dic, file)\n file.close()\n", "def _sample(self, n_samples):\n \"\"\"\nSample\n Generate samples for posterior distribution using Gauss-Newton \n proposal parameters\nInputs : \n n_samples :\n number of samples to generate\nHidden Outputs :\n chain :\n chain of samples\n n_samples :\n length of chain\n n_accepted :\n number of proposals accepted\n step_count :\n count of the steps accepted\n \"\"\"\n try : \n n_samples = int(n_samples)\n except :\n raise TypeError(\"number of samples has to be an integer\")\n exit()\n\n # fetch info\n X = self._proposal_params(self._X)\n k_max = self._max_steps\n\n # initialize \n chain = np.zeros((n_samples, self._n)) \n n_accepted = 0\n step_count = np.zeros(k_max+2)\n\n # begin outer loop\n for i in xrange(n_samples):\n accepted = False # check if sample is accepted\n r_ = [1] # list of step sizes\n Z_ = [X] # initialize list of Z s\n self._r_ = r_ \n log_P_z_x = 0. + X['log_p'] \n\n k = 0 # back-off steps taken so far\n while k <= k_max:\n # get proposal\n chi_z = False\n while not chi_z:\n z = multi_normal(X, r_[-1])\n chi_z, f_z, J_z = self._f(z)\n Z = self._proposal_params({'x':z,'f':f_z,'J':J_z})\n Z_.append(Z)\n self._Z_ = Z_\n\n log_P_z_x += log_K(Z, X, r_[-1])\n\n # N is the Numerator of the acceptance, N = P_x_z\n self._N_is_0 = False # check to see if N = 0, to use in _log_P\n log_N = self._log_P(X, Z, k)\n\n # calculating acceptance probability\n if self._N_is_0 == True :\n A_z_x = 0.\n elif log_N >= log_P_z_x :\n A_z_x = 1.\n else :\n A_z_x = np.exp(log_N - log_P_z_x)\n\n # acceptance rejection\n if np.random.rand() <= A_z_x:\n accepted = True\n break\n else : \n log_P_z_x += np.log(1. - A_z_x)\n self._back_off()\n k += 1 \n # end of steps for loop\n if accepted == True :\n chain[i,:] = z[:,0] \n X = Z\n # for statistics\n n_accepted += 1 \n step_count[k+1] += 1\n else :\n chain[i,:] = X['x'][:,0]\n # for statistics\n step_count[0] += 1\n # end outer loop\n\n # update stored info\n self._X = X\n\n # outputs\n if self._n_samples == 0 :\n self._chain = chain\n self._step_count = step_count\n else :\n self._chain = np.append(self._chain, chain, axis=0)\n self._step_count = np.add(self._step_count, step_count)\n self._n_samples += n_samples\n self._n_accepted += n_accepted\n" ]
class sampler(object): def __init__(self, x, model, args): """ Init Initialize the GNM sampler class Inputs : x : initial guess model : user defined data model function args : arguments for the model """ self._args = args self._f = function(model, args) x = np.reshape(np.array(x), (-1,1)) self._n = np.size(x) # size of input space try: x = np.reshape(np.array(x), (-1)) out_x = model(x, self._args) except TypeError as e: raise TypeError(str(e)[:-7]+" needed)") """ except IndexError as e: raise IndexError("initial guess size does not fit model()\n " +str(e)) except Exception as e: print("Error: Model function could not be evaluated.") print(" - Check size of intial guess.") print(" - Check definition of the model.") print(str(e)) print(type(e)) raise RuntimeError("model() could not be evaluated\n ") """ try: chi_x, f_x, J_x = out_x f_x = np.reshape(np.array(f_x), (-1,1)) except: raise TypeError("model() needs to have 3 outputs: chi_x, f_x, J_x") try: assert chi_x == True except AssertionError: raise ValueError("initial guess out of range") try: self._mn = np.size(f_x) J_x = np.array(J_x) assert np.shape(J_x) == (self._mn, self._n) except: raise TypeError("Shape of Jacobian, " + str(np.shape(J_x)) + ", is not correct, (%d, %d)." % (self._mn, self._n)) x = np.reshape(np.array(x), (-1,1)) self._X = {'x':x,'f':f_x,'J':J_x} # state of x # prior parameters self._prior = False # back-off parameters self._max_steps = 1 self._step_size = 0.1 self._dynamic = False self._opts = {} # sampler outputs self._chain = None self._n_samples = 0 self._n_accepted = 0 def prior(self, m, H): """ Set prior Set prior values Inputs : m : mean of the prior H : precision matrix of the prior Hiddens : ln_H_ : log(det(H))/2 calculate this once to use everytime log prior is called Hm : < H, m > calculate this once to use everytime proposal is called """ if self._prior == True: raise Warning("prior information is already set") else: self._prior = True # mean self._m = np.reshape(np.array(m), (-1,1)) try : assert np.size(self._m) == self._n except : raise TypeError("mean has to be an array of size n") # precision self._H = np.array(H) try : assert np.shape(self._H) == (self._n, self._n) except : raise TypeError("precision has to be a matrix of shape n by n") # precalculations self._ln_H_ = np.log(la.det(self._H))/2. self._Hm = np.dot(self._H, self._m) def Jtest(self, x_min, x_max, dx=0.0002, N=1000, eps_max=0.0001, p=2, l_max=50, r=0.5): """ Gradient Checker Test the function's jacobian against the numerical jacobian """ # check inputs x_min and x_max try : assert np.size(x_min) == self._n except : raise TypeError("dimension of x_min, %d, does not match the " "dimension of input, %d" % (np.size(x_min), self._n)) try : assert np.size(x_max) == self._n except : raise TypeError("dimension of x_max, %d, does not match the " "dimension of input, %d." % (np.size(x_max), self._n)) # end checks and call developer function return self._f.Jtest(x_min, x_max, dx=dx, N=N, eps_max=eps_max, p=p, l_max=l_max, r=r) def static(self, max_steps, step_size): """ Set Back-off to Static Set the sampler parameters for static back off Inputs : max_steps : maximum optimization steps to be taken step_size : the step size of the back-off """ self._dynamic = False # begin checks try : self._max_steps = int(max_steps) except : print("Error: Input 1 (max_steps) has to be an int.") return 0 try : assert self._max_steps >= 0 except : print("Warning: Input 1 (max_steps) has to be non-negative.") print("Setting max_steps to 0.") self._max_steps = 0 if max_steps > 0 : try : assert step_size == float(step_size) except AssertionError : print("Warning: Input 2 (step_size) is not a float. Converted.") step_size = float(step_size) except : print("Error: Input 2 (step_size) has to be a float.") return 0 try : assert 0. < step_size < 1. except : print("Warning: Input 2 (step_size) has to be between 0 and 1.") print("Setting step_size to 0.2.") step_size = 0.2 self._step_size = step_size if step_size**max_steps < 10**(-15): print("Warning: Back-off gets dangerously small.") # end checks def dynamic(self, max_steps, opts={}): """ Dynamic Switch Set the sampler parameters for dynamic back off Inputs : max_steps : maximum back-off steps to be taken Optional Inputs: opts : ({}) dictionary containing fancy options """ self._dynamic = True # begin checks try : self._max_steps = int(max_steps) except : raise TypeError("input 1 (max_steps) has to be an integer") return 0 try : assert self._max_steps >= 0 except : raise Warning("input 1 (max_steps) has to be non-negative. Setting (max_steps) to 0.") self._max_steps = 0 self._opts = opts # end checks def save(self, path="chain.dat"): """ Save Save data to file Inputs : path : specifies the path name of the file to be loaded to """ # create dictionary for data dic = {} dic['chain'] = self._chain.tolist() dic['step_count'] = self._step_count.tolist() dic['n_samples'] = self._n_samples dic['n_accepted'] = self._n_accepted dic['x'] = self._X['x'].tolist() dic['f'] = self._X['f'].tolist() dic['J'] = self._X['J'].tolist() # write data to file file = open(path, 'w') json.dump(dic, file) file.close() def load(self, path="chain.dat"): """ Load Load data from file Inputs : path : specifies the path name of the file to be loaded from """ # read data from file file = open(path, 'r') dic = json.load(file) file.close() # get data from dictionary self._chain = np.array(dic['chain']) self._step_count = np.array(dic['step_count']) self._n_samples = dic['n_samples'] self._n_accepted = dic['n_accepted'] self._X = {} self._X['x'] = dic['x'] self._X['f'] = dic['f'] self._X['J'] = dic['J'] def burn(self, n_burned): """ Burn Burn the inital samples to adjust for convergence of the chain cut the first (n_burned) burn-in samples Inputs : chain : the full Markov chain n_burned : number of samples to cut Hidden Outputs : chain : chain with the firt n_burned samples cut """ self._chain = self._chain[n_burned:] def acor(self, k = 5): """ Autocorrelation time of the chain return the autocorrelation time for each parameters Inputs : k : parameter in self-consistent window Outputs : t : autocorrelation time of the chain """ try: import acor except ImportError: print("Can't import acor, please download it.") return 0 n = np.shape(self._chain)[1] t = np.zeros(n) for i in xrange(n): t[i] = acor.acor(self._chain[:,i],k)[0] return t def posterior(self, x): """ Posterior density ** not normalized ** This is used to plot the theoretical curve for tests. Inputs : x : input value Outputs : p : p(x)=pi(x)*exp{-||f(x)||^2/(2)} posterior probability of x """ x = np.reshape(np.array(x), (-1,1)) chi_x, f_x, J_x = self._f(x) if chi_x : p = np.exp(-la.norm(f_x)**2/2.) if self._prior: m = self._m H = self._H p = p * np.exp(-np.dot((x-m).T,np.dot(H,x-m))/2.) return p else : return 0 def error_bars(self, n_bins, d_min, d_max): """ Error Bars create bars and error bars to plot Inputs : n_bins : number of bins plot_range : (shape) = (number of dimensions, 2) matrix which contain the min and max for each dimension as rows Outputs : x : domain p_x : estimated posterior using the chain on the domain error : estimated error for p_x """ # fetch data chain = self._chain len_chain = len(chain) try: n_dims = np.shape(chain)[1] except: n_dims = 1 # begin checks try: assert n_bins == int(n_bins) except: raise TypeError("number of bins has to be an integer") d_min = np.reshape(np.array(d_min), (-1,1)) d_max = np.reshape(np.array(d_max), (-1,1)) try: assert np.size(d_min) == n_dims except: raise TypeError("domain minimum has wrong size") try: assert np.size(d_max) == n_dims except: raise TypeError("domain maximum has wrong size") # end checks # initialize outputs p_x = np.zeros(n_bins) # esitmate of posterior error = np.zeros(n_bins) # error bars x = np.zeros((n_dims, n_bins)) # centers of bins # set dx v = d_max-d_min v_2 = np.dot(v.T, v)[0][0] # bin count for i in xrange(len_chain): bin_no = int(np.floor(np.dot(chain[i].T-d_min,v)/v_2*n_bins)[0]) if n_bins > bin_no > -1: p_x[bin_no] += 1. # end count dx = np.sqrt(v_2)/n_bins p_x = p_x/(len_chain*dx) # find error for i in xrange(n_bins): p = p_x[i] error[i] = np.sqrt(p*(1./dx-p)/(len_chain)) x[:,i] = (d_min+v*(0.5+i)/n_bins)[0] # end find return x, p_x, error # end error_bars # internal methods def _sample(self, n_samples): """ Sample Generate samples for posterior distribution using Gauss-Newton proposal parameters Inputs : n_samples : number of samples to generate Hidden Outputs : chain : chain of samples n_samples : length of chain n_accepted : number of proposals accepted step_count : count of the steps accepted """ try : n_samples = int(n_samples) except : raise TypeError("number of samples has to be an integer") exit() # fetch info X = self._proposal_params(self._X) k_max = self._max_steps # initialize chain = np.zeros((n_samples, self._n)) n_accepted = 0 step_count = np.zeros(k_max+2) # begin outer loop for i in xrange(n_samples): accepted = False # check if sample is accepted r_ = [1] # list of step sizes Z_ = [X] # initialize list of Z s self._r_ = r_ log_P_z_x = 0. + X['log_p'] k = 0 # back-off steps taken so far while k <= k_max: # get proposal chi_z = False while not chi_z: z = multi_normal(X, r_[-1]) chi_z, f_z, J_z = self._f(z) Z = self._proposal_params({'x':z,'f':f_z,'J':J_z}) Z_.append(Z) self._Z_ = Z_ log_P_z_x += log_K(Z, X, r_[-1]) # N is the Numerator of the acceptance, N = P_x_z self._N_is_0 = False # check to see if N = 0, to use in _log_P log_N = self._log_P(X, Z, k) # calculating acceptance probability if self._N_is_0 == True : A_z_x = 0. elif log_N >= log_P_z_x : A_z_x = 1. else : A_z_x = np.exp(log_N - log_P_z_x) # acceptance rejection if np.random.rand() <= A_z_x: accepted = True break else : log_P_z_x += np.log(1. - A_z_x) self._back_off() k += 1 # end of steps for loop if accepted == True : chain[i,:] = z[:,0] X = Z # for statistics n_accepted += 1 step_count[k+1] += 1 else : chain[i,:] = X['x'][:,0] # for statistics step_count[0] += 1 # end outer loop # update stored info self._X = X # outputs if self._n_samples == 0 : self._chain = chain self._step_count = step_count else : self._chain = np.append(self._chain, chain, axis=0) self._step_count = np.add(self._step_count, step_count) self._n_samples += n_samples self._n_accepted += n_accepted # end sample def _proposal_params(self, state): """ Proposal parameters Calculate parameters needed for the proposal. Inputs : state : x : the present sample, the place to linearize around f : f(x), function value at x J : f'(x), the jacobian of the function evaluated at x Outputs : state : mu : the mean vector L : the lower triangular cholesky factor of P log_p : log(p(x)) log of the posterior density """ x = state['x'] f = state['f'] J = state['J'] JJ = np.dot(J.T,J) if self._prior: m = self._m H = self._H Hm = self._Hm # LL' = P = H+J'J L = la.cholesky(H+JJ) # mu = (P^-1)(Hm-J'f+J'Jx) mu = la.solve(L.T,la.solve(L,Hm-np.dot(J.T,f)+np.dot(JJ,x))) else: # P = J'J L = la.cholesky(JJ) # mu = x-(P^-1)J'f mu = x-la.solve(L.T,la.solve(L,np.dot(J.T,f))) state['L'] = L state['mu'] = mu state['log_p'] = self._log_post(x,f) return state def _log_P(self, X , Z, k): """ Log of the probability of transition from z to x with k steps log ( P_k (x, z) ) Inputs : X : state to be proposed to Z : state to be proposed from k : number of recursions, depth """ r_ = self._r_ Z_ = self._Z_ # zero case if k == 0 : log_P = Z['log_p'] + log_K(X, Z, r_[k]) # recursice case else : P_zk_z = np.exp( self._log_P(Z_[k], Z, k-1) ) P_z_zk = np.exp( self._log_P(Z, Z_[k], k-1) ) # flag if P_zk_z <= P_z_zk : self._N_is_0 = True log_P = -np.inf else : log_P = np.log( P_zk_z - P_z_zk ) + log_K(X, Z, r_[k]) return log_P def _back_off(self): """ Back off Calculate the back off step size Inputs : Z_ : list of states in current proposal r_ : list of back offs in current proposal q : step size reduction dynamic : set to True if you want to use the dynamic back-off Outputs : """ q = self._step_size r = self._r_[-1] Z_ = self._Z_ if self._dynamic: p_0 = la.norm(Z_[0]['f']) dp_0 = p_0*2*la.norm(Z_[0]['J']) p_r = la.norm(Z_[-1]['f']) dp_r = p_0*2*la.norm(Z_[-1]['J']) r_new = optimize(r, p_0**2, dp_0, p_r**2, dp_r) else : r_new = r * q self._r_.append(r_new) def _log_post(self,x,f_x): """ Log of the posterior density This is used to calculete acceptance probability for sampling. Inputs : x : input value f_x : f(x), function value at x Outputs : log(p_x) : log[pi(x)]-||f(x)||^2/(2) log of the posterior probability """ # least squares part -||f(x)||^2/2 log_likelihood = (-la.norm(f_x)**2)/(2.) # prior part -(x-m)'H(x-m)/2 if self._prior: m = self._m H = self._H log_prior = self._ln_H_-np.dot((x-m).T,np.dot(H,x-m))/2. return log_prior+log_likelihood else: return log_likelihood """ Properties: 1. chain 2. n_samples 3. n_accepted 4. accept_rate 5. step_count 6. call_count 7. max_steps 8. step_size """ @property def chain(self): return self._chain @property def n_samples(self): return self._n_samples @property def n_accepted(self): return self._n_accepted @property def accept_rate(self): return float(self._n_accepted)/self._n_samples @property def step_count(self): return self._step_count @property def call_count(self): return self._f.count @property def max_steps(self): return self._max_steps @property def step_size(self): return self._step_size
mugurbil/gnm
gnm/gnm.py
sampler.save
python
def save(self, path="chain.dat"): # create dictionary for data dic = {} dic['chain'] = self._chain.tolist() dic['step_count'] = self._step_count.tolist() dic['n_samples'] = self._n_samples dic['n_accepted'] = self._n_accepted dic['x'] = self._X['x'].tolist() dic['f'] = self._X['f'].tolist() dic['J'] = self._X['J'].tolist() # write data to file file = open(path, 'w') json.dump(dic, file) file.close()
Save Save data to file Inputs : path : specifies the path name of the file to be loaded to
train
https://github.com/mugurbil/gnm/blob/4f9711fb9d78cc02820c25234bc3ab9615014f11/gnm/gnm.py#L254-L275
null
class sampler(object): def __init__(self, x, model, args): """ Init Initialize the GNM sampler class Inputs : x : initial guess model : user defined data model function args : arguments for the model """ self._args = args self._f = function(model, args) x = np.reshape(np.array(x), (-1,1)) self._n = np.size(x) # size of input space try: x = np.reshape(np.array(x), (-1)) out_x = model(x, self._args) except TypeError as e: raise TypeError(str(e)[:-7]+" needed)") """ except IndexError as e: raise IndexError("initial guess size does not fit model()\n " +str(e)) except Exception as e: print("Error: Model function could not be evaluated.") print(" - Check size of intial guess.") print(" - Check definition of the model.") print(str(e)) print(type(e)) raise RuntimeError("model() could not be evaluated\n ") """ try: chi_x, f_x, J_x = out_x f_x = np.reshape(np.array(f_x), (-1,1)) except: raise TypeError("model() needs to have 3 outputs: chi_x, f_x, J_x") try: assert chi_x == True except AssertionError: raise ValueError("initial guess out of range") try: self._mn = np.size(f_x) J_x = np.array(J_x) assert np.shape(J_x) == (self._mn, self._n) except: raise TypeError("Shape of Jacobian, " + str(np.shape(J_x)) + ", is not correct, (%d, %d)." % (self._mn, self._n)) x = np.reshape(np.array(x), (-1,1)) self._X = {'x':x,'f':f_x,'J':J_x} # state of x # prior parameters self._prior = False # back-off parameters self._max_steps = 1 self._step_size = 0.1 self._dynamic = False self._opts = {} # sampler outputs self._chain = None self._n_samples = 0 self._n_accepted = 0 def prior(self, m, H): """ Set prior Set prior values Inputs : m : mean of the prior H : precision matrix of the prior Hiddens : ln_H_ : log(det(H))/2 calculate this once to use everytime log prior is called Hm : < H, m > calculate this once to use everytime proposal is called """ if self._prior == True: raise Warning("prior information is already set") else: self._prior = True # mean self._m = np.reshape(np.array(m), (-1,1)) try : assert np.size(self._m) == self._n except : raise TypeError("mean has to be an array of size n") # precision self._H = np.array(H) try : assert np.shape(self._H) == (self._n, self._n) except : raise TypeError("precision has to be a matrix of shape n by n") # precalculations self._ln_H_ = np.log(la.det(self._H))/2. self._Hm = np.dot(self._H, self._m) def Jtest(self, x_min, x_max, dx=0.0002, N=1000, eps_max=0.0001, p=2, l_max=50, r=0.5): """ Gradient Checker Test the function's jacobian against the numerical jacobian """ # check inputs x_min and x_max try : assert np.size(x_min) == self._n except : raise TypeError("dimension of x_min, %d, does not match the " "dimension of input, %d" % (np.size(x_min), self._n)) try : assert np.size(x_max) == self._n except : raise TypeError("dimension of x_max, %d, does not match the " "dimension of input, %d." % (np.size(x_max), self._n)) # end checks and call developer function return self._f.Jtest(x_min, x_max, dx=dx, N=N, eps_max=eps_max, p=p, l_max=l_max, r=r) def static(self, max_steps, step_size): """ Set Back-off to Static Set the sampler parameters for static back off Inputs : max_steps : maximum optimization steps to be taken step_size : the step size of the back-off """ self._dynamic = False # begin checks try : self._max_steps = int(max_steps) except : print("Error: Input 1 (max_steps) has to be an int.") return 0 try : assert self._max_steps >= 0 except : print("Warning: Input 1 (max_steps) has to be non-negative.") print("Setting max_steps to 0.") self._max_steps = 0 if max_steps > 0 : try : assert step_size == float(step_size) except AssertionError : print("Warning: Input 2 (step_size) is not a float. Converted.") step_size = float(step_size) except : print("Error: Input 2 (step_size) has to be a float.") return 0 try : assert 0. < step_size < 1. except : print("Warning: Input 2 (step_size) has to be between 0 and 1.") print("Setting step_size to 0.2.") step_size = 0.2 self._step_size = step_size if step_size**max_steps < 10**(-15): print("Warning: Back-off gets dangerously small.") # end checks def dynamic(self, max_steps, opts={}): """ Dynamic Switch Set the sampler parameters for dynamic back off Inputs : max_steps : maximum back-off steps to be taken Optional Inputs: opts : ({}) dictionary containing fancy options """ self._dynamic = True # begin checks try : self._max_steps = int(max_steps) except : raise TypeError("input 1 (max_steps) has to be an integer") return 0 try : assert self._max_steps >= 0 except : raise Warning("input 1 (max_steps) has to be non-negative. Setting (max_steps) to 0.") self._max_steps = 0 self._opts = opts # end checks def sample(self, n_samples, divs=1, visual=False, safe=False): """ Sample Sampling Inputs : n_samples : number of samples to generate Optional Inputs : divs : (1) number of divisions visual : show progress safe : save the chain at every division """ if visual: print("Sampling: 0%") for i in xrange(divs): self._sample(int(n_samples/divs)) if visual: sys.stdout.write("\033[F") # curser up print("Sampling: "+str(int(i*100./divs)+1)+'%') if safe: self.save(path="chain_{:}.dat".format(i)) if n_samples % divs != 0: self._sample(n_samples % divs) if safe: self.save(path="chain_{:}.dat".format(divs)) def load(self, path="chain.dat"): """ Load Load data from file Inputs : path : specifies the path name of the file to be loaded from """ # read data from file file = open(path, 'r') dic = json.load(file) file.close() # get data from dictionary self._chain = np.array(dic['chain']) self._step_count = np.array(dic['step_count']) self._n_samples = dic['n_samples'] self._n_accepted = dic['n_accepted'] self._X = {} self._X['x'] = dic['x'] self._X['f'] = dic['f'] self._X['J'] = dic['J'] def burn(self, n_burned): """ Burn Burn the inital samples to adjust for convergence of the chain cut the first (n_burned) burn-in samples Inputs : chain : the full Markov chain n_burned : number of samples to cut Hidden Outputs : chain : chain with the firt n_burned samples cut """ self._chain = self._chain[n_burned:] def acor(self, k = 5): """ Autocorrelation time of the chain return the autocorrelation time for each parameters Inputs : k : parameter in self-consistent window Outputs : t : autocorrelation time of the chain """ try: import acor except ImportError: print("Can't import acor, please download it.") return 0 n = np.shape(self._chain)[1] t = np.zeros(n) for i in xrange(n): t[i] = acor.acor(self._chain[:,i],k)[0] return t def posterior(self, x): """ Posterior density ** not normalized ** This is used to plot the theoretical curve for tests. Inputs : x : input value Outputs : p : p(x)=pi(x)*exp{-||f(x)||^2/(2)} posterior probability of x """ x = np.reshape(np.array(x), (-1,1)) chi_x, f_x, J_x = self._f(x) if chi_x : p = np.exp(-la.norm(f_x)**2/2.) if self._prior: m = self._m H = self._H p = p * np.exp(-np.dot((x-m).T,np.dot(H,x-m))/2.) return p else : return 0 def error_bars(self, n_bins, d_min, d_max): """ Error Bars create bars and error bars to plot Inputs : n_bins : number of bins plot_range : (shape) = (number of dimensions, 2) matrix which contain the min and max for each dimension as rows Outputs : x : domain p_x : estimated posterior using the chain on the domain error : estimated error for p_x """ # fetch data chain = self._chain len_chain = len(chain) try: n_dims = np.shape(chain)[1] except: n_dims = 1 # begin checks try: assert n_bins == int(n_bins) except: raise TypeError("number of bins has to be an integer") d_min = np.reshape(np.array(d_min), (-1,1)) d_max = np.reshape(np.array(d_max), (-1,1)) try: assert np.size(d_min) == n_dims except: raise TypeError("domain minimum has wrong size") try: assert np.size(d_max) == n_dims except: raise TypeError("domain maximum has wrong size") # end checks # initialize outputs p_x = np.zeros(n_bins) # esitmate of posterior error = np.zeros(n_bins) # error bars x = np.zeros((n_dims, n_bins)) # centers of bins # set dx v = d_max-d_min v_2 = np.dot(v.T, v)[0][0] # bin count for i in xrange(len_chain): bin_no = int(np.floor(np.dot(chain[i].T-d_min,v)/v_2*n_bins)[0]) if n_bins > bin_no > -1: p_x[bin_no] += 1. # end count dx = np.sqrt(v_2)/n_bins p_x = p_x/(len_chain*dx) # find error for i in xrange(n_bins): p = p_x[i] error[i] = np.sqrt(p*(1./dx-p)/(len_chain)) x[:,i] = (d_min+v*(0.5+i)/n_bins)[0] # end find return x, p_x, error # end error_bars # internal methods def _sample(self, n_samples): """ Sample Generate samples for posterior distribution using Gauss-Newton proposal parameters Inputs : n_samples : number of samples to generate Hidden Outputs : chain : chain of samples n_samples : length of chain n_accepted : number of proposals accepted step_count : count of the steps accepted """ try : n_samples = int(n_samples) except : raise TypeError("number of samples has to be an integer") exit() # fetch info X = self._proposal_params(self._X) k_max = self._max_steps # initialize chain = np.zeros((n_samples, self._n)) n_accepted = 0 step_count = np.zeros(k_max+2) # begin outer loop for i in xrange(n_samples): accepted = False # check if sample is accepted r_ = [1] # list of step sizes Z_ = [X] # initialize list of Z s self._r_ = r_ log_P_z_x = 0. + X['log_p'] k = 0 # back-off steps taken so far while k <= k_max: # get proposal chi_z = False while not chi_z: z = multi_normal(X, r_[-1]) chi_z, f_z, J_z = self._f(z) Z = self._proposal_params({'x':z,'f':f_z,'J':J_z}) Z_.append(Z) self._Z_ = Z_ log_P_z_x += log_K(Z, X, r_[-1]) # N is the Numerator of the acceptance, N = P_x_z self._N_is_0 = False # check to see if N = 0, to use in _log_P log_N = self._log_P(X, Z, k) # calculating acceptance probability if self._N_is_0 == True : A_z_x = 0. elif log_N >= log_P_z_x : A_z_x = 1. else : A_z_x = np.exp(log_N - log_P_z_x) # acceptance rejection if np.random.rand() <= A_z_x: accepted = True break else : log_P_z_x += np.log(1. - A_z_x) self._back_off() k += 1 # end of steps for loop if accepted == True : chain[i,:] = z[:,0] X = Z # for statistics n_accepted += 1 step_count[k+1] += 1 else : chain[i,:] = X['x'][:,0] # for statistics step_count[0] += 1 # end outer loop # update stored info self._X = X # outputs if self._n_samples == 0 : self._chain = chain self._step_count = step_count else : self._chain = np.append(self._chain, chain, axis=0) self._step_count = np.add(self._step_count, step_count) self._n_samples += n_samples self._n_accepted += n_accepted # end sample def _proposal_params(self, state): """ Proposal parameters Calculate parameters needed for the proposal. Inputs : state : x : the present sample, the place to linearize around f : f(x), function value at x J : f'(x), the jacobian of the function evaluated at x Outputs : state : mu : the mean vector L : the lower triangular cholesky factor of P log_p : log(p(x)) log of the posterior density """ x = state['x'] f = state['f'] J = state['J'] JJ = np.dot(J.T,J) if self._prior: m = self._m H = self._H Hm = self._Hm # LL' = P = H+J'J L = la.cholesky(H+JJ) # mu = (P^-1)(Hm-J'f+J'Jx) mu = la.solve(L.T,la.solve(L,Hm-np.dot(J.T,f)+np.dot(JJ,x))) else: # P = J'J L = la.cholesky(JJ) # mu = x-(P^-1)J'f mu = x-la.solve(L.T,la.solve(L,np.dot(J.T,f))) state['L'] = L state['mu'] = mu state['log_p'] = self._log_post(x,f) return state def _log_P(self, X , Z, k): """ Log of the probability of transition from z to x with k steps log ( P_k (x, z) ) Inputs : X : state to be proposed to Z : state to be proposed from k : number of recursions, depth """ r_ = self._r_ Z_ = self._Z_ # zero case if k == 0 : log_P = Z['log_p'] + log_K(X, Z, r_[k]) # recursice case else : P_zk_z = np.exp( self._log_P(Z_[k], Z, k-1) ) P_z_zk = np.exp( self._log_P(Z, Z_[k], k-1) ) # flag if P_zk_z <= P_z_zk : self._N_is_0 = True log_P = -np.inf else : log_P = np.log( P_zk_z - P_z_zk ) + log_K(X, Z, r_[k]) return log_P def _back_off(self): """ Back off Calculate the back off step size Inputs : Z_ : list of states in current proposal r_ : list of back offs in current proposal q : step size reduction dynamic : set to True if you want to use the dynamic back-off Outputs : """ q = self._step_size r = self._r_[-1] Z_ = self._Z_ if self._dynamic: p_0 = la.norm(Z_[0]['f']) dp_0 = p_0*2*la.norm(Z_[0]['J']) p_r = la.norm(Z_[-1]['f']) dp_r = p_0*2*la.norm(Z_[-1]['J']) r_new = optimize(r, p_0**2, dp_0, p_r**2, dp_r) else : r_new = r * q self._r_.append(r_new) def _log_post(self,x,f_x): """ Log of the posterior density This is used to calculete acceptance probability for sampling. Inputs : x : input value f_x : f(x), function value at x Outputs : log(p_x) : log[pi(x)]-||f(x)||^2/(2) log of the posterior probability """ # least squares part -||f(x)||^2/2 log_likelihood = (-la.norm(f_x)**2)/(2.) # prior part -(x-m)'H(x-m)/2 if self._prior: m = self._m H = self._H log_prior = self._ln_H_-np.dot((x-m).T,np.dot(H,x-m))/2. return log_prior+log_likelihood else: return log_likelihood """ Properties: 1. chain 2. n_samples 3. n_accepted 4. accept_rate 5. step_count 6. call_count 7. max_steps 8. step_size """ @property def chain(self): return self._chain @property def n_samples(self): return self._n_samples @property def n_accepted(self): return self._n_accepted @property def accept_rate(self): return float(self._n_accepted)/self._n_samples @property def step_count(self): return self._step_count @property def call_count(self): return self._f.count @property def max_steps(self): return self._max_steps @property def step_size(self): return self._step_size
mugurbil/gnm
gnm/gnm.py
sampler.load
python
def load(self, path="chain.dat"): # read data from file file = open(path, 'r') dic = json.load(file) file.close() # get data from dictionary self._chain = np.array(dic['chain']) self._step_count = np.array(dic['step_count']) self._n_samples = dic['n_samples'] self._n_accepted = dic['n_accepted'] self._X = {} self._X['x'] = dic['x'] self._X['f'] = dic['f'] self._X['J'] = dic['J']
Load Load data from file Inputs : path : specifies the path name of the file to be loaded from
train
https://github.com/mugurbil/gnm/blob/4f9711fb9d78cc02820c25234bc3ab9615014f11/gnm/gnm.py#L277-L298
null
class sampler(object): def __init__(self, x, model, args): """ Init Initialize the GNM sampler class Inputs : x : initial guess model : user defined data model function args : arguments for the model """ self._args = args self._f = function(model, args) x = np.reshape(np.array(x), (-1,1)) self._n = np.size(x) # size of input space try: x = np.reshape(np.array(x), (-1)) out_x = model(x, self._args) except TypeError as e: raise TypeError(str(e)[:-7]+" needed)") """ except IndexError as e: raise IndexError("initial guess size does not fit model()\n " +str(e)) except Exception as e: print("Error: Model function could not be evaluated.") print(" - Check size of intial guess.") print(" - Check definition of the model.") print(str(e)) print(type(e)) raise RuntimeError("model() could not be evaluated\n ") """ try: chi_x, f_x, J_x = out_x f_x = np.reshape(np.array(f_x), (-1,1)) except: raise TypeError("model() needs to have 3 outputs: chi_x, f_x, J_x") try: assert chi_x == True except AssertionError: raise ValueError("initial guess out of range") try: self._mn = np.size(f_x) J_x = np.array(J_x) assert np.shape(J_x) == (self._mn, self._n) except: raise TypeError("Shape of Jacobian, " + str(np.shape(J_x)) + ", is not correct, (%d, %d)." % (self._mn, self._n)) x = np.reshape(np.array(x), (-1,1)) self._X = {'x':x,'f':f_x,'J':J_x} # state of x # prior parameters self._prior = False # back-off parameters self._max_steps = 1 self._step_size = 0.1 self._dynamic = False self._opts = {} # sampler outputs self._chain = None self._n_samples = 0 self._n_accepted = 0 def prior(self, m, H): """ Set prior Set prior values Inputs : m : mean of the prior H : precision matrix of the prior Hiddens : ln_H_ : log(det(H))/2 calculate this once to use everytime log prior is called Hm : < H, m > calculate this once to use everytime proposal is called """ if self._prior == True: raise Warning("prior information is already set") else: self._prior = True # mean self._m = np.reshape(np.array(m), (-1,1)) try : assert np.size(self._m) == self._n except : raise TypeError("mean has to be an array of size n") # precision self._H = np.array(H) try : assert np.shape(self._H) == (self._n, self._n) except : raise TypeError("precision has to be a matrix of shape n by n") # precalculations self._ln_H_ = np.log(la.det(self._H))/2. self._Hm = np.dot(self._H, self._m) def Jtest(self, x_min, x_max, dx=0.0002, N=1000, eps_max=0.0001, p=2, l_max=50, r=0.5): """ Gradient Checker Test the function's jacobian against the numerical jacobian """ # check inputs x_min and x_max try : assert np.size(x_min) == self._n except : raise TypeError("dimension of x_min, %d, does not match the " "dimension of input, %d" % (np.size(x_min), self._n)) try : assert np.size(x_max) == self._n except : raise TypeError("dimension of x_max, %d, does not match the " "dimension of input, %d." % (np.size(x_max), self._n)) # end checks and call developer function return self._f.Jtest(x_min, x_max, dx=dx, N=N, eps_max=eps_max, p=p, l_max=l_max, r=r) def static(self, max_steps, step_size): """ Set Back-off to Static Set the sampler parameters for static back off Inputs : max_steps : maximum optimization steps to be taken step_size : the step size of the back-off """ self._dynamic = False # begin checks try : self._max_steps = int(max_steps) except : print("Error: Input 1 (max_steps) has to be an int.") return 0 try : assert self._max_steps >= 0 except : print("Warning: Input 1 (max_steps) has to be non-negative.") print("Setting max_steps to 0.") self._max_steps = 0 if max_steps > 0 : try : assert step_size == float(step_size) except AssertionError : print("Warning: Input 2 (step_size) is not a float. Converted.") step_size = float(step_size) except : print("Error: Input 2 (step_size) has to be a float.") return 0 try : assert 0. < step_size < 1. except : print("Warning: Input 2 (step_size) has to be between 0 and 1.") print("Setting step_size to 0.2.") step_size = 0.2 self._step_size = step_size if step_size**max_steps < 10**(-15): print("Warning: Back-off gets dangerously small.") # end checks def dynamic(self, max_steps, opts={}): """ Dynamic Switch Set the sampler parameters for dynamic back off Inputs : max_steps : maximum back-off steps to be taken Optional Inputs: opts : ({}) dictionary containing fancy options """ self._dynamic = True # begin checks try : self._max_steps = int(max_steps) except : raise TypeError("input 1 (max_steps) has to be an integer") return 0 try : assert self._max_steps >= 0 except : raise Warning("input 1 (max_steps) has to be non-negative. Setting (max_steps) to 0.") self._max_steps = 0 self._opts = opts # end checks def sample(self, n_samples, divs=1, visual=False, safe=False): """ Sample Sampling Inputs : n_samples : number of samples to generate Optional Inputs : divs : (1) number of divisions visual : show progress safe : save the chain at every division """ if visual: print("Sampling: 0%") for i in xrange(divs): self._sample(int(n_samples/divs)) if visual: sys.stdout.write("\033[F") # curser up print("Sampling: "+str(int(i*100./divs)+1)+'%') if safe: self.save(path="chain_{:}.dat".format(i)) if n_samples % divs != 0: self._sample(n_samples % divs) if safe: self.save(path="chain_{:}.dat".format(divs)) def save(self, path="chain.dat"): """ Save Save data to file Inputs : path : specifies the path name of the file to be loaded to """ # create dictionary for data dic = {} dic['chain'] = self._chain.tolist() dic['step_count'] = self._step_count.tolist() dic['n_samples'] = self._n_samples dic['n_accepted'] = self._n_accepted dic['x'] = self._X['x'].tolist() dic['f'] = self._X['f'].tolist() dic['J'] = self._X['J'].tolist() # write data to file file = open(path, 'w') json.dump(dic, file) file.close() def burn(self, n_burned): """ Burn Burn the inital samples to adjust for convergence of the chain cut the first (n_burned) burn-in samples Inputs : chain : the full Markov chain n_burned : number of samples to cut Hidden Outputs : chain : chain with the firt n_burned samples cut """ self._chain = self._chain[n_burned:] def acor(self, k = 5): """ Autocorrelation time of the chain return the autocorrelation time for each parameters Inputs : k : parameter in self-consistent window Outputs : t : autocorrelation time of the chain """ try: import acor except ImportError: print("Can't import acor, please download it.") return 0 n = np.shape(self._chain)[1] t = np.zeros(n) for i in xrange(n): t[i] = acor.acor(self._chain[:,i],k)[0] return t def posterior(self, x): """ Posterior density ** not normalized ** This is used to plot the theoretical curve for tests. Inputs : x : input value Outputs : p : p(x)=pi(x)*exp{-||f(x)||^2/(2)} posterior probability of x """ x = np.reshape(np.array(x), (-1,1)) chi_x, f_x, J_x = self._f(x) if chi_x : p = np.exp(-la.norm(f_x)**2/2.) if self._prior: m = self._m H = self._H p = p * np.exp(-np.dot((x-m).T,np.dot(H,x-m))/2.) return p else : return 0 def error_bars(self, n_bins, d_min, d_max): """ Error Bars create bars and error bars to plot Inputs : n_bins : number of bins plot_range : (shape) = (number of dimensions, 2) matrix which contain the min and max for each dimension as rows Outputs : x : domain p_x : estimated posterior using the chain on the domain error : estimated error for p_x """ # fetch data chain = self._chain len_chain = len(chain) try: n_dims = np.shape(chain)[1] except: n_dims = 1 # begin checks try: assert n_bins == int(n_bins) except: raise TypeError("number of bins has to be an integer") d_min = np.reshape(np.array(d_min), (-1,1)) d_max = np.reshape(np.array(d_max), (-1,1)) try: assert np.size(d_min) == n_dims except: raise TypeError("domain minimum has wrong size") try: assert np.size(d_max) == n_dims except: raise TypeError("domain maximum has wrong size") # end checks # initialize outputs p_x = np.zeros(n_bins) # esitmate of posterior error = np.zeros(n_bins) # error bars x = np.zeros((n_dims, n_bins)) # centers of bins # set dx v = d_max-d_min v_2 = np.dot(v.T, v)[0][0] # bin count for i in xrange(len_chain): bin_no = int(np.floor(np.dot(chain[i].T-d_min,v)/v_2*n_bins)[0]) if n_bins > bin_no > -1: p_x[bin_no] += 1. # end count dx = np.sqrt(v_2)/n_bins p_x = p_x/(len_chain*dx) # find error for i in xrange(n_bins): p = p_x[i] error[i] = np.sqrt(p*(1./dx-p)/(len_chain)) x[:,i] = (d_min+v*(0.5+i)/n_bins)[0] # end find return x, p_x, error # end error_bars # internal methods def _sample(self, n_samples): """ Sample Generate samples for posterior distribution using Gauss-Newton proposal parameters Inputs : n_samples : number of samples to generate Hidden Outputs : chain : chain of samples n_samples : length of chain n_accepted : number of proposals accepted step_count : count of the steps accepted """ try : n_samples = int(n_samples) except : raise TypeError("number of samples has to be an integer") exit() # fetch info X = self._proposal_params(self._X) k_max = self._max_steps # initialize chain = np.zeros((n_samples, self._n)) n_accepted = 0 step_count = np.zeros(k_max+2) # begin outer loop for i in xrange(n_samples): accepted = False # check if sample is accepted r_ = [1] # list of step sizes Z_ = [X] # initialize list of Z s self._r_ = r_ log_P_z_x = 0. + X['log_p'] k = 0 # back-off steps taken so far while k <= k_max: # get proposal chi_z = False while not chi_z: z = multi_normal(X, r_[-1]) chi_z, f_z, J_z = self._f(z) Z = self._proposal_params({'x':z,'f':f_z,'J':J_z}) Z_.append(Z) self._Z_ = Z_ log_P_z_x += log_K(Z, X, r_[-1]) # N is the Numerator of the acceptance, N = P_x_z self._N_is_0 = False # check to see if N = 0, to use in _log_P log_N = self._log_P(X, Z, k) # calculating acceptance probability if self._N_is_0 == True : A_z_x = 0. elif log_N >= log_P_z_x : A_z_x = 1. else : A_z_x = np.exp(log_N - log_P_z_x) # acceptance rejection if np.random.rand() <= A_z_x: accepted = True break else : log_P_z_x += np.log(1. - A_z_x) self._back_off() k += 1 # end of steps for loop if accepted == True : chain[i,:] = z[:,0] X = Z # for statistics n_accepted += 1 step_count[k+1] += 1 else : chain[i,:] = X['x'][:,0] # for statistics step_count[0] += 1 # end outer loop # update stored info self._X = X # outputs if self._n_samples == 0 : self._chain = chain self._step_count = step_count else : self._chain = np.append(self._chain, chain, axis=0) self._step_count = np.add(self._step_count, step_count) self._n_samples += n_samples self._n_accepted += n_accepted # end sample def _proposal_params(self, state): """ Proposal parameters Calculate parameters needed for the proposal. Inputs : state : x : the present sample, the place to linearize around f : f(x), function value at x J : f'(x), the jacobian of the function evaluated at x Outputs : state : mu : the mean vector L : the lower triangular cholesky factor of P log_p : log(p(x)) log of the posterior density """ x = state['x'] f = state['f'] J = state['J'] JJ = np.dot(J.T,J) if self._prior: m = self._m H = self._H Hm = self._Hm # LL' = P = H+J'J L = la.cholesky(H+JJ) # mu = (P^-1)(Hm-J'f+J'Jx) mu = la.solve(L.T,la.solve(L,Hm-np.dot(J.T,f)+np.dot(JJ,x))) else: # P = J'J L = la.cholesky(JJ) # mu = x-(P^-1)J'f mu = x-la.solve(L.T,la.solve(L,np.dot(J.T,f))) state['L'] = L state['mu'] = mu state['log_p'] = self._log_post(x,f) return state def _log_P(self, X , Z, k): """ Log of the probability of transition from z to x with k steps log ( P_k (x, z) ) Inputs : X : state to be proposed to Z : state to be proposed from k : number of recursions, depth """ r_ = self._r_ Z_ = self._Z_ # zero case if k == 0 : log_P = Z['log_p'] + log_K(X, Z, r_[k]) # recursice case else : P_zk_z = np.exp( self._log_P(Z_[k], Z, k-1) ) P_z_zk = np.exp( self._log_P(Z, Z_[k], k-1) ) # flag if P_zk_z <= P_z_zk : self._N_is_0 = True log_P = -np.inf else : log_P = np.log( P_zk_z - P_z_zk ) + log_K(X, Z, r_[k]) return log_P def _back_off(self): """ Back off Calculate the back off step size Inputs : Z_ : list of states in current proposal r_ : list of back offs in current proposal q : step size reduction dynamic : set to True if you want to use the dynamic back-off Outputs : """ q = self._step_size r = self._r_[-1] Z_ = self._Z_ if self._dynamic: p_0 = la.norm(Z_[0]['f']) dp_0 = p_0*2*la.norm(Z_[0]['J']) p_r = la.norm(Z_[-1]['f']) dp_r = p_0*2*la.norm(Z_[-1]['J']) r_new = optimize(r, p_0**2, dp_0, p_r**2, dp_r) else : r_new = r * q self._r_.append(r_new) def _log_post(self,x,f_x): """ Log of the posterior density This is used to calculete acceptance probability for sampling. Inputs : x : input value f_x : f(x), function value at x Outputs : log(p_x) : log[pi(x)]-||f(x)||^2/(2) log of the posterior probability """ # least squares part -||f(x)||^2/2 log_likelihood = (-la.norm(f_x)**2)/(2.) # prior part -(x-m)'H(x-m)/2 if self._prior: m = self._m H = self._H log_prior = self._ln_H_-np.dot((x-m).T,np.dot(H,x-m))/2. return log_prior+log_likelihood else: return log_likelihood """ Properties: 1. chain 2. n_samples 3. n_accepted 4. accept_rate 5. step_count 6. call_count 7. max_steps 8. step_size """ @property def chain(self): return self._chain @property def n_samples(self): return self._n_samples @property def n_accepted(self): return self._n_accepted @property def accept_rate(self): return float(self._n_accepted)/self._n_samples @property def step_count(self): return self._step_count @property def call_count(self): return self._f.count @property def max_steps(self): return self._max_steps @property def step_size(self): return self._step_size
mugurbil/gnm
gnm/gnm.py
sampler.acor
python
def acor(self, k = 5): try: import acor except ImportError: print("Can't import acor, please download it.") return 0 n = np.shape(self._chain)[1] t = np.zeros(n) for i in xrange(n): t[i] = acor.acor(self._chain[:,i],k)[0] return t
Autocorrelation time of the chain return the autocorrelation time for each parameters Inputs : k : parameter in self-consistent window Outputs : t : autocorrelation time of the chain
train
https://github.com/mugurbil/gnm/blob/4f9711fb9d78cc02820c25234bc3ab9615014f11/gnm/gnm.py#L316-L336
null
class sampler(object): def __init__(self, x, model, args): """ Init Initialize the GNM sampler class Inputs : x : initial guess model : user defined data model function args : arguments for the model """ self._args = args self._f = function(model, args) x = np.reshape(np.array(x), (-1,1)) self._n = np.size(x) # size of input space try: x = np.reshape(np.array(x), (-1)) out_x = model(x, self._args) except TypeError as e: raise TypeError(str(e)[:-7]+" needed)") """ except IndexError as e: raise IndexError("initial guess size does not fit model()\n " +str(e)) except Exception as e: print("Error: Model function could not be evaluated.") print(" - Check size of intial guess.") print(" - Check definition of the model.") print(str(e)) print(type(e)) raise RuntimeError("model() could not be evaluated\n ") """ try: chi_x, f_x, J_x = out_x f_x = np.reshape(np.array(f_x), (-1,1)) except: raise TypeError("model() needs to have 3 outputs: chi_x, f_x, J_x") try: assert chi_x == True except AssertionError: raise ValueError("initial guess out of range") try: self._mn = np.size(f_x) J_x = np.array(J_x) assert np.shape(J_x) == (self._mn, self._n) except: raise TypeError("Shape of Jacobian, " + str(np.shape(J_x)) + ", is not correct, (%d, %d)." % (self._mn, self._n)) x = np.reshape(np.array(x), (-1,1)) self._X = {'x':x,'f':f_x,'J':J_x} # state of x # prior parameters self._prior = False # back-off parameters self._max_steps = 1 self._step_size = 0.1 self._dynamic = False self._opts = {} # sampler outputs self._chain = None self._n_samples = 0 self._n_accepted = 0 def prior(self, m, H): """ Set prior Set prior values Inputs : m : mean of the prior H : precision matrix of the prior Hiddens : ln_H_ : log(det(H))/2 calculate this once to use everytime log prior is called Hm : < H, m > calculate this once to use everytime proposal is called """ if self._prior == True: raise Warning("prior information is already set") else: self._prior = True # mean self._m = np.reshape(np.array(m), (-1,1)) try : assert np.size(self._m) == self._n except : raise TypeError("mean has to be an array of size n") # precision self._H = np.array(H) try : assert np.shape(self._H) == (self._n, self._n) except : raise TypeError("precision has to be a matrix of shape n by n") # precalculations self._ln_H_ = np.log(la.det(self._H))/2. self._Hm = np.dot(self._H, self._m) def Jtest(self, x_min, x_max, dx=0.0002, N=1000, eps_max=0.0001, p=2, l_max=50, r=0.5): """ Gradient Checker Test the function's jacobian against the numerical jacobian """ # check inputs x_min and x_max try : assert np.size(x_min) == self._n except : raise TypeError("dimension of x_min, %d, does not match the " "dimension of input, %d" % (np.size(x_min), self._n)) try : assert np.size(x_max) == self._n except : raise TypeError("dimension of x_max, %d, does not match the " "dimension of input, %d." % (np.size(x_max), self._n)) # end checks and call developer function return self._f.Jtest(x_min, x_max, dx=dx, N=N, eps_max=eps_max, p=p, l_max=l_max, r=r) def static(self, max_steps, step_size): """ Set Back-off to Static Set the sampler parameters for static back off Inputs : max_steps : maximum optimization steps to be taken step_size : the step size of the back-off """ self._dynamic = False # begin checks try : self._max_steps = int(max_steps) except : print("Error: Input 1 (max_steps) has to be an int.") return 0 try : assert self._max_steps >= 0 except : print("Warning: Input 1 (max_steps) has to be non-negative.") print("Setting max_steps to 0.") self._max_steps = 0 if max_steps > 0 : try : assert step_size == float(step_size) except AssertionError : print("Warning: Input 2 (step_size) is not a float. Converted.") step_size = float(step_size) except : print("Error: Input 2 (step_size) has to be a float.") return 0 try : assert 0. < step_size < 1. except : print("Warning: Input 2 (step_size) has to be between 0 and 1.") print("Setting step_size to 0.2.") step_size = 0.2 self._step_size = step_size if step_size**max_steps < 10**(-15): print("Warning: Back-off gets dangerously small.") # end checks def dynamic(self, max_steps, opts={}): """ Dynamic Switch Set the sampler parameters for dynamic back off Inputs : max_steps : maximum back-off steps to be taken Optional Inputs: opts : ({}) dictionary containing fancy options """ self._dynamic = True # begin checks try : self._max_steps = int(max_steps) except : raise TypeError("input 1 (max_steps) has to be an integer") return 0 try : assert self._max_steps >= 0 except : raise Warning("input 1 (max_steps) has to be non-negative. Setting (max_steps) to 0.") self._max_steps = 0 self._opts = opts # end checks def sample(self, n_samples, divs=1, visual=False, safe=False): """ Sample Sampling Inputs : n_samples : number of samples to generate Optional Inputs : divs : (1) number of divisions visual : show progress safe : save the chain at every division """ if visual: print("Sampling: 0%") for i in xrange(divs): self._sample(int(n_samples/divs)) if visual: sys.stdout.write("\033[F") # curser up print("Sampling: "+str(int(i*100./divs)+1)+'%') if safe: self.save(path="chain_{:}.dat".format(i)) if n_samples % divs != 0: self._sample(n_samples % divs) if safe: self.save(path="chain_{:}.dat".format(divs)) def save(self, path="chain.dat"): """ Save Save data to file Inputs : path : specifies the path name of the file to be loaded to """ # create dictionary for data dic = {} dic['chain'] = self._chain.tolist() dic['step_count'] = self._step_count.tolist() dic['n_samples'] = self._n_samples dic['n_accepted'] = self._n_accepted dic['x'] = self._X['x'].tolist() dic['f'] = self._X['f'].tolist() dic['J'] = self._X['J'].tolist() # write data to file file = open(path, 'w') json.dump(dic, file) file.close() def load(self, path="chain.dat"): """ Load Load data from file Inputs : path : specifies the path name of the file to be loaded from """ # read data from file file = open(path, 'r') dic = json.load(file) file.close() # get data from dictionary self._chain = np.array(dic['chain']) self._step_count = np.array(dic['step_count']) self._n_samples = dic['n_samples'] self._n_accepted = dic['n_accepted'] self._X = {} self._X['x'] = dic['x'] self._X['f'] = dic['f'] self._X['J'] = dic['J'] def burn(self, n_burned): """ Burn Burn the inital samples to adjust for convergence of the chain cut the first (n_burned) burn-in samples Inputs : chain : the full Markov chain n_burned : number of samples to cut Hidden Outputs : chain : chain with the firt n_burned samples cut """ self._chain = self._chain[n_burned:] def posterior(self, x): """ Posterior density ** not normalized ** This is used to plot the theoretical curve for tests. Inputs : x : input value Outputs : p : p(x)=pi(x)*exp{-||f(x)||^2/(2)} posterior probability of x """ x = np.reshape(np.array(x), (-1,1)) chi_x, f_x, J_x = self._f(x) if chi_x : p = np.exp(-la.norm(f_x)**2/2.) if self._prior: m = self._m H = self._H p = p * np.exp(-np.dot((x-m).T,np.dot(H,x-m))/2.) return p else : return 0 def error_bars(self, n_bins, d_min, d_max): """ Error Bars create bars and error bars to plot Inputs : n_bins : number of bins plot_range : (shape) = (number of dimensions, 2) matrix which contain the min and max for each dimension as rows Outputs : x : domain p_x : estimated posterior using the chain on the domain error : estimated error for p_x """ # fetch data chain = self._chain len_chain = len(chain) try: n_dims = np.shape(chain)[1] except: n_dims = 1 # begin checks try: assert n_bins == int(n_bins) except: raise TypeError("number of bins has to be an integer") d_min = np.reshape(np.array(d_min), (-1,1)) d_max = np.reshape(np.array(d_max), (-1,1)) try: assert np.size(d_min) == n_dims except: raise TypeError("domain minimum has wrong size") try: assert np.size(d_max) == n_dims except: raise TypeError("domain maximum has wrong size") # end checks # initialize outputs p_x = np.zeros(n_bins) # esitmate of posterior error = np.zeros(n_bins) # error bars x = np.zeros((n_dims, n_bins)) # centers of bins # set dx v = d_max-d_min v_2 = np.dot(v.T, v)[0][0] # bin count for i in xrange(len_chain): bin_no = int(np.floor(np.dot(chain[i].T-d_min,v)/v_2*n_bins)[0]) if n_bins > bin_no > -1: p_x[bin_no] += 1. # end count dx = np.sqrt(v_2)/n_bins p_x = p_x/(len_chain*dx) # find error for i in xrange(n_bins): p = p_x[i] error[i] = np.sqrt(p*(1./dx-p)/(len_chain)) x[:,i] = (d_min+v*(0.5+i)/n_bins)[0] # end find return x, p_x, error # end error_bars # internal methods def _sample(self, n_samples): """ Sample Generate samples for posterior distribution using Gauss-Newton proposal parameters Inputs : n_samples : number of samples to generate Hidden Outputs : chain : chain of samples n_samples : length of chain n_accepted : number of proposals accepted step_count : count of the steps accepted """ try : n_samples = int(n_samples) except : raise TypeError("number of samples has to be an integer") exit() # fetch info X = self._proposal_params(self._X) k_max = self._max_steps # initialize chain = np.zeros((n_samples, self._n)) n_accepted = 0 step_count = np.zeros(k_max+2) # begin outer loop for i in xrange(n_samples): accepted = False # check if sample is accepted r_ = [1] # list of step sizes Z_ = [X] # initialize list of Z s self._r_ = r_ log_P_z_x = 0. + X['log_p'] k = 0 # back-off steps taken so far while k <= k_max: # get proposal chi_z = False while not chi_z: z = multi_normal(X, r_[-1]) chi_z, f_z, J_z = self._f(z) Z = self._proposal_params({'x':z,'f':f_z,'J':J_z}) Z_.append(Z) self._Z_ = Z_ log_P_z_x += log_K(Z, X, r_[-1]) # N is the Numerator of the acceptance, N = P_x_z self._N_is_0 = False # check to see if N = 0, to use in _log_P log_N = self._log_P(X, Z, k) # calculating acceptance probability if self._N_is_0 == True : A_z_x = 0. elif log_N >= log_P_z_x : A_z_x = 1. else : A_z_x = np.exp(log_N - log_P_z_x) # acceptance rejection if np.random.rand() <= A_z_x: accepted = True break else : log_P_z_x += np.log(1. - A_z_x) self._back_off() k += 1 # end of steps for loop if accepted == True : chain[i,:] = z[:,0] X = Z # for statistics n_accepted += 1 step_count[k+1] += 1 else : chain[i,:] = X['x'][:,0] # for statistics step_count[0] += 1 # end outer loop # update stored info self._X = X # outputs if self._n_samples == 0 : self._chain = chain self._step_count = step_count else : self._chain = np.append(self._chain, chain, axis=0) self._step_count = np.add(self._step_count, step_count) self._n_samples += n_samples self._n_accepted += n_accepted # end sample def _proposal_params(self, state): """ Proposal parameters Calculate parameters needed for the proposal. Inputs : state : x : the present sample, the place to linearize around f : f(x), function value at x J : f'(x), the jacobian of the function evaluated at x Outputs : state : mu : the mean vector L : the lower triangular cholesky factor of P log_p : log(p(x)) log of the posterior density """ x = state['x'] f = state['f'] J = state['J'] JJ = np.dot(J.T,J) if self._prior: m = self._m H = self._H Hm = self._Hm # LL' = P = H+J'J L = la.cholesky(H+JJ) # mu = (P^-1)(Hm-J'f+J'Jx) mu = la.solve(L.T,la.solve(L,Hm-np.dot(J.T,f)+np.dot(JJ,x))) else: # P = J'J L = la.cholesky(JJ) # mu = x-(P^-1)J'f mu = x-la.solve(L.T,la.solve(L,np.dot(J.T,f))) state['L'] = L state['mu'] = mu state['log_p'] = self._log_post(x,f) return state def _log_P(self, X , Z, k): """ Log of the probability of transition from z to x with k steps log ( P_k (x, z) ) Inputs : X : state to be proposed to Z : state to be proposed from k : number of recursions, depth """ r_ = self._r_ Z_ = self._Z_ # zero case if k == 0 : log_P = Z['log_p'] + log_K(X, Z, r_[k]) # recursice case else : P_zk_z = np.exp( self._log_P(Z_[k], Z, k-1) ) P_z_zk = np.exp( self._log_P(Z, Z_[k], k-1) ) # flag if P_zk_z <= P_z_zk : self._N_is_0 = True log_P = -np.inf else : log_P = np.log( P_zk_z - P_z_zk ) + log_K(X, Z, r_[k]) return log_P def _back_off(self): """ Back off Calculate the back off step size Inputs : Z_ : list of states in current proposal r_ : list of back offs in current proposal q : step size reduction dynamic : set to True if you want to use the dynamic back-off Outputs : """ q = self._step_size r = self._r_[-1] Z_ = self._Z_ if self._dynamic: p_0 = la.norm(Z_[0]['f']) dp_0 = p_0*2*la.norm(Z_[0]['J']) p_r = la.norm(Z_[-1]['f']) dp_r = p_0*2*la.norm(Z_[-1]['J']) r_new = optimize(r, p_0**2, dp_0, p_r**2, dp_r) else : r_new = r * q self._r_.append(r_new) def _log_post(self,x,f_x): """ Log of the posterior density This is used to calculete acceptance probability for sampling. Inputs : x : input value f_x : f(x), function value at x Outputs : log(p_x) : log[pi(x)]-||f(x)||^2/(2) log of the posterior probability """ # least squares part -||f(x)||^2/2 log_likelihood = (-la.norm(f_x)**2)/(2.) # prior part -(x-m)'H(x-m)/2 if self._prior: m = self._m H = self._H log_prior = self._ln_H_-np.dot((x-m).T,np.dot(H,x-m))/2. return log_prior+log_likelihood else: return log_likelihood """ Properties: 1. chain 2. n_samples 3. n_accepted 4. accept_rate 5. step_count 6. call_count 7. max_steps 8. step_size """ @property def chain(self): return self._chain @property def n_samples(self): return self._n_samples @property def n_accepted(self): return self._n_accepted @property def accept_rate(self): return float(self._n_accepted)/self._n_samples @property def step_count(self): return self._step_count @property def call_count(self): return self._f.count @property def max_steps(self): return self._max_steps @property def step_size(self): return self._step_size
mugurbil/gnm
gnm/gnm.py
sampler.posterior
python
def posterior(self, x): x = np.reshape(np.array(x), (-1,1)) chi_x, f_x, J_x = self._f(x) if chi_x : p = np.exp(-la.norm(f_x)**2/2.) if self._prior: m = self._m H = self._H p = p * np.exp(-np.dot((x-m).T,np.dot(H,x-m))/2.) return p else : return 0
Posterior density ** not normalized ** This is used to plot the theoretical curve for tests. Inputs : x : input value Outputs : p : p(x)=pi(x)*exp{-||f(x)||^2/(2)} posterior probability of x
train
https://github.com/mugurbil/gnm/blob/4f9711fb9d78cc02820c25234bc3ab9615014f11/gnm/gnm.py#L338-L361
null
class sampler(object): def __init__(self, x, model, args): """ Init Initialize the GNM sampler class Inputs : x : initial guess model : user defined data model function args : arguments for the model """ self._args = args self._f = function(model, args) x = np.reshape(np.array(x), (-1,1)) self._n = np.size(x) # size of input space try: x = np.reshape(np.array(x), (-1)) out_x = model(x, self._args) except TypeError as e: raise TypeError(str(e)[:-7]+" needed)") """ except IndexError as e: raise IndexError("initial guess size does not fit model()\n " +str(e)) except Exception as e: print("Error: Model function could not be evaluated.") print(" - Check size of intial guess.") print(" - Check definition of the model.") print(str(e)) print(type(e)) raise RuntimeError("model() could not be evaluated\n ") """ try: chi_x, f_x, J_x = out_x f_x = np.reshape(np.array(f_x), (-1,1)) except: raise TypeError("model() needs to have 3 outputs: chi_x, f_x, J_x") try: assert chi_x == True except AssertionError: raise ValueError("initial guess out of range") try: self._mn = np.size(f_x) J_x = np.array(J_x) assert np.shape(J_x) == (self._mn, self._n) except: raise TypeError("Shape of Jacobian, " + str(np.shape(J_x)) + ", is not correct, (%d, %d)." % (self._mn, self._n)) x = np.reshape(np.array(x), (-1,1)) self._X = {'x':x,'f':f_x,'J':J_x} # state of x # prior parameters self._prior = False # back-off parameters self._max_steps = 1 self._step_size = 0.1 self._dynamic = False self._opts = {} # sampler outputs self._chain = None self._n_samples = 0 self._n_accepted = 0 def prior(self, m, H): """ Set prior Set prior values Inputs : m : mean of the prior H : precision matrix of the prior Hiddens : ln_H_ : log(det(H))/2 calculate this once to use everytime log prior is called Hm : < H, m > calculate this once to use everytime proposal is called """ if self._prior == True: raise Warning("prior information is already set") else: self._prior = True # mean self._m = np.reshape(np.array(m), (-1,1)) try : assert np.size(self._m) == self._n except : raise TypeError("mean has to be an array of size n") # precision self._H = np.array(H) try : assert np.shape(self._H) == (self._n, self._n) except : raise TypeError("precision has to be a matrix of shape n by n") # precalculations self._ln_H_ = np.log(la.det(self._H))/2. self._Hm = np.dot(self._H, self._m) def Jtest(self, x_min, x_max, dx=0.0002, N=1000, eps_max=0.0001, p=2, l_max=50, r=0.5): """ Gradient Checker Test the function's jacobian against the numerical jacobian """ # check inputs x_min and x_max try : assert np.size(x_min) == self._n except : raise TypeError("dimension of x_min, %d, does not match the " "dimension of input, %d" % (np.size(x_min), self._n)) try : assert np.size(x_max) == self._n except : raise TypeError("dimension of x_max, %d, does not match the " "dimension of input, %d." % (np.size(x_max), self._n)) # end checks and call developer function return self._f.Jtest(x_min, x_max, dx=dx, N=N, eps_max=eps_max, p=p, l_max=l_max, r=r) def static(self, max_steps, step_size): """ Set Back-off to Static Set the sampler parameters for static back off Inputs : max_steps : maximum optimization steps to be taken step_size : the step size of the back-off """ self._dynamic = False # begin checks try : self._max_steps = int(max_steps) except : print("Error: Input 1 (max_steps) has to be an int.") return 0 try : assert self._max_steps >= 0 except : print("Warning: Input 1 (max_steps) has to be non-negative.") print("Setting max_steps to 0.") self._max_steps = 0 if max_steps > 0 : try : assert step_size == float(step_size) except AssertionError : print("Warning: Input 2 (step_size) is not a float. Converted.") step_size = float(step_size) except : print("Error: Input 2 (step_size) has to be a float.") return 0 try : assert 0. < step_size < 1. except : print("Warning: Input 2 (step_size) has to be between 0 and 1.") print("Setting step_size to 0.2.") step_size = 0.2 self._step_size = step_size if step_size**max_steps < 10**(-15): print("Warning: Back-off gets dangerously small.") # end checks def dynamic(self, max_steps, opts={}): """ Dynamic Switch Set the sampler parameters for dynamic back off Inputs : max_steps : maximum back-off steps to be taken Optional Inputs: opts : ({}) dictionary containing fancy options """ self._dynamic = True # begin checks try : self._max_steps = int(max_steps) except : raise TypeError("input 1 (max_steps) has to be an integer") return 0 try : assert self._max_steps >= 0 except : raise Warning("input 1 (max_steps) has to be non-negative. Setting (max_steps) to 0.") self._max_steps = 0 self._opts = opts # end checks def sample(self, n_samples, divs=1, visual=False, safe=False): """ Sample Sampling Inputs : n_samples : number of samples to generate Optional Inputs : divs : (1) number of divisions visual : show progress safe : save the chain at every division """ if visual: print("Sampling: 0%") for i in xrange(divs): self._sample(int(n_samples/divs)) if visual: sys.stdout.write("\033[F") # curser up print("Sampling: "+str(int(i*100./divs)+1)+'%') if safe: self.save(path="chain_{:}.dat".format(i)) if n_samples % divs != 0: self._sample(n_samples % divs) if safe: self.save(path="chain_{:}.dat".format(divs)) def save(self, path="chain.dat"): """ Save Save data to file Inputs : path : specifies the path name of the file to be loaded to """ # create dictionary for data dic = {} dic['chain'] = self._chain.tolist() dic['step_count'] = self._step_count.tolist() dic['n_samples'] = self._n_samples dic['n_accepted'] = self._n_accepted dic['x'] = self._X['x'].tolist() dic['f'] = self._X['f'].tolist() dic['J'] = self._X['J'].tolist() # write data to file file = open(path, 'w') json.dump(dic, file) file.close() def load(self, path="chain.dat"): """ Load Load data from file Inputs : path : specifies the path name of the file to be loaded from """ # read data from file file = open(path, 'r') dic = json.load(file) file.close() # get data from dictionary self._chain = np.array(dic['chain']) self._step_count = np.array(dic['step_count']) self._n_samples = dic['n_samples'] self._n_accepted = dic['n_accepted'] self._X = {} self._X['x'] = dic['x'] self._X['f'] = dic['f'] self._X['J'] = dic['J'] def burn(self, n_burned): """ Burn Burn the inital samples to adjust for convergence of the chain cut the first (n_burned) burn-in samples Inputs : chain : the full Markov chain n_burned : number of samples to cut Hidden Outputs : chain : chain with the firt n_burned samples cut """ self._chain = self._chain[n_burned:] def acor(self, k = 5): """ Autocorrelation time of the chain return the autocorrelation time for each parameters Inputs : k : parameter in self-consistent window Outputs : t : autocorrelation time of the chain """ try: import acor except ImportError: print("Can't import acor, please download it.") return 0 n = np.shape(self._chain)[1] t = np.zeros(n) for i in xrange(n): t[i] = acor.acor(self._chain[:,i],k)[0] return t def error_bars(self, n_bins, d_min, d_max): """ Error Bars create bars and error bars to plot Inputs : n_bins : number of bins plot_range : (shape) = (number of dimensions, 2) matrix which contain the min and max for each dimension as rows Outputs : x : domain p_x : estimated posterior using the chain on the domain error : estimated error for p_x """ # fetch data chain = self._chain len_chain = len(chain) try: n_dims = np.shape(chain)[1] except: n_dims = 1 # begin checks try: assert n_bins == int(n_bins) except: raise TypeError("number of bins has to be an integer") d_min = np.reshape(np.array(d_min), (-1,1)) d_max = np.reshape(np.array(d_max), (-1,1)) try: assert np.size(d_min) == n_dims except: raise TypeError("domain minimum has wrong size") try: assert np.size(d_max) == n_dims except: raise TypeError("domain maximum has wrong size") # end checks # initialize outputs p_x = np.zeros(n_bins) # esitmate of posterior error = np.zeros(n_bins) # error bars x = np.zeros((n_dims, n_bins)) # centers of bins # set dx v = d_max-d_min v_2 = np.dot(v.T, v)[0][0] # bin count for i in xrange(len_chain): bin_no = int(np.floor(np.dot(chain[i].T-d_min,v)/v_2*n_bins)[0]) if n_bins > bin_no > -1: p_x[bin_no] += 1. # end count dx = np.sqrt(v_2)/n_bins p_x = p_x/(len_chain*dx) # find error for i in xrange(n_bins): p = p_x[i] error[i] = np.sqrt(p*(1./dx-p)/(len_chain)) x[:,i] = (d_min+v*(0.5+i)/n_bins)[0] # end find return x, p_x, error # end error_bars # internal methods def _sample(self, n_samples): """ Sample Generate samples for posterior distribution using Gauss-Newton proposal parameters Inputs : n_samples : number of samples to generate Hidden Outputs : chain : chain of samples n_samples : length of chain n_accepted : number of proposals accepted step_count : count of the steps accepted """ try : n_samples = int(n_samples) except : raise TypeError("number of samples has to be an integer") exit() # fetch info X = self._proposal_params(self._X) k_max = self._max_steps # initialize chain = np.zeros((n_samples, self._n)) n_accepted = 0 step_count = np.zeros(k_max+2) # begin outer loop for i in xrange(n_samples): accepted = False # check if sample is accepted r_ = [1] # list of step sizes Z_ = [X] # initialize list of Z s self._r_ = r_ log_P_z_x = 0. + X['log_p'] k = 0 # back-off steps taken so far while k <= k_max: # get proposal chi_z = False while not chi_z: z = multi_normal(X, r_[-1]) chi_z, f_z, J_z = self._f(z) Z = self._proposal_params({'x':z,'f':f_z,'J':J_z}) Z_.append(Z) self._Z_ = Z_ log_P_z_x += log_K(Z, X, r_[-1]) # N is the Numerator of the acceptance, N = P_x_z self._N_is_0 = False # check to see if N = 0, to use in _log_P log_N = self._log_P(X, Z, k) # calculating acceptance probability if self._N_is_0 == True : A_z_x = 0. elif log_N >= log_P_z_x : A_z_x = 1. else : A_z_x = np.exp(log_N - log_P_z_x) # acceptance rejection if np.random.rand() <= A_z_x: accepted = True break else : log_P_z_x += np.log(1. - A_z_x) self._back_off() k += 1 # end of steps for loop if accepted == True : chain[i,:] = z[:,0] X = Z # for statistics n_accepted += 1 step_count[k+1] += 1 else : chain[i,:] = X['x'][:,0] # for statistics step_count[0] += 1 # end outer loop # update stored info self._X = X # outputs if self._n_samples == 0 : self._chain = chain self._step_count = step_count else : self._chain = np.append(self._chain, chain, axis=0) self._step_count = np.add(self._step_count, step_count) self._n_samples += n_samples self._n_accepted += n_accepted # end sample def _proposal_params(self, state): """ Proposal parameters Calculate parameters needed for the proposal. Inputs : state : x : the present sample, the place to linearize around f : f(x), function value at x J : f'(x), the jacobian of the function evaluated at x Outputs : state : mu : the mean vector L : the lower triangular cholesky factor of P log_p : log(p(x)) log of the posterior density """ x = state['x'] f = state['f'] J = state['J'] JJ = np.dot(J.T,J) if self._prior: m = self._m H = self._H Hm = self._Hm # LL' = P = H+J'J L = la.cholesky(H+JJ) # mu = (P^-1)(Hm-J'f+J'Jx) mu = la.solve(L.T,la.solve(L,Hm-np.dot(J.T,f)+np.dot(JJ,x))) else: # P = J'J L = la.cholesky(JJ) # mu = x-(P^-1)J'f mu = x-la.solve(L.T,la.solve(L,np.dot(J.T,f))) state['L'] = L state['mu'] = mu state['log_p'] = self._log_post(x,f) return state def _log_P(self, X , Z, k): """ Log of the probability of transition from z to x with k steps log ( P_k (x, z) ) Inputs : X : state to be proposed to Z : state to be proposed from k : number of recursions, depth """ r_ = self._r_ Z_ = self._Z_ # zero case if k == 0 : log_P = Z['log_p'] + log_K(X, Z, r_[k]) # recursice case else : P_zk_z = np.exp( self._log_P(Z_[k], Z, k-1) ) P_z_zk = np.exp( self._log_P(Z, Z_[k], k-1) ) # flag if P_zk_z <= P_z_zk : self._N_is_0 = True log_P = -np.inf else : log_P = np.log( P_zk_z - P_z_zk ) + log_K(X, Z, r_[k]) return log_P def _back_off(self): """ Back off Calculate the back off step size Inputs : Z_ : list of states in current proposal r_ : list of back offs in current proposal q : step size reduction dynamic : set to True if you want to use the dynamic back-off Outputs : """ q = self._step_size r = self._r_[-1] Z_ = self._Z_ if self._dynamic: p_0 = la.norm(Z_[0]['f']) dp_0 = p_0*2*la.norm(Z_[0]['J']) p_r = la.norm(Z_[-1]['f']) dp_r = p_0*2*la.norm(Z_[-1]['J']) r_new = optimize(r, p_0**2, dp_0, p_r**2, dp_r) else : r_new = r * q self._r_.append(r_new) def _log_post(self,x,f_x): """ Log of the posterior density This is used to calculete acceptance probability for sampling. Inputs : x : input value f_x : f(x), function value at x Outputs : log(p_x) : log[pi(x)]-||f(x)||^2/(2) log of the posterior probability """ # least squares part -||f(x)||^2/2 log_likelihood = (-la.norm(f_x)**2)/(2.) # prior part -(x-m)'H(x-m)/2 if self._prior: m = self._m H = self._H log_prior = self._ln_H_-np.dot((x-m).T,np.dot(H,x-m))/2. return log_prior+log_likelihood else: return log_likelihood """ Properties: 1. chain 2. n_samples 3. n_accepted 4. accept_rate 5. step_count 6. call_count 7. max_steps 8. step_size """ @property def chain(self): return self._chain @property def n_samples(self): return self._n_samples @property def n_accepted(self): return self._n_accepted @property def accept_rate(self): return float(self._n_accepted)/self._n_samples @property def step_count(self): return self._step_count @property def call_count(self): return self._f.count @property def max_steps(self): return self._max_steps @property def step_size(self): return self._step_size
mugurbil/gnm
gnm/gnm.py
sampler.error_bars
python
def error_bars(self, n_bins, d_min, d_max): # fetch data chain = self._chain len_chain = len(chain) try: n_dims = np.shape(chain)[1] except: n_dims = 1 # begin checks try: assert n_bins == int(n_bins) except: raise TypeError("number of bins has to be an integer") d_min = np.reshape(np.array(d_min), (-1,1)) d_max = np.reshape(np.array(d_max), (-1,1)) try: assert np.size(d_min) == n_dims except: raise TypeError("domain minimum has wrong size") try: assert np.size(d_max) == n_dims except: raise TypeError("domain maximum has wrong size") # end checks # initialize outputs p_x = np.zeros(n_bins) # esitmate of posterior error = np.zeros(n_bins) # error bars x = np.zeros((n_dims, n_bins)) # centers of bins # set dx v = d_max-d_min v_2 = np.dot(v.T, v)[0][0] # bin count for i in xrange(len_chain): bin_no = int(np.floor(np.dot(chain[i].T-d_min,v)/v_2*n_bins)[0]) if n_bins > bin_no > -1: p_x[bin_no] += 1. # end count dx = np.sqrt(v_2)/n_bins p_x = p_x/(len_chain*dx) # find error for i in xrange(n_bins): p = p_x[i] error[i] = np.sqrt(p*(1./dx-p)/(len_chain)) x[:,i] = (d_min+v*(0.5+i)/n_bins)[0] # end find return x, p_x, error
Error Bars create bars and error bars to plot Inputs : n_bins : number of bins plot_range : (shape) = (number of dimensions, 2) matrix which contain the min and max for each dimension as rows Outputs : x : domain p_x : estimated posterior using the chain on the domain error : estimated error for p_x
train
https://github.com/mugurbil/gnm/blob/4f9711fb9d78cc02820c25234bc3ab9615014f11/gnm/gnm.py#L363-L428
null
class sampler(object): def __init__(self, x, model, args): """ Init Initialize the GNM sampler class Inputs : x : initial guess model : user defined data model function args : arguments for the model """ self._args = args self._f = function(model, args) x = np.reshape(np.array(x), (-1,1)) self._n = np.size(x) # size of input space try: x = np.reshape(np.array(x), (-1)) out_x = model(x, self._args) except TypeError as e: raise TypeError(str(e)[:-7]+" needed)") """ except IndexError as e: raise IndexError("initial guess size does not fit model()\n " +str(e)) except Exception as e: print("Error: Model function could not be evaluated.") print(" - Check size of intial guess.") print(" - Check definition of the model.") print(str(e)) print(type(e)) raise RuntimeError("model() could not be evaluated\n ") """ try: chi_x, f_x, J_x = out_x f_x = np.reshape(np.array(f_x), (-1,1)) except: raise TypeError("model() needs to have 3 outputs: chi_x, f_x, J_x") try: assert chi_x == True except AssertionError: raise ValueError("initial guess out of range") try: self._mn = np.size(f_x) J_x = np.array(J_x) assert np.shape(J_x) == (self._mn, self._n) except: raise TypeError("Shape of Jacobian, " + str(np.shape(J_x)) + ", is not correct, (%d, %d)." % (self._mn, self._n)) x = np.reshape(np.array(x), (-1,1)) self._X = {'x':x,'f':f_x,'J':J_x} # state of x # prior parameters self._prior = False # back-off parameters self._max_steps = 1 self._step_size = 0.1 self._dynamic = False self._opts = {} # sampler outputs self._chain = None self._n_samples = 0 self._n_accepted = 0 def prior(self, m, H): """ Set prior Set prior values Inputs : m : mean of the prior H : precision matrix of the prior Hiddens : ln_H_ : log(det(H))/2 calculate this once to use everytime log prior is called Hm : < H, m > calculate this once to use everytime proposal is called """ if self._prior == True: raise Warning("prior information is already set") else: self._prior = True # mean self._m = np.reshape(np.array(m), (-1,1)) try : assert np.size(self._m) == self._n except : raise TypeError("mean has to be an array of size n") # precision self._H = np.array(H) try : assert np.shape(self._H) == (self._n, self._n) except : raise TypeError("precision has to be a matrix of shape n by n") # precalculations self._ln_H_ = np.log(la.det(self._H))/2. self._Hm = np.dot(self._H, self._m) def Jtest(self, x_min, x_max, dx=0.0002, N=1000, eps_max=0.0001, p=2, l_max=50, r=0.5): """ Gradient Checker Test the function's jacobian against the numerical jacobian """ # check inputs x_min and x_max try : assert np.size(x_min) == self._n except : raise TypeError("dimension of x_min, %d, does not match the " "dimension of input, %d" % (np.size(x_min), self._n)) try : assert np.size(x_max) == self._n except : raise TypeError("dimension of x_max, %d, does not match the " "dimension of input, %d." % (np.size(x_max), self._n)) # end checks and call developer function return self._f.Jtest(x_min, x_max, dx=dx, N=N, eps_max=eps_max, p=p, l_max=l_max, r=r) def static(self, max_steps, step_size): """ Set Back-off to Static Set the sampler parameters for static back off Inputs : max_steps : maximum optimization steps to be taken step_size : the step size of the back-off """ self._dynamic = False # begin checks try : self._max_steps = int(max_steps) except : print("Error: Input 1 (max_steps) has to be an int.") return 0 try : assert self._max_steps >= 0 except : print("Warning: Input 1 (max_steps) has to be non-negative.") print("Setting max_steps to 0.") self._max_steps = 0 if max_steps > 0 : try : assert step_size == float(step_size) except AssertionError : print("Warning: Input 2 (step_size) is not a float. Converted.") step_size = float(step_size) except : print("Error: Input 2 (step_size) has to be a float.") return 0 try : assert 0. < step_size < 1. except : print("Warning: Input 2 (step_size) has to be between 0 and 1.") print("Setting step_size to 0.2.") step_size = 0.2 self._step_size = step_size if step_size**max_steps < 10**(-15): print("Warning: Back-off gets dangerously small.") # end checks def dynamic(self, max_steps, opts={}): """ Dynamic Switch Set the sampler parameters for dynamic back off Inputs : max_steps : maximum back-off steps to be taken Optional Inputs: opts : ({}) dictionary containing fancy options """ self._dynamic = True # begin checks try : self._max_steps = int(max_steps) except : raise TypeError("input 1 (max_steps) has to be an integer") return 0 try : assert self._max_steps >= 0 except : raise Warning("input 1 (max_steps) has to be non-negative. Setting (max_steps) to 0.") self._max_steps = 0 self._opts = opts # end checks def sample(self, n_samples, divs=1, visual=False, safe=False): """ Sample Sampling Inputs : n_samples : number of samples to generate Optional Inputs : divs : (1) number of divisions visual : show progress safe : save the chain at every division """ if visual: print("Sampling: 0%") for i in xrange(divs): self._sample(int(n_samples/divs)) if visual: sys.stdout.write("\033[F") # curser up print("Sampling: "+str(int(i*100./divs)+1)+'%') if safe: self.save(path="chain_{:}.dat".format(i)) if n_samples % divs != 0: self._sample(n_samples % divs) if safe: self.save(path="chain_{:}.dat".format(divs)) def save(self, path="chain.dat"): """ Save Save data to file Inputs : path : specifies the path name of the file to be loaded to """ # create dictionary for data dic = {} dic['chain'] = self._chain.tolist() dic['step_count'] = self._step_count.tolist() dic['n_samples'] = self._n_samples dic['n_accepted'] = self._n_accepted dic['x'] = self._X['x'].tolist() dic['f'] = self._X['f'].tolist() dic['J'] = self._X['J'].tolist() # write data to file file = open(path, 'w') json.dump(dic, file) file.close() def load(self, path="chain.dat"): """ Load Load data from file Inputs : path : specifies the path name of the file to be loaded from """ # read data from file file = open(path, 'r') dic = json.load(file) file.close() # get data from dictionary self._chain = np.array(dic['chain']) self._step_count = np.array(dic['step_count']) self._n_samples = dic['n_samples'] self._n_accepted = dic['n_accepted'] self._X = {} self._X['x'] = dic['x'] self._X['f'] = dic['f'] self._X['J'] = dic['J'] def burn(self, n_burned): """ Burn Burn the inital samples to adjust for convergence of the chain cut the first (n_burned) burn-in samples Inputs : chain : the full Markov chain n_burned : number of samples to cut Hidden Outputs : chain : chain with the firt n_burned samples cut """ self._chain = self._chain[n_burned:] def acor(self, k = 5): """ Autocorrelation time of the chain return the autocorrelation time for each parameters Inputs : k : parameter in self-consistent window Outputs : t : autocorrelation time of the chain """ try: import acor except ImportError: print("Can't import acor, please download it.") return 0 n = np.shape(self._chain)[1] t = np.zeros(n) for i in xrange(n): t[i] = acor.acor(self._chain[:,i],k)[0] return t def posterior(self, x): """ Posterior density ** not normalized ** This is used to plot the theoretical curve for tests. Inputs : x : input value Outputs : p : p(x)=pi(x)*exp{-||f(x)||^2/(2)} posterior probability of x """ x = np.reshape(np.array(x), (-1,1)) chi_x, f_x, J_x = self._f(x) if chi_x : p = np.exp(-la.norm(f_x)**2/2.) if self._prior: m = self._m H = self._H p = p * np.exp(-np.dot((x-m).T,np.dot(H,x-m))/2.) return p else : return 0 # end error_bars # internal methods def _sample(self, n_samples): """ Sample Generate samples for posterior distribution using Gauss-Newton proposal parameters Inputs : n_samples : number of samples to generate Hidden Outputs : chain : chain of samples n_samples : length of chain n_accepted : number of proposals accepted step_count : count of the steps accepted """ try : n_samples = int(n_samples) except : raise TypeError("number of samples has to be an integer") exit() # fetch info X = self._proposal_params(self._X) k_max = self._max_steps # initialize chain = np.zeros((n_samples, self._n)) n_accepted = 0 step_count = np.zeros(k_max+2) # begin outer loop for i in xrange(n_samples): accepted = False # check if sample is accepted r_ = [1] # list of step sizes Z_ = [X] # initialize list of Z s self._r_ = r_ log_P_z_x = 0. + X['log_p'] k = 0 # back-off steps taken so far while k <= k_max: # get proposal chi_z = False while not chi_z: z = multi_normal(X, r_[-1]) chi_z, f_z, J_z = self._f(z) Z = self._proposal_params({'x':z,'f':f_z,'J':J_z}) Z_.append(Z) self._Z_ = Z_ log_P_z_x += log_K(Z, X, r_[-1]) # N is the Numerator of the acceptance, N = P_x_z self._N_is_0 = False # check to see if N = 0, to use in _log_P log_N = self._log_P(X, Z, k) # calculating acceptance probability if self._N_is_0 == True : A_z_x = 0. elif log_N >= log_P_z_x : A_z_x = 1. else : A_z_x = np.exp(log_N - log_P_z_x) # acceptance rejection if np.random.rand() <= A_z_x: accepted = True break else : log_P_z_x += np.log(1. - A_z_x) self._back_off() k += 1 # end of steps for loop if accepted == True : chain[i,:] = z[:,0] X = Z # for statistics n_accepted += 1 step_count[k+1] += 1 else : chain[i,:] = X['x'][:,0] # for statistics step_count[0] += 1 # end outer loop # update stored info self._X = X # outputs if self._n_samples == 0 : self._chain = chain self._step_count = step_count else : self._chain = np.append(self._chain, chain, axis=0) self._step_count = np.add(self._step_count, step_count) self._n_samples += n_samples self._n_accepted += n_accepted # end sample def _proposal_params(self, state): """ Proposal parameters Calculate parameters needed for the proposal. Inputs : state : x : the present sample, the place to linearize around f : f(x), function value at x J : f'(x), the jacobian of the function evaluated at x Outputs : state : mu : the mean vector L : the lower triangular cholesky factor of P log_p : log(p(x)) log of the posterior density """ x = state['x'] f = state['f'] J = state['J'] JJ = np.dot(J.T,J) if self._prior: m = self._m H = self._H Hm = self._Hm # LL' = P = H+J'J L = la.cholesky(H+JJ) # mu = (P^-1)(Hm-J'f+J'Jx) mu = la.solve(L.T,la.solve(L,Hm-np.dot(J.T,f)+np.dot(JJ,x))) else: # P = J'J L = la.cholesky(JJ) # mu = x-(P^-1)J'f mu = x-la.solve(L.T,la.solve(L,np.dot(J.T,f))) state['L'] = L state['mu'] = mu state['log_p'] = self._log_post(x,f) return state def _log_P(self, X , Z, k): """ Log of the probability of transition from z to x with k steps log ( P_k (x, z) ) Inputs : X : state to be proposed to Z : state to be proposed from k : number of recursions, depth """ r_ = self._r_ Z_ = self._Z_ # zero case if k == 0 : log_P = Z['log_p'] + log_K(X, Z, r_[k]) # recursice case else : P_zk_z = np.exp( self._log_P(Z_[k], Z, k-1) ) P_z_zk = np.exp( self._log_P(Z, Z_[k], k-1) ) # flag if P_zk_z <= P_z_zk : self._N_is_0 = True log_P = -np.inf else : log_P = np.log( P_zk_z - P_z_zk ) + log_K(X, Z, r_[k]) return log_P def _back_off(self): """ Back off Calculate the back off step size Inputs : Z_ : list of states in current proposal r_ : list of back offs in current proposal q : step size reduction dynamic : set to True if you want to use the dynamic back-off Outputs : """ q = self._step_size r = self._r_[-1] Z_ = self._Z_ if self._dynamic: p_0 = la.norm(Z_[0]['f']) dp_0 = p_0*2*la.norm(Z_[0]['J']) p_r = la.norm(Z_[-1]['f']) dp_r = p_0*2*la.norm(Z_[-1]['J']) r_new = optimize(r, p_0**2, dp_0, p_r**2, dp_r) else : r_new = r * q self._r_.append(r_new) def _log_post(self,x,f_x): """ Log of the posterior density This is used to calculete acceptance probability for sampling. Inputs : x : input value f_x : f(x), function value at x Outputs : log(p_x) : log[pi(x)]-||f(x)||^2/(2) log of the posterior probability """ # least squares part -||f(x)||^2/2 log_likelihood = (-la.norm(f_x)**2)/(2.) # prior part -(x-m)'H(x-m)/2 if self._prior: m = self._m H = self._H log_prior = self._ln_H_-np.dot((x-m).T,np.dot(H,x-m))/2. return log_prior+log_likelihood else: return log_likelihood """ Properties: 1. chain 2. n_samples 3. n_accepted 4. accept_rate 5. step_count 6. call_count 7. max_steps 8. step_size """ @property def chain(self): return self._chain @property def n_samples(self): return self._n_samples @property def n_accepted(self): return self._n_accepted @property def accept_rate(self): return float(self._n_accepted)/self._n_samples @property def step_count(self): return self._step_count @property def call_count(self): return self._f.count @property def max_steps(self): return self._max_steps @property def step_size(self): return self._step_size
mugurbil/gnm
gnm/gnm.py
sampler._sample
python
def _sample(self, n_samples): try : n_samples = int(n_samples) except : raise TypeError("number of samples has to be an integer") exit() # fetch info X = self._proposal_params(self._X) k_max = self._max_steps # initialize chain = np.zeros((n_samples, self._n)) n_accepted = 0 step_count = np.zeros(k_max+2) # begin outer loop for i in xrange(n_samples): accepted = False # check if sample is accepted r_ = [1] # list of step sizes Z_ = [X] # initialize list of Z s self._r_ = r_ log_P_z_x = 0. + X['log_p'] k = 0 # back-off steps taken so far while k <= k_max: # get proposal chi_z = False while not chi_z: z = multi_normal(X, r_[-1]) chi_z, f_z, J_z = self._f(z) Z = self._proposal_params({'x':z,'f':f_z,'J':J_z}) Z_.append(Z) self._Z_ = Z_ log_P_z_x += log_K(Z, X, r_[-1]) # N is the Numerator of the acceptance, N = P_x_z self._N_is_0 = False # check to see if N = 0, to use in _log_P log_N = self._log_P(X, Z, k) # calculating acceptance probability if self._N_is_0 == True : A_z_x = 0. elif log_N >= log_P_z_x : A_z_x = 1. else : A_z_x = np.exp(log_N - log_P_z_x) # acceptance rejection if np.random.rand() <= A_z_x: accepted = True break else : log_P_z_x += np.log(1. - A_z_x) self._back_off() k += 1 # end of steps for loop if accepted == True : chain[i,:] = z[:,0] X = Z # for statistics n_accepted += 1 step_count[k+1] += 1 else : chain[i,:] = X['x'][:,0] # for statistics step_count[0] += 1 # end outer loop # update stored info self._X = X # outputs if self._n_samples == 0 : self._chain = chain self._step_count = step_count else : self._chain = np.append(self._chain, chain, axis=0) self._step_count = np.add(self._step_count, step_count) self._n_samples += n_samples self._n_accepted += n_accepted
Sample Generate samples for posterior distribution using Gauss-Newton proposal parameters Inputs : n_samples : number of samples to generate Hidden Outputs : chain : chain of samples n_samples : length of chain n_accepted : number of proposals accepted step_count : count of the steps accepted
train
https://github.com/mugurbil/gnm/blob/4f9711fb9d78cc02820c25234bc3ab9615014f11/gnm/gnm.py#L432-L530
[ "def multi_normal(X, t):\n \"\"\"\n Multivariate normal sampler:\n Generates normal samples with mean m, precision matrix LL' \n Inputs:\n x : \n propose from \n Outputs:\n normal with mean m and precision LL'\n \"\"\"\n m, L = update_params(X, t)\n z = np.random.standard_normal(np.shape(m)) # generate i.i.d N(0,1)\n return la.solve(L.T,z)+m\n", "def log_K(Z, X, t):\n \"\"\"\n Log K\n Log of the proposal probability density function for gnm\n Inputs :\n Z :\n proposed to\n x : \n proposed from \n Outputs : \n log of the probability density function\n \"\"\"\n m, L = update_params(X, t)\n z = Z['x']\n return np.log(det(L))-la.norm(np.dot(L.T,z-m))**2/2. \n", "def _proposal_params(self, state):\n \"\"\"\nProposal parameters\n Calculate parameters needed for the proposal. \nInputs :\n state : \n x : \n the present sample, the place to linearize around\n f : f(x), \n function value at x\n J : f'(x), \n the jacobian of the function evaluated at x\nOutputs :\n state :\n mu : \n the mean vector\n L :\n the lower triangular cholesky factor of P \n log_p : log(p(x))\n log of the posterior density\n \"\"\"\n x = state['x']\n f = state['f']\n J = state['J']\n JJ = np.dot(J.T,J) \n\n if self._prior: \n m = self._m\n H = self._H\n Hm = self._Hm\n # LL' = P = H+J'J \n L = la.cholesky(H+JJ) \n # mu = (P^-1)(Hm-J'f+J'Jx)\n mu = la.solve(L.T,la.solve(L,Hm-np.dot(J.T,f)+np.dot(JJ,x))) \n else: \n # P = J'J\n L = la.cholesky(JJ)\n # mu = x-(P^-1)J'f\n mu = x-la.solve(L.T,la.solve(L,np.dot(J.T,f)))\n\n state['L'] = L \n state['mu'] = mu\n state['log_p'] = self._log_post(x,f)\n return state\n", "def _log_P(self, X , Z, k):\n \"\"\"\nLog of the probability of transition from z to x with k steps\n log ( P_k (x, z) )\nInputs : \n X :\n state to be proposed to\n Z : \n state to be proposed from\n k : \n number of recursions, depth\n \"\"\"\n r_ = self._r_\n Z_ = self._Z_\n # zero case\n if k == 0 :\n log_P = Z['log_p'] + log_K(X, Z, r_[k])\n # recursice case\n else :\n P_zk_z = np.exp( self._log_P(Z_[k], Z, k-1) )\n P_z_zk = np.exp( self._log_P(Z, Z_[k], k-1) ) \n # flag\n if P_zk_z <= P_z_zk :\n self._N_is_0 = True\n log_P = -np.inf\n else : \n log_P = np.log( P_zk_z - P_z_zk ) + log_K(X, Z, r_[k])\n return log_P\n" ]
class sampler(object): def __init__(self, x, model, args): """ Init Initialize the GNM sampler class Inputs : x : initial guess model : user defined data model function args : arguments for the model """ self._args = args self._f = function(model, args) x = np.reshape(np.array(x), (-1,1)) self._n = np.size(x) # size of input space try: x = np.reshape(np.array(x), (-1)) out_x = model(x, self._args) except TypeError as e: raise TypeError(str(e)[:-7]+" needed)") """ except IndexError as e: raise IndexError("initial guess size does not fit model()\n " +str(e)) except Exception as e: print("Error: Model function could not be evaluated.") print(" - Check size of intial guess.") print(" - Check definition of the model.") print(str(e)) print(type(e)) raise RuntimeError("model() could not be evaluated\n ") """ try: chi_x, f_x, J_x = out_x f_x = np.reshape(np.array(f_x), (-1,1)) except: raise TypeError("model() needs to have 3 outputs: chi_x, f_x, J_x") try: assert chi_x == True except AssertionError: raise ValueError("initial guess out of range") try: self._mn = np.size(f_x) J_x = np.array(J_x) assert np.shape(J_x) == (self._mn, self._n) except: raise TypeError("Shape of Jacobian, " + str(np.shape(J_x)) + ", is not correct, (%d, %d)." % (self._mn, self._n)) x = np.reshape(np.array(x), (-1,1)) self._X = {'x':x,'f':f_x,'J':J_x} # state of x # prior parameters self._prior = False # back-off parameters self._max_steps = 1 self._step_size = 0.1 self._dynamic = False self._opts = {} # sampler outputs self._chain = None self._n_samples = 0 self._n_accepted = 0 def prior(self, m, H): """ Set prior Set prior values Inputs : m : mean of the prior H : precision matrix of the prior Hiddens : ln_H_ : log(det(H))/2 calculate this once to use everytime log prior is called Hm : < H, m > calculate this once to use everytime proposal is called """ if self._prior == True: raise Warning("prior information is already set") else: self._prior = True # mean self._m = np.reshape(np.array(m), (-1,1)) try : assert np.size(self._m) == self._n except : raise TypeError("mean has to be an array of size n") # precision self._H = np.array(H) try : assert np.shape(self._H) == (self._n, self._n) except : raise TypeError("precision has to be a matrix of shape n by n") # precalculations self._ln_H_ = np.log(la.det(self._H))/2. self._Hm = np.dot(self._H, self._m) def Jtest(self, x_min, x_max, dx=0.0002, N=1000, eps_max=0.0001, p=2, l_max=50, r=0.5): """ Gradient Checker Test the function's jacobian against the numerical jacobian """ # check inputs x_min and x_max try : assert np.size(x_min) == self._n except : raise TypeError("dimension of x_min, %d, does not match the " "dimension of input, %d" % (np.size(x_min), self._n)) try : assert np.size(x_max) == self._n except : raise TypeError("dimension of x_max, %d, does not match the " "dimension of input, %d." % (np.size(x_max), self._n)) # end checks and call developer function return self._f.Jtest(x_min, x_max, dx=dx, N=N, eps_max=eps_max, p=p, l_max=l_max, r=r) def static(self, max_steps, step_size): """ Set Back-off to Static Set the sampler parameters for static back off Inputs : max_steps : maximum optimization steps to be taken step_size : the step size of the back-off """ self._dynamic = False # begin checks try : self._max_steps = int(max_steps) except : print("Error: Input 1 (max_steps) has to be an int.") return 0 try : assert self._max_steps >= 0 except : print("Warning: Input 1 (max_steps) has to be non-negative.") print("Setting max_steps to 0.") self._max_steps = 0 if max_steps > 0 : try : assert step_size == float(step_size) except AssertionError : print("Warning: Input 2 (step_size) is not a float. Converted.") step_size = float(step_size) except : print("Error: Input 2 (step_size) has to be a float.") return 0 try : assert 0. < step_size < 1. except : print("Warning: Input 2 (step_size) has to be between 0 and 1.") print("Setting step_size to 0.2.") step_size = 0.2 self._step_size = step_size if step_size**max_steps < 10**(-15): print("Warning: Back-off gets dangerously small.") # end checks def dynamic(self, max_steps, opts={}): """ Dynamic Switch Set the sampler parameters for dynamic back off Inputs : max_steps : maximum back-off steps to be taken Optional Inputs: opts : ({}) dictionary containing fancy options """ self._dynamic = True # begin checks try : self._max_steps = int(max_steps) except : raise TypeError("input 1 (max_steps) has to be an integer") return 0 try : assert self._max_steps >= 0 except : raise Warning("input 1 (max_steps) has to be non-negative. Setting (max_steps) to 0.") self._max_steps = 0 self._opts = opts # end checks def sample(self, n_samples, divs=1, visual=False, safe=False): """ Sample Sampling Inputs : n_samples : number of samples to generate Optional Inputs : divs : (1) number of divisions visual : show progress safe : save the chain at every division """ if visual: print("Sampling: 0%") for i in xrange(divs): self._sample(int(n_samples/divs)) if visual: sys.stdout.write("\033[F") # curser up print("Sampling: "+str(int(i*100./divs)+1)+'%') if safe: self.save(path="chain_{:}.dat".format(i)) if n_samples % divs != 0: self._sample(n_samples % divs) if safe: self.save(path="chain_{:}.dat".format(divs)) def save(self, path="chain.dat"): """ Save Save data to file Inputs : path : specifies the path name of the file to be loaded to """ # create dictionary for data dic = {} dic['chain'] = self._chain.tolist() dic['step_count'] = self._step_count.tolist() dic['n_samples'] = self._n_samples dic['n_accepted'] = self._n_accepted dic['x'] = self._X['x'].tolist() dic['f'] = self._X['f'].tolist() dic['J'] = self._X['J'].tolist() # write data to file file = open(path, 'w') json.dump(dic, file) file.close() def load(self, path="chain.dat"): """ Load Load data from file Inputs : path : specifies the path name of the file to be loaded from """ # read data from file file = open(path, 'r') dic = json.load(file) file.close() # get data from dictionary self._chain = np.array(dic['chain']) self._step_count = np.array(dic['step_count']) self._n_samples = dic['n_samples'] self._n_accepted = dic['n_accepted'] self._X = {} self._X['x'] = dic['x'] self._X['f'] = dic['f'] self._X['J'] = dic['J'] def burn(self, n_burned): """ Burn Burn the inital samples to adjust for convergence of the chain cut the first (n_burned) burn-in samples Inputs : chain : the full Markov chain n_burned : number of samples to cut Hidden Outputs : chain : chain with the firt n_burned samples cut """ self._chain = self._chain[n_burned:] def acor(self, k = 5): """ Autocorrelation time of the chain return the autocorrelation time for each parameters Inputs : k : parameter in self-consistent window Outputs : t : autocorrelation time of the chain """ try: import acor except ImportError: print("Can't import acor, please download it.") return 0 n = np.shape(self._chain)[1] t = np.zeros(n) for i in xrange(n): t[i] = acor.acor(self._chain[:,i],k)[0] return t def posterior(self, x): """ Posterior density ** not normalized ** This is used to plot the theoretical curve for tests. Inputs : x : input value Outputs : p : p(x)=pi(x)*exp{-||f(x)||^2/(2)} posterior probability of x """ x = np.reshape(np.array(x), (-1,1)) chi_x, f_x, J_x = self._f(x) if chi_x : p = np.exp(-la.norm(f_x)**2/2.) if self._prior: m = self._m H = self._H p = p * np.exp(-np.dot((x-m).T,np.dot(H,x-m))/2.) return p else : return 0 def error_bars(self, n_bins, d_min, d_max): """ Error Bars create bars and error bars to plot Inputs : n_bins : number of bins plot_range : (shape) = (number of dimensions, 2) matrix which contain the min and max for each dimension as rows Outputs : x : domain p_x : estimated posterior using the chain on the domain error : estimated error for p_x """ # fetch data chain = self._chain len_chain = len(chain) try: n_dims = np.shape(chain)[1] except: n_dims = 1 # begin checks try: assert n_bins == int(n_bins) except: raise TypeError("number of bins has to be an integer") d_min = np.reshape(np.array(d_min), (-1,1)) d_max = np.reshape(np.array(d_max), (-1,1)) try: assert np.size(d_min) == n_dims except: raise TypeError("domain minimum has wrong size") try: assert np.size(d_max) == n_dims except: raise TypeError("domain maximum has wrong size") # end checks # initialize outputs p_x = np.zeros(n_bins) # esitmate of posterior error = np.zeros(n_bins) # error bars x = np.zeros((n_dims, n_bins)) # centers of bins # set dx v = d_max-d_min v_2 = np.dot(v.T, v)[0][0] # bin count for i in xrange(len_chain): bin_no = int(np.floor(np.dot(chain[i].T-d_min,v)/v_2*n_bins)[0]) if n_bins > bin_no > -1: p_x[bin_no] += 1. # end count dx = np.sqrt(v_2)/n_bins p_x = p_x/(len_chain*dx) # find error for i in xrange(n_bins): p = p_x[i] error[i] = np.sqrt(p*(1./dx-p)/(len_chain)) x[:,i] = (d_min+v*(0.5+i)/n_bins)[0] # end find return x, p_x, error # end error_bars # internal methods # end sample def _proposal_params(self, state): """ Proposal parameters Calculate parameters needed for the proposal. Inputs : state : x : the present sample, the place to linearize around f : f(x), function value at x J : f'(x), the jacobian of the function evaluated at x Outputs : state : mu : the mean vector L : the lower triangular cholesky factor of P log_p : log(p(x)) log of the posterior density """ x = state['x'] f = state['f'] J = state['J'] JJ = np.dot(J.T,J) if self._prior: m = self._m H = self._H Hm = self._Hm # LL' = P = H+J'J L = la.cholesky(H+JJ) # mu = (P^-1)(Hm-J'f+J'Jx) mu = la.solve(L.T,la.solve(L,Hm-np.dot(J.T,f)+np.dot(JJ,x))) else: # P = J'J L = la.cholesky(JJ) # mu = x-(P^-1)J'f mu = x-la.solve(L.T,la.solve(L,np.dot(J.T,f))) state['L'] = L state['mu'] = mu state['log_p'] = self._log_post(x,f) return state def _log_P(self, X , Z, k): """ Log of the probability of transition from z to x with k steps log ( P_k (x, z) ) Inputs : X : state to be proposed to Z : state to be proposed from k : number of recursions, depth """ r_ = self._r_ Z_ = self._Z_ # zero case if k == 0 : log_P = Z['log_p'] + log_K(X, Z, r_[k]) # recursice case else : P_zk_z = np.exp( self._log_P(Z_[k], Z, k-1) ) P_z_zk = np.exp( self._log_P(Z, Z_[k], k-1) ) # flag if P_zk_z <= P_z_zk : self._N_is_0 = True log_P = -np.inf else : log_P = np.log( P_zk_z - P_z_zk ) + log_K(X, Z, r_[k]) return log_P def _back_off(self): """ Back off Calculate the back off step size Inputs : Z_ : list of states in current proposal r_ : list of back offs in current proposal q : step size reduction dynamic : set to True if you want to use the dynamic back-off Outputs : """ q = self._step_size r = self._r_[-1] Z_ = self._Z_ if self._dynamic: p_0 = la.norm(Z_[0]['f']) dp_0 = p_0*2*la.norm(Z_[0]['J']) p_r = la.norm(Z_[-1]['f']) dp_r = p_0*2*la.norm(Z_[-1]['J']) r_new = optimize(r, p_0**2, dp_0, p_r**2, dp_r) else : r_new = r * q self._r_.append(r_new) def _log_post(self,x,f_x): """ Log of the posterior density This is used to calculete acceptance probability for sampling. Inputs : x : input value f_x : f(x), function value at x Outputs : log(p_x) : log[pi(x)]-||f(x)||^2/(2) log of the posterior probability """ # least squares part -||f(x)||^2/2 log_likelihood = (-la.norm(f_x)**2)/(2.) # prior part -(x-m)'H(x-m)/2 if self._prior: m = self._m H = self._H log_prior = self._ln_H_-np.dot((x-m).T,np.dot(H,x-m))/2. return log_prior+log_likelihood else: return log_likelihood """ Properties: 1. chain 2. n_samples 3. n_accepted 4. accept_rate 5. step_count 6. call_count 7. max_steps 8. step_size """ @property def chain(self): return self._chain @property def n_samples(self): return self._n_samples @property def n_accepted(self): return self._n_accepted @property def accept_rate(self): return float(self._n_accepted)/self._n_samples @property def step_count(self): return self._step_count @property def call_count(self): return self._f.count @property def max_steps(self): return self._max_steps @property def step_size(self): return self._step_size
mugurbil/gnm
gnm/gnm.py
sampler._proposal_params
python
def _proposal_params(self, state): x = state['x'] f = state['f'] J = state['J'] JJ = np.dot(J.T,J) if self._prior: m = self._m H = self._H Hm = self._Hm # LL' = P = H+J'J L = la.cholesky(H+JJ) # mu = (P^-1)(Hm-J'f+J'Jx) mu = la.solve(L.T,la.solve(L,Hm-np.dot(J.T,f)+np.dot(JJ,x))) else: # P = J'J L = la.cholesky(JJ) # mu = x-(P^-1)J'f mu = x-la.solve(L.T,la.solve(L,np.dot(J.T,f))) state['L'] = L state['mu'] = mu state['log_p'] = self._log_post(x,f) return state
Proposal parameters Calculate parameters needed for the proposal. Inputs : state : x : the present sample, the place to linearize around f : f(x), function value at x J : f'(x), the jacobian of the function evaluated at x Outputs : state : mu : the mean vector L : the lower triangular cholesky factor of P log_p : log(p(x)) log of the posterior density
train
https://github.com/mugurbil/gnm/blob/4f9711fb9d78cc02820c25234bc3ab9615014f11/gnm/gnm.py#L533-L576
null
class sampler(object): def __init__(self, x, model, args): """ Init Initialize the GNM sampler class Inputs : x : initial guess model : user defined data model function args : arguments for the model """ self._args = args self._f = function(model, args) x = np.reshape(np.array(x), (-1,1)) self._n = np.size(x) # size of input space try: x = np.reshape(np.array(x), (-1)) out_x = model(x, self._args) except TypeError as e: raise TypeError(str(e)[:-7]+" needed)") """ except IndexError as e: raise IndexError("initial guess size does not fit model()\n " +str(e)) except Exception as e: print("Error: Model function could not be evaluated.") print(" - Check size of intial guess.") print(" - Check definition of the model.") print(str(e)) print(type(e)) raise RuntimeError("model() could not be evaluated\n ") """ try: chi_x, f_x, J_x = out_x f_x = np.reshape(np.array(f_x), (-1,1)) except: raise TypeError("model() needs to have 3 outputs: chi_x, f_x, J_x") try: assert chi_x == True except AssertionError: raise ValueError("initial guess out of range") try: self._mn = np.size(f_x) J_x = np.array(J_x) assert np.shape(J_x) == (self._mn, self._n) except: raise TypeError("Shape of Jacobian, " + str(np.shape(J_x)) + ", is not correct, (%d, %d)." % (self._mn, self._n)) x = np.reshape(np.array(x), (-1,1)) self._X = {'x':x,'f':f_x,'J':J_x} # state of x # prior parameters self._prior = False # back-off parameters self._max_steps = 1 self._step_size = 0.1 self._dynamic = False self._opts = {} # sampler outputs self._chain = None self._n_samples = 0 self._n_accepted = 0 def prior(self, m, H): """ Set prior Set prior values Inputs : m : mean of the prior H : precision matrix of the prior Hiddens : ln_H_ : log(det(H))/2 calculate this once to use everytime log prior is called Hm : < H, m > calculate this once to use everytime proposal is called """ if self._prior == True: raise Warning("prior information is already set") else: self._prior = True # mean self._m = np.reshape(np.array(m), (-1,1)) try : assert np.size(self._m) == self._n except : raise TypeError("mean has to be an array of size n") # precision self._H = np.array(H) try : assert np.shape(self._H) == (self._n, self._n) except : raise TypeError("precision has to be a matrix of shape n by n") # precalculations self._ln_H_ = np.log(la.det(self._H))/2. self._Hm = np.dot(self._H, self._m) def Jtest(self, x_min, x_max, dx=0.0002, N=1000, eps_max=0.0001, p=2, l_max=50, r=0.5): """ Gradient Checker Test the function's jacobian against the numerical jacobian """ # check inputs x_min and x_max try : assert np.size(x_min) == self._n except : raise TypeError("dimension of x_min, %d, does not match the " "dimension of input, %d" % (np.size(x_min), self._n)) try : assert np.size(x_max) == self._n except : raise TypeError("dimension of x_max, %d, does not match the " "dimension of input, %d." % (np.size(x_max), self._n)) # end checks and call developer function return self._f.Jtest(x_min, x_max, dx=dx, N=N, eps_max=eps_max, p=p, l_max=l_max, r=r) def static(self, max_steps, step_size): """ Set Back-off to Static Set the sampler parameters for static back off Inputs : max_steps : maximum optimization steps to be taken step_size : the step size of the back-off """ self._dynamic = False # begin checks try : self._max_steps = int(max_steps) except : print("Error: Input 1 (max_steps) has to be an int.") return 0 try : assert self._max_steps >= 0 except : print("Warning: Input 1 (max_steps) has to be non-negative.") print("Setting max_steps to 0.") self._max_steps = 0 if max_steps > 0 : try : assert step_size == float(step_size) except AssertionError : print("Warning: Input 2 (step_size) is not a float. Converted.") step_size = float(step_size) except : print("Error: Input 2 (step_size) has to be a float.") return 0 try : assert 0. < step_size < 1. except : print("Warning: Input 2 (step_size) has to be between 0 and 1.") print("Setting step_size to 0.2.") step_size = 0.2 self._step_size = step_size if step_size**max_steps < 10**(-15): print("Warning: Back-off gets dangerously small.") # end checks def dynamic(self, max_steps, opts={}): """ Dynamic Switch Set the sampler parameters for dynamic back off Inputs : max_steps : maximum back-off steps to be taken Optional Inputs: opts : ({}) dictionary containing fancy options """ self._dynamic = True # begin checks try : self._max_steps = int(max_steps) except : raise TypeError("input 1 (max_steps) has to be an integer") return 0 try : assert self._max_steps >= 0 except : raise Warning("input 1 (max_steps) has to be non-negative. Setting (max_steps) to 0.") self._max_steps = 0 self._opts = opts # end checks def sample(self, n_samples, divs=1, visual=False, safe=False): """ Sample Sampling Inputs : n_samples : number of samples to generate Optional Inputs : divs : (1) number of divisions visual : show progress safe : save the chain at every division """ if visual: print("Sampling: 0%") for i in xrange(divs): self._sample(int(n_samples/divs)) if visual: sys.stdout.write("\033[F") # curser up print("Sampling: "+str(int(i*100./divs)+1)+'%') if safe: self.save(path="chain_{:}.dat".format(i)) if n_samples % divs != 0: self._sample(n_samples % divs) if safe: self.save(path="chain_{:}.dat".format(divs)) def save(self, path="chain.dat"): """ Save Save data to file Inputs : path : specifies the path name of the file to be loaded to """ # create dictionary for data dic = {} dic['chain'] = self._chain.tolist() dic['step_count'] = self._step_count.tolist() dic['n_samples'] = self._n_samples dic['n_accepted'] = self._n_accepted dic['x'] = self._X['x'].tolist() dic['f'] = self._X['f'].tolist() dic['J'] = self._X['J'].tolist() # write data to file file = open(path, 'w') json.dump(dic, file) file.close() def load(self, path="chain.dat"): """ Load Load data from file Inputs : path : specifies the path name of the file to be loaded from """ # read data from file file = open(path, 'r') dic = json.load(file) file.close() # get data from dictionary self._chain = np.array(dic['chain']) self._step_count = np.array(dic['step_count']) self._n_samples = dic['n_samples'] self._n_accepted = dic['n_accepted'] self._X = {} self._X['x'] = dic['x'] self._X['f'] = dic['f'] self._X['J'] = dic['J'] def burn(self, n_burned): """ Burn Burn the inital samples to adjust for convergence of the chain cut the first (n_burned) burn-in samples Inputs : chain : the full Markov chain n_burned : number of samples to cut Hidden Outputs : chain : chain with the firt n_burned samples cut """ self._chain = self._chain[n_burned:] def acor(self, k = 5): """ Autocorrelation time of the chain return the autocorrelation time for each parameters Inputs : k : parameter in self-consistent window Outputs : t : autocorrelation time of the chain """ try: import acor except ImportError: print("Can't import acor, please download it.") return 0 n = np.shape(self._chain)[1] t = np.zeros(n) for i in xrange(n): t[i] = acor.acor(self._chain[:,i],k)[0] return t def posterior(self, x): """ Posterior density ** not normalized ** This is used to plot the theoretical curve for tests. Inputs : x : input value Outputs : p : p(x)=pi(x)*exp{-||f(x)||^2/(2)} posterior probability of x """ x = np.reshape(np.array(x), (-1,1)) chi_x, f_x, J_x = self._f(x) if chi_x : p = np.exp(-la.norm(f_x)**2/2.) if self._prior: m = self._m H = self._H p = p * np.exp(-np.dot((x-m).T,np.dot(H,x-m))/2.) return p else : return 0 def error_bars(self, n_bins, d_min, d_max): """ Error Bars create bars and error bars to plot Inputs : n_bins : number of bins plot_range : (shape) = (number of dimensions, 2) matrix which contain the min and max for each dimension as rows Outputs : x : domain p_x : estimated posterior using the chain on the domain error : estimated error for p_x """ # fetch data chain = self._chain len_chain = len(chain) try: n_dims = np.shape(chain)[1] except: n_dims = 1 # begin checks try: assert n_bins == int(n_bins) except: raise TypeError("number of bins has to be an integer") d_min = np.reshape(np.array(d_min), (-1,1)) d_max = np.reshape(np.array(d_max), (-1,1)) try: assert np.size(d_min) == n_dims except: raise TypeError("domain minimum has wrong size") try: assert np.size(d_max) == n_dims except: raise TypeError("domain maximum has wrong size") # end checks # initialize outputs p_x = np.zeros(n_bins) # esitmate of posterior error = np.zeros(n_bins) # error bars x = np.zeros((n_dims, n_bins)) # centers of bins # set dx v = d_max-d_min v_2 = np.dot(v.T, v)[0][0] # bin count for i in xrange(len_chain): bin_no = int(np.floor(np.dot(chain[i].T-d_min,v)/v_2*n_bins)[0]) if n_bins > bin_no > -1: p_x[bin_no] += 1. # end count dx = np.sqrt(v_2)/n_bins p_x = p_x/(len_chain*dx) # find error for i in xrange(n_bins): p = p_x[i] error[i] = np.sqrt(p*(1./dx-p)/(len_chain)) x[:,i] = (d_min+v*(0.5+i)/n_bins)[0] # end find return x, p_x, error # end error_bars # internal methods def _sample(self, n_samples): """ Sample Generate samples for posterior distribution using Gauss-Newton proposal parameters Inputs : n_samples : number of samples to generate Hidden Outputs : chain : chain of samples n_samples : length of chain n_accepted : number of proposals accepted step_count : count of the steps accepted """ try : n_samples = int(n_samples) except : raise TypeError("number of samples has to be an integer") exit() # fetch info X = self._proposal_params(self._X) k_max = self._max_steps # initialize chain = np.zeros((n_samples, self._n)) n_accepted = 0 step_count = np.zeros(k_max+2) # begin outer loop for i in xrange(n_samples): accepted = False # check if sample is accepted r_ = [1] # list of step sizes Z_ = [X] # initialize list of Z s self._r_ = r_ log_P_z_x = 0. + X['log_p'] k = 0 # back-off steps taken so far while k <= k_max: # get proposal chi_z = False while not chi_z: z = multi_normal(X, r_[-1]) chi_z, f_z, J_z = self._f(z) Z = self._proposal_params({'x':z,'f':f_z,'J':J_z}) Z_.append(Z) self._Z_ = Z_ log_P_z_x += log_K(Z, X, r_[-1]) # N is the Numerator of the acceptance, N = P_x_z self._N_is_0 = False # check to see if N = 0, to use in _log_P log_N = self._log_P(X, Z, k) # calculating acceptance probability if self._N_is_0 == True : A_z_x = 0. elif log_N >= log_P_z_x : A_z_x = 1. else : A_z_x = np.exp(log_N - log_P_z_x) # acceptance rejection if np.random.rand() <= A_z_x: accepted = True break else : log_P_z_x += np.log(1. - A_z_x) self._back_off() k += 1 # end of steps for loop if accepted == True : chain[i,:] = z[:,0] X = Z # for statistics n_accepted += 1 step_count[k+1] += 1 else : chain[i,:] = X['x'][:,0] # for statistics step_count[0] += 1 # end outer loop # update stored info self._X = X # outputs if self._n_samples == 0 : self._chain = chain self._step_count = step_count else : self._chain = np.append(self._chain, chain, axis=0) self._step_count = np.add(self._step_count, step_count) self._n_samples += n_samples self._n_accepted += n_accepted # end sample def _log_P(self, X , Z, k): """ Log of the probability of transition from z to x with k steps log ( P_k (x, z) ) Inputs : X : state to be proposed to Z : state to be proposed from k : number of recursions, depth """ r_ = self._r_ Z_ = self._Z_ # zero case if k == 0 : log_P = Z['log_p'] + log_K(X, Z, r_[k]) # recursice case else : P_zk_z = np.exp( self._log_P(Z_[k], Z, k-1) ) P_z_zk = np.exp( self._log_P(Z, Z_[k], k-1) ) # flag if P_zk_z <= P_z_zk : self._N_is_0 = True log_P = -np.inf else : log_P = np.log( P_zk_z - P_z_zk ) + log_K(X, Z, r_[k]) return log_P def _back_off(self): """ Back off Calculate the back off step size Inputs : Z_ : list of states in current proposal r_ : list of back offs in current proposal q : step size reduction dynamic : set to True if you want to use the dynamic back-off Outputs : """ q = self._step_size r = self._r_[-1] Z_ = self._Z_ if self._dynamic: p_0 = la.norm(Z_[0]['f']) dp_0 = p_0*2*la.norm(Z_[0]['J']) p_r = la.norm(Z_[-1]['f']) dp_r = p_0*2*la.norm(Z_[-1]['J']) r_new = optimize(r, p_0**2, dp_0, p_r**2, dp_r) else : r_new = r * q self._r_.append(r_new) def _log_post(self,x,f_x): """ Log of the posterior density This is used to calculete acceptance probability for sampling. Inputs : x : input value f_x : f(x), function value at x Outputs : log(p_x) : log[pi(x)]-||f(x)||^2/(2) log of the posterior probability """ # least squares part -||f(x)||^2/2 log_likelihood = (-la.norm(f_x)**2)/(2.) # prior part -(x-m)'H(x-m)/2 if self._prior: m = self._m H = self._H log_prior = self._ln_H_-np.dot((x-m).T,np.dot(H,x-m))/2. return log_prior+log_likelihood else: return log_likelihood """ Properties: 1. chain 2. n_samples 3. n_accepted 4. accept_rate 5. step_count 6. call_count 7. max_steps 8. step_size """ @property def chain(self): return self._chain @property def n_samples(self): return self._n_samples @property def n_accepted(self): return self._n_accepted @property def accept_rate(self): return float(self._n_accepted)/self._n_samples @property def step_count(self): return self._step_count @property def call_count(self): return self._f.count @property def max_steps(self): return self._max_steps @property def step_size(self): return self._step_size
mugurbil/gnm
gnm/gnm.py
sampler._log_P
python
def _log_P(self, X , Z, k): r_ = self._r_ Z_ = self._Z_ # zero case if k == 0 : log_P = Z['log_p'] + log_K(X, Z, r_[k]) # recursice case else : P_zk_z = np.exp( self._log_P(Z_[k], Z, k-1) ) P_z_zk = np.exp( self._log_P(Z, Z_[k], k-1) ) # flag if P_zk_z <= P_z_zk : self._N_is_0 = True log_P = -np.inf else : log_P = np.log( P_zk_z - P_z_zk ) + log_K(X, Z, r_[k]) return log_P
Log of the probability of transition from z to x with k steps log ( P_k (x, z) ) Inputs : X : state to be proposed to Z : state to be proposed from k : number of recursions, depth
train
https://github.com/mugurbil/gnm/blob/4f9711fb9d78cc02820c25234bc3ab9615014f11/gnm/gnm.py#L578-L605
null
class sampler(object): def __init__(self, x, model, args): """ Init Initialize the GNM sampler class Inputs : x : initial guess model : user defined data model function args : arguments for the model """ self._args = args self._f = function(model, args) x = np.reshape(np.array(x), (-1,1)) self._n = np.size(x) # size of input space try: x = np.reshape(np.array(x), (-1)) out_x = model(x, self._args) except TypeError as e: raise TypeError(str(e)[:-7]+" needed)") """ except IndexError as e: raise IndexError("initial guess size does not fit model()\n " +str(e)) except Exception as e: print("Error: Model function could not be evaluated.") print(" - Check size of intial guess.") print(" - Check definition of the model.") print(str(e)) print(type(e)) raise RuntimeError("model() could not be evaluated\n ") """ try: chi_x, f_x, J_x = out_x f_x = np.reshape(np.array(f_x), (-1,1)) except: raise TypeError("model() needs to have 3 outputs: chi_x, f_x, J_x") try: assert chi_x == True except AssertionError: raise ValueError("initial guess out of range") try: self._mn = np.size(f_x) J_x = np.array(J_x) assert np.shape(J_x) == (self._mn, self._n) except: raise TypeError("Shape of Jacobian, " + str(np.shape(J_x)) + ", is not correct, (%d, %d)." % (self._mn, self._n)) x = np.reshape(np.array(x), (-1,1)) self._X = {'x':x,'f':f_x,'J':J_x} # state of x # prior parameters self._prior = False # back-off parameters self._max_steps = 1 self._step_size = 0.1 self._dynamic = False self._opts = {} # sampler outputs self._chain = None self._n_samples = 0 self._n_accepted = 0 def prior(self, m, H): """ Set prior Set prior values Inputs : m : mean of the prior H : precision matrix of the prior Hiddens : ln_H_ : log(det(H))/2 calculate this once to use everytime log prior is called Hm : < H, m > calculate this once to use everytime proposal is called """ if self._prior == True: raise Warning("prior information is already set") else: self._prior = True # mean self._m = np.reshape(np.array(m), (-1,1)) try : assert np.size(self._m) == self._n except : raise TypeError("mean has to be an array of size n") # precision self._H = np.array(H) try : assert np.shape(self._H) == (self._n, self._n) except : raise TypeError("precision has to be a matrix of shape n by n") # precalculations self._ln_H_ = np.log(la.det(self._H))/2. self._Hm = np.dot(self._H, self._m) def Jtest(self, x_min, x_max, dx=0.0002, N=1000, eps_max=0.0001, p=2, l_max=50, r=0.5): """ Gradient Checker Test the function's jacobian against the numerical jacobian """ # check inputs x_min and x_max try : assert np.size(x_min) == self._n except : raise TypeError("dimension of x_min, %d, does not match the " "dimension of input, %d" % (np.size(x_min), self._n)) try : assert np.size(x_max) == self._n except : raise TypeError("dimension of x_max, %d, does not match the " "dimension of input, %d." % (np.size(x_max), self._n)) # end checks and call developer function return self._f.Jtest(x_min, x_max, dx=dx, N=N, eps_max=eps_max, p=p, l_max=l_max, r=r) def static(self, max_steps, step_size): """ Set Back-off to Static Set the sampler parameters for static back off Inputs : max_steps : maximum optimization steps to be taken step_size : the step size of the back-off """ self._dynamic = False # begin checks try : self._max_steps = int(max_steps) except : print("Error: Input 1 (max_steps) has to be an int.") return 0 try : assert self._max_steps >= 0 except : print("Warning: Input 1 (max_steps) has to be non-negative.") print("Setting max_steps to 0.") self._max_steps = 0 if max_steps > 0 : try : assert step_size == float(step_size) except AssertionError : print("Warning: Input 2 (step_size) is not a float. Converted.") step_size = float(step_size) except : print("Error: Input 2 (step_size) has to be a float.") return 0 try : assert 0. < step_size < 1. except : print("Warning: Input 2 (step_size) has to be between 0 and 1.") print("Setting step_size to 0.2.") step_size = 0.2 self._step_size = step_size if step_size**max_steps < 10**(-15): print("Warning: Back-off gets dangerously small.") # end checks def dynamic(self, max_steps, opts={}): """ Dynamic Switch Set the sampler parameters for dynamic back off Inputs : max_steps : maximum back-off steps to be taken Optional Inputs: opts : ({}) dictionary containing fancy options """ self._dynamic = True # begin checks try : self._max_steps = int(max_steps) except : raise TypeError("input 1 (max_steps) has to be an integer") return 0 try : assert self._max_steps >= 0 except : raise Warning("input 1 (max_steps) has to be non-negative. Setting (max_steps) to 0.") self._max_steps = 0 self._opts = opts # end checks def sample(self, n_samples, divs=1, visual=False, safe=False): """ Sample Sampling Inputs : n_samples : number of samples to generate Optional Inputs : divs : (1) number of divisions visual : show progress safe : save the chain at every division """ if visual: print("Sampling: 0%") for i in xrange(divs): self._sample(int(n_samples/divs)) if visual: sys.stdout.write("\033[F") # curser up print("Sampling: "+str(int(i*100./divs)+1)+'%') if safe: self.save(path="chain_{:}.dat".format(i)) if n_samples % divs != 0: self._sample(n_samples % divs) if safe: self.save(path="chain_{:}.dat".format(divs)) def save(self, path="chain.dat"): """ Save Save data to file Inputs : path : specifies the path name of the file to be loaded to """ # create dictionary for data dic = {} dic['chain'] = self._chain.tolist() dic['step_count'] = self._step_count.tolist() dic['n_samples'] = self._n_samples dic['n_accepted'] = self._n_accepted dic['x'] = self._X['x'].tolist() dic['f'] = self._X['f'].tolist() dic['J'] = self._X['J'].tolist() # write data to file file = open(path, 'w') json.dump(dic, file) file.close() def load(self, path="chain.dat"): """ Load Load data from file Inputs : path : specifies the path name of the file to be loaded from """ # read data from file file = open(path, 'r') dic = json.load(file) file.close() # get data from dictionary self._chain = np.array(dic['chain']) self._step_count = np.array(dic['step_count']) self._n_samples = dic['n_samples'] self._n_accepted = dic['n_accepted'] self._X = {} self._X['x'] = dic['x'] self._X['f'] = dic['f'] self._X['J'] = dic['J'] def burn(self, n_burned): """ Burn Burn the inital samples to adjust for convergence of the chain cut the first (n_burned) burn-in samples Inputs : chain : the full Markov chain n_burned : number of samples to cut Hidden Outputs : chain : chain with the firt n_burned samples cut """ self._chain = self._chain[n_burned:] def acor(self, k = 5): """ Autocorrelation time of the chain return the autocorrelation time for each parameters Inputs : k : parameter in self-consistent window Outputs : t : autocorrelation time of the chain """ try: import acor except ImportError: print("Can't import acor, please download it.") return 0 n = np.shape(self._chain)[1] t = np.zeros(n) for i in xrange(n): t[i] = acor.acor(self._chain[:,i],k)[0] return t def posterior(self, x): """ Posterior density ** not normalized ** This is used to plot the theoretical curve for tests. Inputs : x : input value Outputs : p : p(x)=pi(x)*exp{-||f(x)||^2/(2)} posterior probability of x """ x = np.reshape(np.array(x), (-1,1)) chi_x, f_x, J_x = self._f(x) if chi_x : p = np.exp(-la.norm(f_x)**2/2.) if self._prior: m = self._m H = self._H p = p * np.exp(-np.dot((x-m).T,np.dot(H,x-m))/2.) return p else : return 0 def error_bars(self, n_bins, d_min, d_max): """ Error Bars create bars and error bars to plot Inputs : n_bins : number of bins plot_range : (shape) = (number of dimensions, 2) matrix which contain the min and max for each dimension as rows Outputs : x : domain p_x : estimated posterior using the chain on the domain error : estimated error for p_x """ # fetch data chain = self._chain len_chain = len(chain) try: n_dims = np.shape(chain)[1] except: n_dims = 1 # begin checks try: assert n_bins == int(n_bins) except: raise TypeError("number of bins has to be an integer") d_min = np.reshape(np.array(d_min), (-1,1)) d_max = np.reshape(np.array(d_max), (-1,1)) try: assert np.size(d_min) == n_dims except: raise TypeError("domain minimum has wrong size") try: assert np.size(d_max) == n_dims except: raise TypeError("domain maximum has wrong size") # end checks # initialize outputs p_x = np.zeros(n_bins) # esitmate of posterior error = np.zeros(n_bins) # error bars x = np.zeros((n_dims, n_bins)) # centers of bins # set dx v = d_max-d_min v_2 = np.dot(v.T, v)[0][0] # bin count for i in xrange(len_chain): bin_no = int(np.floor(np.dot(chain[i].T-d_min,v)/v_2*n_bins)[0]) if n_bins > bin_no > -1: p_x[bin_no] += 1. # end count dx = np.sqrt(v_2)/n_bins p_x = p_x/(len_chain*dx) # find error for i in xrange(n_bins): p = p_x[i] error[i] = np.sqrt(p*(1./dx-p)/(len_chain)) x[:,i] = (d_min+v*(0.5+i)/n_bins)[0] # end find return x, p_x, error # end error_bars # internal methods def _sample(self, n_samples): """ Sample Generate samples for posterior distribution using Gauss-Newton proposal parameters Inputs : n_samples : number of samples to generate Hidden Outputs : chain : chain of samples n_samples : length of chain n_accepted : number of proposals accepted step_count : count of the steps accepted """ try : n_samples = int(n_samples) except : raise TypeError("number of samples has to be an integer") exit() # fetch info X = self._proposal_params(self._X) k_max = self._max_steps # initialize chain = np.zeros((n_samples, self._n)) n_accepted = 0 step_count = np.zeros(k_max+2) # begin outer loop for i in xrange(n_samples): accepted = False # check if sample is accepted r_ = [1] # list of step sizes Z_ = [X] # initialize list of Z s self._r_ = r_ log_P_z_x = 0. + X['log_p'] k = 0 # back-off steps taken so far while k <= k_max: # get proposal chi_z = False while not chi_z: z = multi_normal(X, r_[-1]) chi_z, f_z, J_z = self._f(z) Z = self._proposal_params({'x':z,'f':f_z,'J':J_z}) Z_.append(Z) self._Z_ = Z_ log_P_z_x += log_K(Z, X, r_[-1]) # N is the Numerator of the acceptance, N = P_x_z self._N_is_0 = False # check to see if N = 0, to use in _log_P log_N = self._log_P(X, Z, k) # calculating acceptance probability if self._N_is_0 == True : A_z_x = 0. elif log_N >= log_P_z_x : A_z_x = 1. else : A_z_x = np.exp(log_N - log_P_z_x) # acceptance rejection if np.random.rand() <= A_z_x: accepted = True break else : log_P_z_x += np.log(1. - A_z_x) self._back_off() k += 1 # end of steps for loop if accepted == True : chain[i,:] = z[:,0] X = Z # for statistics n_accepted += 1 step_count[k+1] += 1 else : chain[i,:] = X['x'][:,0] # for statistics step_count[0] += 1 # end outer loop # update stored info self._X = X # outputs if self._n_samples == 0 : self._chain = chain self._step_count = step_count else : self._chain = np.append(self._chain, chain, axis=0) self._step_count = np.add(self._step_count, step_count) self._n_samples += n_samples self._n_accepted += n_accepted # end sample def _proposal_params(self, state): """ Proposal parameters Calculate parameters needed for the proposal. Inputs : state : x : the present sample, the place to linearize around f : f(x), function value at x J : f'(x), the jacobian of the function evaluated at x Outputs : state : mu : the mean vector L : the lower triangular cholesky factor of P log_p : log(p(x)) log of the posterior density """ x = state['x'] f = state['f'] J = state['J'] JJ = np.dot(J.T,J) if self._prior: m = self._m H = self._H Hm = self._Hm # LL' = P = H+J'J L = la.cholesky(H+JJ) # mu = (P^-1)(Hm-J'f+J'Jx) mu = la.solve(L.T,la.solve(L,Hm-np.dot(J.T,f)+np.dot(JJ,x))) else: # P = J'J L = la.cholesky(JJ) # mu = x-(P^-1)J'f mu = x-la.solve(L.T,la.solve(L,np.dot(J.T,f))) state['L'] = L state['mu'] = mu state['log_p'] = self._log_post(x,f) return state def _back_off(self): """ Back off Calculate the back off step size Inputs : Z_ : list of states in current proposal r_ : list of back offs in current proposal q : step size reduction dynamic : set to True if you want to use the dynamic back-off Outputs : """ q = self._step_size r = self._r_[-1] Z_ = self._Z_ if self._dynamic: p_0 = la.norm(Z_[0]['f']) dp_0 = p_0*2*la.norm(Z_[0]['J']) p_r = la.norm(Z_[-1]['f']) dp_r = p_0*2*la.norm(Z_[-1]['J']) r_new = optimize(r, p_0**2, dp_0, p_r**2, dp_r) else : r_new = r * q self._r_.append(r_new) def _log_post(self,x,f_x): """ Log of the posterior density This is used to calculete acceptance probability for sampling. Inputs : x : input value f_x : f(x), function value at x Outputs : log(p_x) : log[pi(x)]-||f(x)||^2/(2) log of the posterior probability """ # least squares part -||f(x)||^2/2 log_likelihood = (-la.norm(f_x)**2)/(2.) # prior part -(x-m)'H(x-m)/2 if self._prior: m = self._m H = self._H log_prior = self._ln_H_-np.dot((x-m).T,np.dot(H,x-m))/2. return log_prior+log_likelihood else: return log_likelihood """ Properties: 1. chain 2. n_samples 3. n_accepted 4. accept_rate 5. step_count 6. call_count 7. max_steps 8. step_size """ @property def chain(self): return self._chain @property def n_samples(self): return self._n_samples @property def n_accepted(self): return self._n_accepted @property def accept_rate(self): return float(self._n_accepted)/self._n_samples @property def step_count(self): return self._step_count @property def call_count(self): return self._f.count @property def max_steps(self): return self._max_steps @property def step_size(self): return self._step_size
mugurbil/gnm
gnm/gnm.py
sampler._back_off
python
def _back_off(self): q = self._step_size r = self._r_[-1] Z_ = self._Z_ if self._dynamic: p_0 = la.norm(Z_[0]['f']) dp_0 = p_0*2*la.norm(Z_[0]['J']) p_r = la.norm(Z_[-1]['f']) dp_r = p_0*2*la.norm(Z_[-1]['J']) r_new = optimize(r, p_0**2, dp_0, p_r**2, dp_r) else : r_new = r * q self._r_.append(r_new)
Back off Calculate the back off step size Inputs : Z_ : list of states in current proposal r_ : list of back offs in current proposal q : step size reduction dynamic : set to True if you want to use the dynamic back-off Outputs :
train
https://github.com/mugurbil/gnm/blob/4f9711fb9d78cc02820c25234bc3ab9615014f11/gnm/gnm.py#L607-L633
null
class sampler(object): def __init__(self, x, model, args): """ Init Initialize the GNM sampler class Inputs : x : initial guess model : user defined data model function args : arguments for the model """ self._args = args self._f = function(model, args) x = np.reshape(np.array(x), (-1,1)) self._n = np.size(x) # size of input space try: x = np.reshape(np.array(x), (-1)) out_x = model(x, self._args) except TypeError as e: raise TypeError(str(e)[:-7]+" needed)") """ except IndexError as e: raise IndexError("initial guess size does not fit model()\n " +str(e)) except Exception as e: print("Error: Model function could not be evaluated.") print(" - Check size of intial guess.") print(" - Check definition of the model.") print(str(e)) print(type(e)) raise RuntimeError("model() could not be evaluated\n ") """ try: chi_x, f_x, J_x = out_x f_x = np.reshape(np.array(f_x), (-1,1)) except: raise TypeError("model() needs to have 3 outputs: chi_x, f_x, J_x") try: assert chi_x == True except AssertionError: raise ValueError("initial guess out of range") try: self._mn = np.size(f_x) J_x = np.array(J_x) assert np.shape(J_x) == (self._mn, self._n) except: raise TypeError("Shape of Jacobian, " + str(np.shape(J_x)) + ", is not correct, (%d, %d)." % (self._mn, self._n)) x = np.reshape(np.array(x), (-1,1)) self._X = {'x':x,'f':f_x,'J':J_x} # state of x # prior parameters self._prior = False # back-off parameters self._max_steps = 1 self._step_size = 0.1 self._dynamic = False self._opts = {} # sampler outputs self._chain = None self._n_samples = 0 self._n_accepted = 0 def prior(self, m, H): """ Set prior Set prior values Inputs : m : mean of the prior H : precision matrix of the prior Hiddens : ln_H_ : log(det(H))/2 calculate this once to use everytime log prior is called Hm : < H, m > calculate this once to use everytime proposal is called """ if self._prior == True: raise Warning("prior information is already set") else: self._prior = True # mean self._m = np.reshape(np.array(m), (-1,1)) try : assert np.size(self._m) == self._n except : raise TypeError("mean has to be an array of size n") # precision self._H = np.array(H) try : assert np.shape(self._H) == (self._n, self._n) except : raise TypeError("precision has to be a matrix of shape n by n") # precalculations self._ln_H_ = np.log(la.det(self._H))/2. self._Hm = np.dot(self._H, self._m) def Jtest(self, x_min, x_max, dx=0.0002, N=1000, eps_max=0.0001, p=2, l_max=50, r=0.5): """ Gradient Checker Test the function's jacobian against the numerical jacobian """ # check inputs x_min and x_max try : assert np.size(x_min) == self._n except : raise TypeError("dimension of x_min, %d, does not match the " "dimension of input, %d" % (np.size(x_min), self._n)) try : assert np.size(x_max) == self._n except : raise TypeError("dimension of x_max, %d, does not match the " "dimension of input, %d." % (np.size(x_max), self._n)) # end checks and call developer function return self._f.Jtest(x_min, x_max, dx=dx, N=N, eps_max=eps_max, p=p, l_max=l_max, r=r) def static(self, max_steps, step_size): """ Set Back-off to Static Set the sampler parameters for static back off Inputs : max_steps : maximum optimization steps to be taken step_size : the step size of the back-off """ self._dynamic = False # begin checks try : self._max_steps = int(max_steps) except : print("Error: Input 1 (max_steps) has to be an int.") return 0 try : assert self._max_steps >= 0 except : print("Warning: Input 1 (max_steps) has to be non-negative.") print("Setting max_steps to 0.") self._max_steps = 0 if max_steps > 0 : try : assert step_size == float(step_size) except AssertionError : print("Warning: Input 2 (step_size) is not a float. Converted.") step_size = float(step_size) except : print("Error: Input 2 (step_size) has to be a float.") return 0 try : assert 0. < step_size < 1. except : print("Warning: Input 2 (step_size) has to be between 0 and 1.") print("Setting step_size to 0.2.") step_size = 0.2 self._step_size = step_size if step_size**max_steps < 10**(-15): print("Warning: Back-off gets dangerously small.") # end checks def dynamic(self, max_steps, opts={}): """ Dynamic Switch Set the sampler parameters for dynamic back off Inputs : max_steps : maximum back-off steps to be taken Optional Inputs: opts : ({}) dictionary containing fancy options """ self._dynamic = True # begin checks try : self._max_steps = int(max_steps) except : raise TypeError("input 1 (max_steps) has to be an integer") return 0 try : assert self._max_steps >= 0 except : raise Warning("input 1 (max_steps) has to be non-negative. Setting (max_steps) to 0.") self._max_steps = 0 self._opts = opts # end checks def sample(self, n_samples, divs=1, visual=False, safe=False): """ Sample Sampling Inputs : n_samples : number of samples to generate Optional Inputs : divs : (1) number of divisions visual : show progress safe : save the chain at every division """ if visual: print("Sampling: 0%") for i in xrange(divs): self._sample(int(n_samples/divs)) if visual: sys.stdout.write("\033[F") # curser up print("Sampling: "+str(int(i*100./divs)+1)+'%') if safe: self.save(path="chain_{:}.dat".format(i)) if n_samples % divs != 0: self._sample(n_samples % divs) if safe: self.save(path="chain_{:}.dat".format(divs)) def save(self, path="chain.dat"): """ Save Save data to file Inputs : path : specifies the path name of the file to be loaded to """ # create dictionary for data dic = {} dic['chain'] = self._chain.tolist() dic['step_count'] = self._step_count.tolist() dic['n_samples'] = self._n_samples dic['n_accepted'] = self._n_accepted dic['x'] = self._X['x'].tolist() dic['f'] = self._X['f'].tolist() dic['J'] = self._X['J'].tolist() # write data to file file = open(path, 'w') json.dump(dic, file) file.close() def load(self, path="chain.dat"): """ Load Load data from file Inputs : path : specifies the path name of the file to be loaded from """ # read data from file file = open(path, 'r') dic = json.load(file) file.close() # get data from dictionary self._chain = np.array(dic['chain']) self._step_count = np.array(dic['step_count']) self._n_samples = dic['n_samples'] self._n_accepted = dic['n_accepted'] self._X = {} self._X['x'] = dic['x'] self._X['f'] = dic['f'] self._X['J'] = dic['J'] def burn(self, n_burned): """ Burn Burn the inital samples to adjust for convergence of the chain cut the first (n_burned) burn-in samples Inputs : chain : the full Markov chain n_burned : number of samples to cut Hidden Outputs : chain : chain with the firt n_burned samples cut """ self._chain = self._chain[n_burned:] def acor(self, k = 5): """ Autocorrelation time of the chain return the autocorrelation time for each parameters Inputs : k : parameter in self-consistent window Outputs : t : autocorrelation time of the chain """ try: import acor except ImportError: print("Can't import acor, please download it.") return 0 n = np.shape(self._chain)[1] t = np.zeros(n) for i in xrange(n): t[i] = acor.acor(self._chain[:,i],k)[0] return t def posterior(self, x): """ Posterior density ** not normalized ** This is used to plot the theoretical curve for tests. Inputs : x : input value Outputs : p : p(x)=pi(x)*exp{-||f(x)||^2/(2)} posterior probability of x """ x = np.reshape(np.array(x), (-1,1)) chi_x, f_x, J_x = self._f(x) if chi_x : p = np.exp(-la.norm(f_x)**2/2.) if self._prior: m = self._m H = self._H p = p * np.exp(-np.dot((x-m).T,np.dot(H,x-m))/2.) return p else : return 0 def error_bars(self, n_bins, d_min, d_max): """ Error Bars create bars and error bars to plot Inputs : n_bins : number of bins plot_range : (shape) = (number of dimensions, 2) matrix which contain the min and max for each dimension as rows Outputs : x : domain p_x : estimated posterior using the chain on the domain error : estimated error for p_x """ # fetch data chain = self._chain len_chain = len(chain) try: n_dims = np.shape(chain)[1] except: n_dims = 1 # begin checks try: assert n_bins == int(n_bins) except: raise TypeError("number of bins has to be an integer") d_min = np.reshape(np.array(d_min), (-1,1)) d_max = np.reshape(np.array(d_max), (-1,1)) try: assert np.size(d_min) == n_dims except: raise TypeError("domain minimum has wrong size") try: assert np.size(d_max) == n_dims except: raise TypeError("domain maximum has wrong size") # end checks # initialize outputs p_x = np.zeros(n_bins) # esitmate of posterior error = np.zeros(n_bins) # error bars x = np.zeros((n_dims, n_bins)) # centers of bins # set dx v = d_max-d_min v_2 = np.dot(v.T, v)[0][0] # bin count for i in xrange(len_chain): bin_no = int(np.floor(np.dot(chain[i].T-d_min,v)/v_2*n_bins)[0]) if n_bins > bin_no > -1: p_x[bin_no] += 1. # end count dx = np.sqrt(v_2)/n_bins p_x = p_x/(len_chain*dx) # find error for i in xrange(n_bins): p = p_x[i] error[i] = np.sqrt(p*(1./dx-p)/(len_chain)) x[:,i] = (d_min+v*(0.5+i)/n_bins)[0] # end find return x, p_x, error # end error_bars # internal methods def _sample(self, n_samples): """ Sample Generate samples for posterior distribution using Gauss-Newton proposal parameters Inputs : n_samples : number of samples to generate Hidden Outputs : chain : chain of samples n_samples : length of chain n_accepted : number of proposals accepted step_count : count of the steps accepted """ try : n_samples = int(n_samples) except : raise TypeError("number of samples has to be an integer") exit() # fetch info X = self._proposal_params(self._X) k_max = self._max_steps # initialize chain = np.zeros((n_samples, self._n)) n_accepted = 0 step_count = np.zeros(k_max+2) # begin outer loop for i in xrange(n_samples): accepted = False # check if sample is accepted r_ = [1] # list of step sizes Z_ = [X] # initialize list of Z s self._r_ = r_ log_P_z_x = 0. + X['log_p'] k = 0 # back-off steps taken so far while k <= k_max: # get proposal chi_z = False while not chi_z: z = multi_normal(X, r_[-1]) chi_z, f_z, J_z = self._f(z) Z = self._proposal_params({'x':z,'f':f_z,'J':J_z}) Z_.append(Z) self._Z_ = Z_ log_P_z_x += log_K(Z, X, r_[-1]) # N is the Numerator of the acceptance, N = P_x_z self._N_is_0 = False # check to see if N = 0, to use in _log_P log_N = self._log_P(X, Z, k) # calculating acceptance probability if self._N_is_0 == True : A_z_x = 0. elif log_N >= log_P_z_x : A_z_x = 1. else : A_z_x = np.exp(log_N - log_P_z_x) # acceptance rejection if np.random.rand() <= A_z_x: accepted = True break else : log_P_z_x += np.log(1. - A_z_x) self._back_off() k += 1 # end of steps for loop if accepted == True : chain[i,:] = z[:,0] X = Z # for statistics n_accepted += 1 step_count[k+1] += 1 else : chain[i,:] = X['x'][:,0] # for statistics step_count[0] += 1 # end outer loop # update stored info self._X = X # outputs if self._n_samples == 0 : self._chain = chain self._step_count = step_count else : self._chain = np.append(self._chain, chain, axis=0) self._step_count = np.add(self._step_count, step_count) self._n_samples += n_samples self._n_accepted += n_accepted # end sample def _proposal_params(self, state): """ Proposal parameters Calculate parameters needed for the proposal. Inputs : state : x : the present sample, the place to linearize around f : f(x), function value at x J : f'(x), the jacobian of the function evaluated at x Outputs : state : mu : the mean vector L : the lower triangular cholesky factor of P log_p : log(p(x)) log of the posterior density """ x = state['x'] f = state['f'] J = state['J'] JJ = np.dot(J.T,J) if self._prior: m = self._m H = self._H Hm = self._Hm # LL' = P = H+J'J L = la.cholesky(H+JJ) # mu = (P^-1)(Hm-J'f+J'Jx) mu = la.solve(L.T,la.solve(L,Hm-np.dot(J.T,f)+np.dot(JJ,x))) else: # P = J'J L = la.cholesky(JJ) # mu = x-(P^-1)J'f mu = x-la.solve(L.T,la.solve(L,np.dot(J.T,f))) state['L'] = L state['mu'] = mu state['log_p'] = self._log_post(x,f) return state def _log_P(self, X , Z, k): """ Log of the probability of transition from z to x with k steps log ( P_k (x, z) ) Inputs : X : state to be proposed to Z : state to be proposed from k : number of recursions, depth """ r_ = self._r_ Z_ = self._Z_ # zero case if k == 0 : log_P = Z['log_p'] + log_K(X, Z, r_[k]) # recursice case else : P_zk_z = np.exp( self._log_P(Z_[k], Z, k-1) ) P_z_zk = np.exp( self._log_P(Z, Z_[k], k-1) ) # flag if P_zk_z <= P_z_zk : self._N_is_0 = True log_P = -np.inf else : log_P = np.log( P_zk_z - P_z_zk ) + log_K(X, Z, r_[k]) return log_P def _log_post(self,x,f_x): """ Log of the posterior density This is used to calculete acceptance probability for sampling. Inputs : x : input value f_x : f(x), function value at x Outputs : log(p_x) : log[pi(x)]-||f(x)||^2/(2) log of the posterior probability """ # least squares part -||f(x)||^2/2 log_likelihood = (-la.norm(f_x)**2)/(2.) # prior part -(x-m)'H(x-m)/2 if self._prior: m = self._m H = self._H log_prior = self._ln_H_-np.dot((x-m).T,np.dot(H,x-m))/2. return log_prior+log_likelihood else: return log_likelihood """ Properties: 1. chain 2. n_samples 3. n_accepted 4. accept_rate 5. step_count 6. call_count 7. max_steps 8. step_size """ @property def chain(self): return self._chain @property def n_samples(self): return self._n_samples @property def n_accepted(self): return self._n_accepted @property def accept_rate(self): return float(self._n_accepted)/self._n_samples @property def step_count(self): return self._step_count @property def call_count(self): return self._f.count @property def max_steps(self): return self._max_steps @property def step_size(self): return self._step_size
mugurbil/gnm
gnm/gnm.py
sampler._log_post
python
def _log_post(self,x,f_x): # least squares part -||f(x)||^2/2 log_likelihood = (-la.norm(f_x)**2)/(2.) # prior part -(x-m)'H(x-m)/2 if self._prior: m = self._m H = self._H log_prior = self._ln_H_-np.dot((x-m).T,np.dot(H,x-m))/2. return log_prior+log_likelihood else: return log_likelihood
Log of the posterior density This is used to calculete acceptance probability for sampling. Inputs : x : input value f_x : f(x), function value at x Outputs : log(p_x) : log[pi(x)]-||f(x)||^2/(2) log of the posterior probability
train
https://github.com/mugurbil/gnm/blob/4f9711fb9d78cc02820c25234bc3ab9615014f11/gnm/gnm.py#L635-L658
null
class sampler(object): def __init__(self, x, model, args): """ Init Initialize the GNM sampler class Inputs : x : initial guess model : user defined data model function args : arguments for the model """ self._args = args self._f = function(model, args) x = np.reshape(np.array(x), (-1,1)) self._n = np.size(x) # size of input space try: x = np.reshape(np.array(x), (-1)) out_x = model(x, self._args) except TypeError as e: raise TypeError(str(e)[:-7]+" needed)") """ except IndexError as e: raise IndexError("initial guess size does not fit model()\n " +str(e)) except Exception as e: print("Error: Model function could not be evaluated.") print(" - Check size of intial guess.") print(" - Check definition of the model.") print(str(e)) print(type(e)) raise RuntimeError("model() could not be evaluated\n ") """ try: chi_x, f_x, J_x = out_x f_x = np.reshape(np.array(f_x), (-1,1)) except: raise TypeError("model() needs to have 3 outputs: chi_x, f_x, J_x") try: assert chi_x == True except AssertionError: raise ValueError("initial guess out of range") try: self._mn = np.size(f_x) J_x = np.array(J_x) assert np.shape(J_x) == (self._mn, self._n) except: raise TypeError("Shape of Jacobian, " + str(np.shape(J_x)) + ", is not correct, (%d, %d)." % (self._mn, self._n)) x = np.reshape(np.array(x), (-1,1)) self._X = {'x':x,'f':f_x,'J':J_x} # state of x # prior parameters self._prior = False # back-off parameters self._max_steps = 1 self._step_size = 0.1 self._dynamic = False self._opts = {} # sampler outputs self._chain = None self._n_samples = 0 self._n_accepted = 0 def prior(self, m, H): """ Set prior Set prior values Inputs : m : mean of the prior H : precision matrix of the prior Hiddens : ln_H_ : log(det(H))/2 calculate this once to use everytime log prior is called Hm : < H, m > calculate this once to use everytime proposal is called """ if self._prior == True: raise Warning("prior information is already set") else: self._prior = True # mean self._m = np.reshape(np.array(m), (-1,1)) try : assert np.size(self._m) == self._n except : raise TypeError("mean has to be an array of size n") # precision self._H = np.array(H) try : assert np.shape(self._H) == (self._n, self._n) except : raise TypeError("precision has to be a matrix of shape n by n") # precalculations self._ln_H_ = np.log(la.det(self._H))/2. self._Hm = np.dot(self._H, self._m) def Jtest(self, x_min, x_max, dx=0.0002, N=1000, eps_max=0.0001, p=2, l_max=50, r=0.5): """ Gradient Checker Test the function's jacobian against the numerical jacobian """ # check inputs x_min and x_max try : assert np.size(x_min) == self._n except : raise TypeError("dimension of x_min, %d, does not match the " "dimension of input, %d" % (np.size(x_min), self._n)) try : assert np.size(x_max) == self._n except : raise TypeError("dimension of x_max, %d, does not match the " "dimension of input, %d." % (np.size(x_max), self._n)) # end checks and call developer function return self._f.Jtest(x_min, x_max, dx=dx, N=N, eps_max=eps_max, p=p, l_max=l_max, r=r) def static(self, max_steps, step_size): """ Set Back-off to Static Set the sampler parameters for static back off Inputs : max_steps : maximum optimization steps to be taken step_size : the step size of the back-off """ self._dynamic = False # begin checks try : self._max_steps = int(max_steps) except : print("Error: Input 1 (max_steps) has to be an int.") return 0 try : assert self._max_steps >= 0 except : print("Warning: Input 1 (max_steps) has to be non-negative.") print("Setting max_steps to 0.") self._max_steps = 0 if max_steps > 0 : try : assert step_size == float(step_size) except AssertionError : print("Warning: Input 2 (step_size) is not a float. Converted.") step_size = float(step_size) except : print("Error: Input 2 (step_size) has to be a float.") return 0 try : assert 0. < step_size < 1. except : print("Warning: Input 2 (step_size) has to be between 0 and 1.") print("Setting step_size to 0.2.") step_size = 0.2 self._step_size = step_size if step_size**max_steps < 10**(-15): print("Warning: Back-off gets dangerously small.") # end checks def dynamic(self, max_steps, opts={}): """ Dynamic Switch Set the sampler parameters for dynamic back off Inputs : max_steps : maximum back-off steps to be taken Optional Inputs: opts : ({}) dictionary containing fancy options """ self._dynamic = True # begin checks try : self._max_steps = int(max_steps) except : raise TypeError("input 1 (max_steps) has to be an integer") return 0 try : assert self._max_steps >= 0 except : raise Warning("input 1 (max_steps) has to be non-negative. Setting (max_steps) to 0.") self._max_steps = 0 self._opts = opts # end checks def sample(self, n_samples, divs=1, visual=False, safe=False): """ Sample Sampling Inputs : n_samples : number of samples to generate Optional Inputs : divs : (1) number of divisions visual : show progress safe : save the chain at every division """ if visual: print("Sampling: 0%") for i in xrange(divs): self._sample(int(n_samples/divs)) if visual: sys.stdout.write("\033[F") # curser up print("Sampling: "+str(int(i*100./divs)+1)+'%') if safe: self.save(path="chain_{:}.dat".format(i)) if n_samples % divs != 0: self._sample(n_samples % divs) if safe: self.save(path="chain_{:}.dat".format(divs)) def save(self, path="chain.dat"): """ Save Save data to file Inputs : path : specifies the path name of the file to be loaded to """ # create dictionary for data dic = {} dic['chain'] = self._chain.tolist() dic['step_count'] = self._step_count.tolist() dic['n_samples'] = self._n_samples dic['n_accepted'] = self._n_accepted dic['x'] = self._X['x'].tolist() dic['f'] = self._X['f'].tolist() dic['J'] = self._X['J'].tolist() # write data to file file = open(path, 'w') json.dump(dic, file) file.close() def load(self, path="chain.dat"): """ Load Load data from file Inputs : path : specifies the path name of the file to be loaded from """ # read data from file file = open(path, 'r') dic = json.load(file) file.close() # get data from dictionary self._chain = np.array(dic['chain']) self._step_count = np.array(dic['step_count']) self._n_samples = dic['n_samples'] self._n_accepted = dic['n_accepted'] self._X = {} self._X['x'] = dic['x'] self._X['f'] = dic['f'] self._X['J'] = dic['J'] def burn(self, n_burned): """ Burn Burn the inital samples to adjust for convergence of the chain cut the first (n_burned) burn-in samples Inputs : chain : the full Markov chain n_burned : number of samples to cut Hidden Outputs : chain : chain with the firt n_burned samples cut """ self._chain = self._chain[n_burned:] def acor(self, k = 5): """ Autocorrelation time of the chain return the autocorrelation time for each parameters Inputs : k : parameter in self-consistent window Outputs : t : autocorrelation time of the chain """ try: import acor except ImportError: print("Can't import acor, please download it.") return 0 n = np.shape(self._chain)[1] t = np.zeros(n) for i in xrange(n): t[i] = acor.acor(self._chain[:,i],k)[0] return t def posterior(self, x): """ Posterior density ** not normalized ** This is used to plot the theoretical curve for tests. Inputs : x : input value Outputs : p : p(x)=pi(x)*exp{-||f(x)||^2/(2)} posterior probability of x """ x = np.reshape(np.array(x), (-1,1)) chi_x, f_x, J_x = self._f(x) if chi_x : p = np.exp(-la.norm(f_x)**2/2.) if self._prior: m = self._m H = self._H p = p * np.exp(-np.dot((x-m).T,np.dot(H,x-m))/2.) return p else : return 0 def error_bars(self, n_bins, d_min, d_max): """ Error Bars create bars and error bars to plot Inputs : n_bins : number of bins plot_range : (shape) = (number of dimensions, 2) matrix which contain the min and max for each dimension as rows Outputs : x : domain p_x : estimated posterior using the chain on the domain error : estimated error for p_x """ # fetch data chain = self._chain len_chain = len(chain) try: n_dims = np.shape(chain)[1] except: n_dims = 1 # begin checks try: assert n_bins == int(n_bins) except: raise TypeError("number of bins has to be an integer") d_min = np.reshape(np.array(d_min), (-1,1)) d_max = np.reshape(np.array(d_max), (-1,1)) try: assert np.size(d_min) == n_dims except: raise TypeError("domain minimum has wrong size") try: assert np.size(d_max) == n_dims except: raise TypeError("domain maximum has wrong size") # end checks # initialize outputs p_x = np.zeros(n_bins) # esitmate of posterior error = np.zeros(n_bins) # error bars x = np.zeros((n_dims, n_bins)) # centers of bins # set dx v = d_max-d_min v_2 = np.dot(v.T, v)[0][0] # bin count for i in xrange(len_chain): bin_no = int(np.floor(np.dot(chain[i].T-d_min,v)/v_2*n_bins)[0]) if n_bins > bin_no > -1: p_x[bin_no] += 1. # end count dx = np.sqrt(v_2)/n_bins p_x = p_x/(len_chain*dx) # find error for i in xrange(n_bins): p = p_x[i] error[i] = np.sqrt(p*(1./dx-p)/(len_chain)) x[:,i] = (d_min+v*(0.5+i)/n_bins)[0] # end find return x, p_x, error # end error_bars # internal methods def _sample(self, n_samples): """ Sample Generate samples for posterior distribution using Gauss-Newton proposal parameters Inputs : n_samples : number of samples to generate Hidden Outputs : chain : chain of samples n_samples : length of chain n_accepted : number of proposals accepted step_count : count of the steps accepted """ try : n_samples = int(n_samples) except : raise TypeError("number of samples has to be an integer") exit() # fetch info X = self._proposal_params(self._X) k_max = self._max_steps # initialize chain = np.zeros((n_samples, self._n)) n_accepted = 0 step_count = np.zeros(k_max+2) # begin outer loop for i in xrange(n_samples): accepted = False # check if sample is accepted r_ = [1] # list of step sizes Z_ = [X] # initialize list of Z s self._r_ = r_ log_P_z_x = 0. + X['log_p'] k = 0 # back-off steps taken so far while k <= k_max: # get proposal chi_z = False while not chi_z: z = multi_normal(X, r_[-1]) chi_z, f_z, J_z = self._f(z) Z = self._proposal_params({'x':z,'f':f_z,'J':J_z}) Z_.append(Z) self._Z_ = Z_ log_P_z_x += log_K(Z, X, r_[-1]) # N is the Numerator of the acceptance, N = P_x_z self._N_is_0 = False # check to see if N = 0, to use in _log_P log_N = self._log_P(X, Z, k) # calculating acceptance probability if self._N_is_0 == True : A_z_x = 0. elif log_N >= log_P_z_x : A_z_x = 1. else : A_z_x = np.exp(log_N - log_P_z_x) # acceptance rejection if np.random.rand() <= A_z_x: accepted = True break else : log_P_z_x += np.log(1. - A_z_x) self._back_off() k += 1 # end of steps for loop if accepted == True : chain[i,:] = z[:,0] X = Z # for statistics n_accepted += 1 step_count[k+1] += 1 else : chain[i,:] = X['x'][:,0] # for statistics step_count[0] += 1 # end outer loop # update stored info self._X = X # outputs if self._n_samples == 0 : self._chain = chain self._step_count = step_count else : self._chain = np.append(self._chain, chain, axis=0) self._step_count = np.add(self._step_count, step_count) self._n_samples += n_samples self._n_accepted += n_accepted # end sample def _proposal_params(self, state): """ Proposal parameters Calculate parameters needed for the proposal. Inputs : state : x : the present sample, the place to linearize around f : f(x), function value at x J : f'(x), the jacobian of the function evaluated at x Outputs : state : mu : the mean vector L : the lower triangular cholesky factor of P log_p : log(p(x)) log of the posterior density """ x = state['x'] f = state['f'] J = state['J'] JJ = np.dot(J.T,J) if self._prior: m = self._m H = self._H Hm = self._Hm # LL' = P = H+J'J L = la.cholesky(H+JJ) # mu = (P^-1)(Hm-J'f+J'Jx) mu = la.solve(L.T,la.solve(L,Hm-np.dot(J.T,f)+np.dot(JJ,x))) else: # P = J'J L = la.cholesky(JJ) # mu = x-(P^-1)J'f mu = x-la.solve(L.T,la.solve(L,np.dot(J.T,f))) state['L'] = L state['mu'] = mu state['log_p'] = self._log_post(x,f) return state def _log_P(self, X , Z, k): """ Log of the probability of transition from z to x with k steps log ( P_k (x, z) ) Inputs : X : state to be proposed to Z : state to be proposed from k : number of recursions, depth """ r_ = self._r_ Z_ = self._Z_ # zero case if k == 0 : log_P = Z['log_p'] + log_K(X, Z, r_[k]) # recursice case else : P_zk_z = np.exp( self._log_P(Z_[k], Z, k-1) ) P_z_zk = np.exp( self._log_P(Z, Z_[k], k-1) ) # flag if P_zk_z <= P_z_zk : self._N_is_0 = True log_P = -np.inf else : log_P = np.log( P_zk_z - P_z_zk ) + log_K(X, Z, r_[k]) return log_P def _back_off(self): """ Back off Calculate the back off step size Inputs : Z_ : list of states in current proposal r_ : list of back offs in current proposal q : step size reduction dynamic : set to True if you want to use the dynamic back-off Outputs : """ q = self._step_size r = self._r_[-1] Z_ = self._Z_ if self._dynamic: p_0 = la.norm(Z_[0]['f']) dp_0 = p_0*2*la.norm(Z_[0]['J']) p_r = la.norm(Z_[-1]['f']) dp_r = p_0*2*la.norm(Z_[-1]['J']) r_new = optimize(r, p_0**2, dp_0, p_r**2, dp_r) else : r_new = r * q self._r_.append(r_new) """ Properties: 1. chain 2. n_samples 3. n_accepted 4. accept_rate 5. step_count 6. call_count 7. max_steps 8. step_size """ @property def chain(self): return self._chain @property def n_samples(self): return self._n_samples @property def n_accepted(self): return self._n_accepted @property def accept_rate(self): return float(self._n_accepted)/self._n_samples @property def step_count(self): return self._step_count @property def call_count(self): return self._f.count @property def max_steps(self): return self._max_steps @property def step_size(self): return self._step_size
akissa/sachannelupdate
sachannelupdate/base.py
getfiles
python
def getfiles(qfiles, dirname, names): for name in names: fullname = os.path.join(dirname, name) if os.path.isfile(fullname) and \ fullname.endswith('.cf') or \ fullname.endswith('.post'): qfiles.put(fullname)
Get rule files in a directory
train
https://github.com/akissa/sachannelupdate/blob/a1c3c3d86b874f9c92c2407e2608963165d3ae98/sachannelupdate/base.py#L49-L56
null
# -*- coding: utf-8 -*- # vim: ai ts=4 sts=4 et sw=4 # sachannelupdate - Utility for pushing updates to Spamassassin update channels # Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Affero General Public License as published # by the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Affero General Public License for more details. # # You should have received a copy of the GNU Affero General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. """ sachannelupdate: Utility for pushing updates to Spamassassin update channels Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> """ import os import tarfile import datetime from Queue import Queue from hashlib import sha1 from datetime import datetime from gnupg import GPG from dns.exception import DNSException from dns import tsig, query, tsigkeyring, update from sachannelupdate.utils import info from sachannelupdate.exceptions import SaChannelUpdateConfigError \ as CfgError, SaChannelUpdateDNSError, SaChannelUpdateError from sachannelupdate.transports import get_sftp_conn, get_remote_path BLOCKSIZE = 65536 HASHTMPL = """%s %s\n""" def create_file(name, content): "Generic to write file" with open(name, 'w') as writefile: writefile.write(content) def deploy_file(source, dest): """Deploy a file""" date = datetime.utcnow().strftime('%Y-%m-%d') shandle = open(source) with open(dest, 'w') as handle: for line in shandle: if line == '# Updated: %date%\n': newline = '# Updated: %s\n' % date else: newline = line handle.write(newline) handle.flush() shandle.close() def package(dest, tardir, p_version): """Package files""" os.chdir(dest) p_filename = '%s.tar.gz' % p_version p_path = os.path.join(tardir, p_filename) tar = tarfile.open(p_path, mode='w:gz') for cf_file in os.listdir('.'): if os.path.isfile(cf_file): tar.add(cf_file) tar.close() def process(dest, rulefiles): """process rules""" deploy = False while not rulefiles.empty(): rulefile = rulefiles.get() base = os.path.basename(rulefile) dest = os.path.join(dest, base) if os.path.exists(dest): # check if older oldtime = os.stat(rulefile).st_mtime newtime = os.stat(dest).st_mtime if oldtime > newtime: deploy = True deploy_file(rulefile, dest) else: deploy = True deploy_file(rulefile, dest) return deploy def get_counter(counterfile): """Get the counter value""" try: version_num = open(counterfile).read() version_num = int(version_num) + 1 except (ValueError, IOError): version_num = 1 create_file(counterfile, "%d" % version_num) except BaseException as msg: raise SaChannelUpdateError(msg) return version_num def update_dns(config, record, sa_version): "Update the DNS record" try: domain = config.get('domain_name', 'sa.baruwa.com.') dns_key = config.get('domain_key') dns_ip = config.get('domain_ip', '127.0.0.1') keyring = tsigkeyring.from_text({domain: dns_key}) transaction = update.Update( domain, keyring=keyring, keyalgorithm=tsig.HMAC_SHA512) txtrecord = '%s.%s' % (sa_version, domain) transaction.replace(txtrecord, 120, 'txt', record) query.tcp(transaction, dns_ip) return True except DNSException, msg: raise SaChannelUpdateDNSError(msg) def sign(config, s_filename): """sign the package""" gpg_home = config.get('gpg_dir', '/var/lib/sachannelupdate/gnupg') gpg_pass = config.get('gpg_passphrase') gpg_keyid = config.get('gpg_keyid') gpg = GPG(gnupghome=gpg_home) try: plaintext = open(s_filename, 'rb') signature = gpg.sign_file( plaintext, keyid=gpg_keyid, passphrase=gpg_pass, detach=True) with open('%s.asc' % s_filename, 'wb') as handle: handle.write(str(signature)) finally: if 'plaintext' in locals(): plaintext.close() def hash_file(tar_filename): """hash the file""" hasher = sha1() with open(tar_filename, 'rb') as afile: buf = afile.read(BLOCKSIZE) while len(buf) > 0: hasher.update(buf) buf = afile.read(BLOCKSIZE) data = HASHTMPL % (hasher.hexdigest(), os.path.basename(tar_filename)) create_file('%s.sha1' % tar_filename, data) def upload(config, remote_loc, u_filename): """Upload the files""" rcode = False try: sftp, transport = get_sftp_conn(config) remote_dir = get_remote_path(remote_loc) for part in ['sha1', 'asc']: local_file = '%s.%s' % (u_filename, part) remote_file = os.path.join(remote_dir, local_file) sftp.put(local_file, remote_file) sftp.put(remote_dir, os.path.join(remote_dir, u_filename)) rcode = True except BaseException: pass finally: if 'transport' in locals(): transport.close() return rcode def queue_files(dirpath, queue): """Add files in a directory to a queue""" for root, _, files in os.walk(os.path.abspath(dirpath)): if not files: continue for filename in files: queue.put(os.path.join(root, filename)) def get_cf_files(path, queue): """Get rule files in a directory and put them in a queue""" for root, _, files in os.walk(os.path.abspath(path)): if not files: continue for filename in files: fullname = os.path.join(root, filename) if os.path.isfile(fullname) and fullname.endswith('.cf') or \ fullname.endswith('.post'): queue.put(fullname) def cleanup(dest, tardir, counterfile): """Remove existing rules""" thefiles = Queue() # dest directory files queue_files(dest, thefiles) # tar directory files queue_files(tardir, thefiles) while not thefiles.empty(): d_file = thefiles.get() info("Deleting file: %s" % d_file) os.unlink(d_file) if os.path.exists(counterfile): info("Deleting the counter file %s" % counterfile) os.unlink(counterfile) def check_required(config): """Validate the input""" if config.get('domain_key') is None: raise CfgError("The domain_key option is required") if config.get('remote_loc') is None: raise CfgError("The remote_location option is required") if config.get('gpg_keyid') is None: raise CfgError("The gpg_keyid option is required") def entry(config, delete_files=None): """Main function""" home_dir = config.get('home_dir', '/var/lib/sachannelupdate') dns_ver = config.get('spamassassin_version', '1.4.3') remote_loc = config.get('remote_location') rule_dir = os.path.join(home_dir, 'rules') dest = os.path.join(home_dir, 'deploy') tardir = os.path.join(home_dir, 'archives') counterfile = os.path.join(home_dir, 'db', 'counters') check_required(config) if delete_files: cleanup(dest, tardir, counterfile) return cffiles = Queue() get_cf_files(rule_dir, cffiles) if process(dest, cffiles): version = get_counter(counterfile) filename = '%s.tar.gz' % version path = os.path.join(tardir, filename) package(dest, tardir, version) sign(config, path) hash_file(path) if upload(config, remote_loc, path): if update_dns(config, str(version), dns_ver): create_file(counterfile, "%d" % version)
akissa/sachannelupdate
sachannelupdate/base.py
deploy_file
python
def deploy_file(source, dest): date = datetime.utcnow().strftime('%Y-%m-%d') shandle = open(source) with open(dest, 'w') as handle: for line in shandle: if line == '# Updated: %date%\n': newline = '# Updated: %s\n' % date else: newline = line handle.write(newline) handle.flush() shandle.close()
Deploy a file
train
https://github.com/akissa/sachannelupdate/blob/a1c3c3d86b874f9c92c2407e2608963165d3ae98/sachannelupdate/base.py#L59-L71
null
# -*- coding: utf-8 -*- # vim: ai ts=4 sts=4 et sw=4 # sachannelupdate - Utility for pushing updates to Spamassassin update channels # Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Affero General Public License as published # by the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Affero General Public License for more details. # # You should have received a copy of the GNU Affero General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. """ sachannelupdate: Utility for pushing updates to Spamassassin update channels Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> """ import os import tarfile import datetime from Queue import Queue from hashlib import sha1 from datetime import datetime from gnupg import GPG from dns.exception import DNSException from dns import tsig, query, tsigkeyring, update from sachannelupdate.utils import info from sachannelupdate.exceptions import SaChannelUpdateConfigError \ as CfgError, SaChannelUpdateDNSError, SaChannelUpdateError from sachannelupdate.transports import get_sftp_conn, get_remote_path BLOCKSIZE = 65536 HASHTMPL = """%s %s\n""" def create_file(name, content): "Generic to write file" with open(name, 'w') as writefile: writefile.write(content) def getfiles(qfiles, dirname, names): """Get rule files in a directory""" for name in names: fullname = os.path.join(dirname, name) if os.path.isfile(fullname) and \ fullname.endswith('.cf') or \ fullname.endswith('.post'): qfiles.put(fullname) def package(dest, tardir, p_version): """Package files""" os.chdir(dest) p_filename = '%s.tar.gz' % p_version p_path = os.path.join(tardir, p_filename) tar = tarfile.open(p_path, mode='w:gz') for cf_file in os.listdir('.'): if os.path.isfile(cf_file): tar.add(cf_file) tar.close() def process(dest, rulefiles): """process rules""" deploy = False while not rulefiles.empty(): rulefile = rulefiles.get() base = os.path.basename(rulefile) dest = os.path.join(dest, base) if os.path.exists(dest): # check if older oldtime = os.stat(rulefile).st_mtime newtime = os.stat(dest).st_mtime if oldtime > newtime: deploy = True deploy_file(rulefile, dest) else: deploy = True deploy_file(rulefile, dest) return deploy def get_counter(counterfile): """Get the counter value""" try: version_num = open(counterfile).read() version_num = int(version_num) + 1 except (ValueError, IOError): version_num = 1 create_file(counterfile, "%d" % version_num) except BaseException as msg: raise SaChannelUpdateError(msg) return version_num def update_dns(config, record, sa_version): "Update the DNS record" try: domain = config.get('domain_name', 'sa.baruwa.com.') dns_key = config.get('domain_key') dns_ip = config.get('domain_ip', '127.0.0.1') keyring = tsigkeyring.from_text({domain: dns_key}) transaction = update.Update( domain, keyring=keyring, keyalgorithm=tsig.HMAC_SHA512) txtrecord = '%s.%s' % (sa_version, domain) transaction.replace(txtrecord, 120, 'txt', record) query.tcp(transaction, dns_ip) return True except DNSException, msg: raise SaChannelUpdateDNSError(msg) def sign(config, s_filename): """sign the package""" gpg_home = config.get('gpg_dir', '/var/lib/sachannelupdate/gnupg') gpg_pass = config.get('gpg_passphrase') gpg_keyid = config.get('gpg_keyid') gpg = GPG(gnupghome=gpg_home) try: plaintext = open(s_filename, 'rb') signature = gpg.sign_file( plaintext, keyid=gpg_keyid, passphrase=gpg_pass, detach=True) with open('%s.asc' % s_filename, 'wb') as handle: handle.write(str(signature)) finally: if 'plaintext' in locals(): plaintext.close() def hash_file(tar_filename): """hash the file""" hasher = sha1() with open(tar_filename, 'rb') as afile: buf = afile.read(BLOCKSIZE) while len(buf) > 0: hasher.update(buf) buf = afile.read(BLOCKSIZE) data = HASHTMPL % (hasher.hexdigest(), os.path.basename(tar_filename)) create_file('%s.sha1' % tar_filename, data) def upload(config, remote_loc, u_filename): """Upload the files""" rcode = False try: sftp, transport = get_sftp_conn(config) remote_dir = get_remote_path(remote_loc) for part in ['sha1', 'asc']: local_file = '%s.%s' % (u_filename, part) remote_file = os.path.join(remote_dir, local_file) sftp.put(local_file, remote_file) sftp.put(remote_dir, os.path.join(remote_dir, u_filename)) rcode = True except BaseException: pass finally: if 'transport' in locals(): transport.close() return rcode def queue_files(dirpath, queue): """Add files in a directory to a queue""" for root, _, files in os.walk(os.path.abspath(dirpath)): if not files: continue for filename in files: queue.put(os.path.join(root, filename)) def get_cf_files(path, queue): """Get rule files in a directory and put them in a queue""" for root, _, files in os.walk(os.path.abspath(path)): if not files: continue for filename in files: fullname = os.path.join(root, filename) if os.path.isfile(fullname) and fullname.endswith('.cf') or \ fullname.endswith('.post'): queue.put(fullname) def cleanup(dest, tardir, counterfile): """Remove existing rules""" thefiles = Queue() # dest directory files queue_files(dest, thefiles) # tar directory files queue_files(tardir, thefiles) while not thefiles.empty(): d_file = thefiles.get() info("Deleting file: %s" % d_file) os.unlink(d_file) if os.path.exists(counterfile): info("Deleting the counter file %s" % counterfile) os.unlink(counterfile) def check_required(config): """Validate the input""" if config.get('domain_key') is None: raise CfgError("The domain_key option is required") if config.get('remote_loc') is None: raise CfgError("The remote_location option is required") if config.get('gpg_keyid') is None: raise CfgError("The gpg_keyid option is required") def entry(config, delete_files=None): """Main function""" home_dir = config.get('home_dir', '/var/lib/sachannelupdate') dns_ver = config.get('spamassassin_version', '1.4.3') remote_loc = config.get('remote_location') rule_dir = os.path.join(home_dir, 'rules') dest = os.path.join(home_dir, 'deploy') tardir = os.path.join(home_dir, 'archives') counterfile = os.path.join(home_dir, 'db', 'counters') check_required(config) if delete_files: cleanup(dest, tardir, counterfile) return cffiles = Queue() get_cf_files(rule_dir, cffiles) if process(dest, cffiles): version = get_counter(counterfile) filename = '%s.tar.gz' % version path = os.path.join(tardir, filename) package(dest, tardir, version) sign(config, path) hash_file(path) if upload(config, remote_loc, path): if update_dns(config, str(version), dns_ver): create_file(counterfile, "%d" % version)
akissa/sachannelupdate
sachannelupdate/base.py
package
python
def package(dest, tardir, p_version): os.chdir(dest) p_filename = '%s.tar.gz' % p_version p_path = os.path.join(tardir, p_filename) tar = tarfile.open(p_path, mode='w:gz') for cf_file in os.listdir('.'): if os.path.isfile(cf_file): tar.add(cf_file) tar.close()
Package files
train
https://github.com/akissa/sachannelupdate/blob/a1c3c3d86b874f9c92c2407e2608963165d3ae98/sachannelupdate/base.py#L74-L83
null
# -*- coding: utf-8 -*- # vim: ai ts=4 sts=4 et sw=4 # sachannelupdate - Utility for pushing updates to Spamassassin update channels # Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Affero General Public License as published # by the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Affero General Public License for more details. # # You should have received a copy of the GNU Affero General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. """ sachannelupdate: Utility for pushing updates to Spamassassin update channels Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> """ import os import tarfile import datetime from Queue import Queue from hashlib import sha1 from datetime import datetime from gnupg import GPG from dns.exception import DNSException from dns import tsig, query, tsigkeyring, update from sachannelupdate.utils import info from sachannelupdate.exceptions import SaChannelUpdateConfigError \ as CfgError, SaChannelUpdateDNSError, SaChannelUpdateError from sachannelupdate.transports import get_sftp_conn, get_remote_path BLOCKSIZE = 65536 HASHTMPL = """%s %s\n""" def create_file(name, content): "Generic to write file" with open(name, 'w') as writefile: writefile.write(content) def getfiles(qfiles, dirname, names): """Get rule files in a directory""" for name in names: fullname = os.path.join(dirname, name) if os.path.isfile(fullname) and \ fullname.endswith('.cf') or \ fullname.endswith('.post'): qfiles.put(fullname) def deploy_file(source, dest): """Deploy a file""" date = datetime.utcnow().strftime('%Y-%m-%d') shandle = open(source) with open(dest, 'w') as handle: for line in shandle: if line == '# Updated: %date%\n': newline = '# Updated: %s\n' % date else: newline = line handle.write(newline) handle.flush() shandle.close() def process(dest, rulefiles): """process rules""" deploy = False while not rulefiles.empty(): rulefile = rulefiles.get() base = os.path.basename(rulefile) dest = os.path.join(dest, base) if os.path.exists(dest): # check if older oldtime = os.stat(rulefile).st_mtime newtime = os.stat(dest).st_mtime if oldtime > newtime: deploy = True deploy_file(rulefile, dest) else: deploy = True deploy_file(rulefile, dest) return deploy def get_counter(counterfile): """Get the counter value""" try: version_num = open(counterfile).read() version_num = int(version_num) + 1 except (ValueError, IOError): version_num = 1 create_file(counterfile, "%d" % version_num) except BaseException as msg: raise SaChannelUpdateError(msg) return version_num def update_dns(config, record, sa_version): "Update the DNS record" try: domain = config.get('domain_name', 'sa.baruwa.com.') dns_key = config.get('domain_key') dns_ip = config.get('domain_ip', '127.0.0.1') keyring = tsigkeyring.from_text({domain: dns_key}) transaction = update.Update( domain, keyring=keyring, keyalgorithm=tsig.HMAC_SHA512) txtrecord = '%s.%s' % (sa_version, domain) transaction.replace(txtrecord, 120, 'txt', record) query.tcp(transaction, dns_ip) return True except DNSException, msg: raise SaChannelUpdateDNSError(msg) def sign(config, s_filename): """sign the package""" gpg_home = config.get('gpg_dir', '/var/lib/sachannelupdate/gnupg') gpg_pass = config.get('gpg_passphrase') gpg_keyid = config.get('gpg_keyid') gpg = GPG(gnupghome=gpg_home) try: plaintext = open(s_filename, 'rb') signature = gpg.sign_file( plaintext, keyid=gpg_keyid, passphrase=gpg_pass, detach=True) with open('%s.asc' % s_filename, 'wb') as handle: handle.write(str(signature)) finally: if 'plaintext' in locals(): plaintext.close() def hash_file(tar_filename): """hash the file""" hasher = sha1() with open(tar_filename, 'rb') as afile: buf = afile.read(BLOCKSIZE) while len(buf) > 0: hasher.update(buf) buf = afile.read(BLOCKSIZE) data = HASHTMPL % (hasher.hexdigest(), os.path.basename(tar_filename)) create_file('%s.sha1' % tar_filename, data) def upload(config, remote_loc, u_filename): """Upload the files""" rcode = False try: sftp, transport = get_sftp_conn(config) remote_dir = get_remote_path(remote_loc) for part in ['sha1', 'asc']: local_file = '%s.%s' % (u_filename, part) remote_file = os.path.join(remote_dir, local_file) sftp.put(local_file, remote_file) sftp.put(remote_dir, os.path.join(remote_dir, u_filename)) rcode = True except BaseException: pass finally: if 'transport' in locals(): transport.close() return rcode def queue_files(dirpath, queue): """Add files in a directory to a queue""" for root, _, files in os.walk(os.path.abspath(dirpath)): if not files: continue for filename in files: queue.put(os.path.join(root, filename)) def get_cf_files(path, queue): """Get rule files in a directory and put them in a queue""" for root, _, files in os.walk(os.path.abspath(path)): if not files: continue for filename in files: fullname = os.path.join(root, filename) if os.path.isfile(fullname) and fullname.endswith('.cf') or \ fullname.endswith('.post'): queue.put(fullname) def cleanup(dest, tardir, counterfile): """Remove existing rules""" thefiles = Queue() # dest directory files queue_files(dest, thefiles) # tar directory files queue_files(tardir, thefiles) while not thefiles.empty(): d_file = thefiles.get() info("Deleting file: %s" % d_file) os.unlink(d_file) if os.path.exists(counterfile): info("Deleting the counter file %s" % counterfile) os.unlink(counterfile) def check_required(config): """Validate the input""" if config.get('domain_key') is None: raise CfgError("The domain_key option is required") if config.get('remote_loc') is None: raise CfgError("The remote_location option is required") if config.get('gpg_keyid') is None: raise CfgError("The gpg_keyid option is required") def entry(config, delete_files=None): """Main function""" home_dir = config.get('home_dir', '/var/lib/sachannelupdate') dns_ver = config.get('spamassassin_version', '1.4.3') remote_loc = config.get('remote_location') rule_dir = os.path.join(home_dir, 'rules') dest = os.path.join(home_dir, 'deploy') tardir = os.path.join(home_dir, 'archives') counterfile = os.path.join(home_dir, 'db', 'counters') check_required(config) if delete_files: cleanup(dest, tardir, counterfile) return cffiles = Queue() get_cf_files(rule_dir, cffiles) if process(dest, cffiles): version = get_counter(counterfile) filename = '%s.tar.gz' % version path = os.path.join(tardir, filename) package(dest, tardir, version) sign(config, path) hash_file(path) if upload(config, remote_loc, path): if update_dns(config, str(version), dns_ver): create_file(counterfile, "%d" % version)
akissa/sachannelupdate
sachannelupdate/base.py
process
python
def process(dest, rulefiles): deploy = False while not rulefiles.empty(): rulefile = rulefiles.get() base = os.path.basename(rulefile) dest = os.path.join(dest, base) if os.path.exists(dest): # check if older oldtime = os.stat(rulefile).st_mtime newtime = os.stat(dest).st_mtime if oldtime > newtime: deploy = True deploy_file(rulefile, dest) else: deploy = True deploy_file(rulefile, dest) return deploy
process rules
train
https://github.com/akissa/sachannelupdate/blob/a1c3c3d86b874f9c92c2407e2608963165d3ae98/sachannelupdate/base.py#L86-L103
[ "def deploy_file(source, dest):\n \"\"\"Deploy a file\"\"\"\n date = datetime.utcnow().strftime('%Y-%m-%d')\n shandle = open(source)\n with open(dest, 'w') as handle:\n for line in shandle:\n if line == '# Updated: %date%\\n':\n newline = '# Updated: %s\\n' % date\n else:\n newline = line\n handle.write(newline)\n handle.flush()\n shandle.close()\n" ]
# -*- coding: utf-8 -*- # vim: ai ts=4 sts=4 et sw=4 # sachannelupdate - Utility for pushing updates to Spamassassin update channels # Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Affero General Public License as published # by the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Affero General Public License for more details. # # You should have received a copy of the GNU Affero General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. """ sachannelupdate: Utility for pushing updates to Spamassassin update channels Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> """ import os import tarfile import datetime from Queue import Queue from hashlib import sha1 from datetime import datetime from gnupg import GPG from dns.exception import DNSException from dns import tsig, query, tsigkeyring, update from sachannelupdate.utils import info from sachannelupdate.exceptions import SaChannelUpdateConfigError \ as CfgError, SaChannelUpdateDNSError, SaChannelUpdateError from sachannelupdate.transports import get_sftp_conn, get_remote_path BLOCKSIZE = 65536 HASHTMPL = """%s %s\n""" def create_file(name, content): "Generic to write file" with open(name, 'w') as writefile: writefile.write(content) def getfiles(qfiles, dirname, names): """Get rule files in a directory""" for name in names: fullname = os.path.join(dirname, name) if os.path.isfile(fullname) and \ fullname.endswith('.cf') or \ fullname.endswith('.post'): qfiles.put(fullname) def deploy_file(source, dest): """Deploy a file""" date = datetime.utcnow().strftime('%Y-%m-%d') shandle = open(source) with open(dest, 'w') as handle: for line in shandle: if line == '# Updated: %date%\n': newline = '# Updated: %s\n' % date else: newline = line handle.write(newline) handle.flush() shandle.close() def package(dest, tardir, p_version): """Package files""" os.chdir(dest) p_filename = '%s.tar.gz' % p_version p_path = os.path.join(tardir, p_filename) tar = tarfile.open(p_path, mode='w:gz') for cf_file in os.listdir('.'): if os.path.isfile(cf_file): tar.add(cf_file) tar.close() def get_counter(counterfile): """Get the counter value""" try: version_num = open(counterfile).read() version_num = int(version_num) + 1 except (ValueError, IOError): version_num = 1 create_file(counterfile, "%d" % version_num) except BaseException as msg: raise SaChannelUpdateError(msg) return version_num def update_dns(config, record, sa_version): "Update the DNS record" try: domain = config.get('domain_name', 'sa.baruwa.com.') dns_key = config.get('domain_key') dns_ip = config.get('domain_ip', '127.0.0.1') keyring = tsigkeyring.from_text({domain: dns_key}) transaction = update.Update( domain, keyring=keyring, keyalgorithm=tsig.HMAC_SHA512) txtrecord = '%s.%s' % (sa_version, domain) transaction.replace(txtrecord, 120, 'txt', record) query.tcp(transaction, dns_ip) return True except DNSException, msg: raise SaChannelUpdateDNSError(msg) def sign(config, s_filename): """sign the package""" gpg_home = config.get('gpg_dir', '/var/lib/sachannelupdate/gnupg') gpg_pass = config.get('gpg_passphrase') gpg_keyid = config.get('gpg_keyid') gpg = GPG(gnupghome=gpg_home) try: plaintext = open(s_filename, 'rb') signature = gpg.sign_file( plaintext, keyid=gpg_keyid, passphrase=gpg_pass, detach=True) with open('%s.asc' % s_filename, 'wb') as handle: handle.write(str(signature)) finally: if 'plaintext' in locals(): plaintext.close() def hash_file(tar_filename): """hash the file""" hasher = sha1() with open(tar_filename, 'rb') as afile: buf = afile.read(BLOCKSIZE) while len(buf) > 0: hasher.update(buf) buf = afile.read(BLOCKSIZE) data = HASHTMPL % (hasher.hexdigest(), os.path.basename(tar_filename)) create_file('%s.sha1' % tar_filename, data) def upload(config, remote_loc, u_filename): """Upload the files""" rcode = False try: sftp, transport = get_sftp_conn(config) remote_dir = get_remote_path(remote_loc) for part in ['sha1', 'asc']: local_file = '%s.%s' % (u_filename, part) remote_file = os.path.join(remote_dir, local_file) sftp.put(local_file, remote_file) sftp.put(remote_dir, os.path.join(remote_dir, u_filename)) rcode = True except BaseException: pass finally: if 'transport' in locals(): transport.close() return rcode def queue_files(dirpath, queue): """Add files in a directory to a queue""" for root, _, files in os.walk(os.path.abspath(dirpath)): if not files: continue for filename in files: queue.put(os.path.join(root, filename)) def get_cf_files(path, queue): """Get rule files in a directory and put them in a queue""" for root, _, files in os.walk(os.path.abspath(path)): if not files: continue for filename in files: fullname = os.path.join(root, filename) if os.path.isfile(fullname) and fullname.endswith('.cf') or \ fullname.endswith('.post'): queue.put(fullname) def cleanup(dest, tardir, counterfile): """Remove existing rules""" thefiles = Queue() # dest directory files queue_files(dest, thefiles) # tar directory files queue_files(tardir, thefiles) while not thefiles.empty(): d_file = thefiles.get() info("Deleting file: %s" % d_file) os.unlink(d_file) if os.path.exists(counterfile): info("Deleting the counter file %s" % counterfile) os.unlink(counterfile) def check_required(config): """Validate the input""" if config.get('domain_key') is None: raise CfgError("The domain_key option is required") if config.get('remote_loc') is None: raise CfgError("The remote_location option is required") if config.get('gpg_keyid') is None: raise CfgError("The gpg_keyid option is required") def entry(config, delete_files=None): """Main function""" home_dir = config.get('home_dir', '/var/lib/sachannelupdate') dns_ver = config.get('spamassassin_version', '1.4.3') remote_loc = config.get('remote_location') rule_dir = os.path.join(home_dir, 'rules') dest = os.path.join(home_dir, 'deploy') tardir = os.path.join(home_dir, 'archives') counterfile = os.path.join(home_dir, 'db', 'counters') check_required(config) if delete_files: cleanup(dest, tardir, counterfile) return cffiles = Queue() get_cf_files(rule_dir, cffiles) if process(dest, cffiles): version = get_counter(counterfile) filename = '%s.tar.gz' % version path = os.path.join(tardir, filename) package(dest, tardir, version) sign(config, path) hash_file(path) if upload(config, remote_loc, path): if update_dns(config, str(version), dns_ver): create_file(counterfile, "%d" % version)
akissa/sachannelupdate
sachannelupdate/base.py
get_counter
python
def get_counter(counterfile): try: version_num = open(counterfile).read() version_num = int(version_num) + 1 except (ValueError, IOError): version_num = 1 create_file(counterfile, "%d" % version_num) except BaseException as msg: raise SaChannelUpdateError(msg) return version_num
Get the counter value
train
https://github.com/akissa/sachannelupdate/blob/a1c3c3d86b874f9c92c2407e2608963165d3ae98/sachannelupdate/base.py#L106-L116
[ "def create_file(name, content):\n \"Generic to write file\"\n with open(name, 'w') as writefile:\n writefile.write(content)\n" ]
# -*- coding: utf-8 -*- # vim: ai ts=4 sts=4 et sw=4 # sachannelupdate - Utility for pushing updates to Spamassassin update channels # Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Affero General Public License as published # by the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Affero General Public License for more details. # # You should have received a copy of the GNU Affero General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. """ sachannelupdate: Utility for pushing updates to Spamassassin update channels Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> """ import os import tarfile import datetime from Queue import Queue from hashlib import sha1 from datetime import datetime from gnupg import GPG from dns.exception import DNSException from dns import tsig, query, tsigkeyring, update from sachannelupdate.utils import info from sachannelupdate.exceptions import SaChannelUpdateConfigError \ as CfgError, SaChannelUpdateDNSError, SaChannelUpdateError from sachannelupdate.transports import get_sftp_conn, get_remote_path BLOCKSIZE = 65536 HASHTMPL = """%s %s\n""" def create_file(name, content): "Generic to write file" with open(name, 'w') as writefile: writefile.write(content) def getfiles(qfiles, dirname, names): """Get rule files in a directory""" for name in names: fullname = os.path.join(dirname, name) if os.path.isfile(fullname) and \ fullname.endswith('.cf') or \ fullname.endswith('.post'): qfiles.put(fullname) def deploy_file(source, dest): """Deploy a file""" date = datetime.utcnow().strftime('%Y-%m-%d') shandle = open(source) with open(dest, 'w') as handle: for line in shandle: if line == '# Updated: %date%\n': newline = '# Updated: %s\n' % date else: newline = line handle.write(newline) handle.flush() shandle.close() def package(dest, tardir, p_version): """Package files""" os.chdir(dest) p_filename = '%s.tar.gz' % p_version p_path = os.path.join(tardir, p_filename) tar = tarfile.open(p_path, mode='w:gz') for cf_file in os.listdir('.'): if os.path.isfile(cf_file): tar.add(cf_file) tar.close() def process(dest, rulefiles): """process rules""" deploy = False while not rulefiles.empty(): rulefile = rulefiles.get() base = os.path.basename(rulefile) dest = os.path.join(dest, base) if os.path.exists(dest): # check if older oldtime = os.stat(rulefile).st_mtime newtime = os.stat(dest).st_mtime if oldtime > newtime: deploy = True deploy_file(rulefile, dest) else: deploy = True deploy_file(rulefile, dest) return deploy def update_dns(config, record, sa_version): "Update the DNS record" try: domain = config.get('domain_name', 'sa.baruwa.com.') dns_key = config.get('domain_key') dns_ip = config.get('domain_ip', '127.0.0.1') keyring = tsigkeyring.from_text({domain: dns_key}) transaction = update.Update( domain, keyring=keyring, keyalgorithm=tsig.HMAC_SHA512) txtrecord = '%s.%s' % (sa_version, domain) transaction.replace(txtrecord, 120, 'txt', record) query.tcp(transaction, dns_ip) return True except DNSException, msg: raise SaChannelUpdateDNSError(msg) def sign(config, s_filename): """sign the package""" gpg_home = config.get('gpg_dir', '/var/lib/sachannelupdate/gnupg') gpg_pass = config.get('gpg_passphrase') gpg_keyid = config.get('gpg_keyid') gpg = GPG(gnupghome=gpg_home) try: plaintext = open(s_filename, 'rb') signature = gpg.sign_file( plaintext, keyid=gpg_keyid, passphrase=gpg_pass, detach=True) with open('%s.asc' % s_filename, 'wb') as handle: handle.write(str(signature)) finally: if 'plaintext' in locals(): plaintext.close() def hash_file(tar_filename): """hash the file""" hasher = sha1() with open(tar_filename, 'rb') as afile: buf = afile.read(BLOCKSIZE) while len(buf) > 0: hasher.update(buf) buf = afile.read(BLOCKSIZE) data = HASHTMPL % (hasher.hexdigest(), os.path.basename(tar_filename)) create_file('%s.sha1' % tar_filename, data) def upload(config, remote_loc, u_filename): """Upload the files""" rcode = False try: sftp, transport = get_sftp_conn(config) remote_dir = get_remote_path(remote_loc) for part in ['sha1', 'asc']: local_file = '%s.%s' % (u_filename, part) remote_file = os.path.join(remote_dir, local_file) sftp.put(local_file, remote_file) sftp.put(remote_dir, os.path.join(remote_dir, u_filename)) rcode = True except BaseException: pass finally: if 'transport' in locals(): transport.close() return rcode def queue_files(dirpath, queue): """Add files in a directory to a queue""" for root, _, files in os.walk(os.path.abspath(dirpath)): if not files: continue for filename in files: queue.put(os.path.join(root, filename)) def get_cf_files(path, queue): """Get rule files in a directory and put them in a queue""" for root, _, files in os.walk(os.path.abspath(path)): if not files: continue for filename in files: fullname = os.path.join(root, filename) if os.path.isfile(fullname) and fullname.endswith('.cf') or \ fullname.endswith('.post'): queue.put(fullname) def cleanup(dest, tardir, counterfile): """Remove existing rules""" thefiles = Queue() # dest directory files queue_files(dest, thefiles) # tar directory files queue_files(tardir, thefiles) while not thefiles.empty(): d_file = thefiles.get() info("Deleting file: %s" % d_file) os.unlink(d_file) if os.path.exists(counterfile): info("Deleting the counter file %s" % counterfile) os.unlink(counterfile) def check_required(config): """Validate the input""" if config.get('domain_key') is None: raise CfgError("The domain_key option is required") if config.get('remote_loc') is None: raise CfgError("The remote_location option is required") if config.get('gpg_keyid') is None: raise CfgError("The gpg_keyid option is required") def entry(config, delete_files=None): """Main function""" home_dir = config.get('home_dir', '/var/lib/sachannelupdate') dns_ver = config.get('spamassassin_version', '1.4.3') remote_loc = config.get('remote_location') rule_dir = os.path.join(home_dir, 'rules') dest = os.path.join(home_dir, 'deploy') tardir = os.path.join(home_dir, 'archives') counterfile = os.path.join(home_dir, 'db', 'counters') check_required(config) if delete_files: cleanup(dest, tardir, counterfile) return cffiles = Queue() get_cf_files(rule_dir, cffiles) if process(dest, cffiles): version = get_counter(counterfile) filename = '%s.tar.gz' % version path = os.path.join(tardir, filename) package(dest, tardir, version) sign(config, path) hash_file(path) if upload(config, remote_loc, path): if update_dns(config, str(version), dns_ver): create_file(counterfile, "%d" % version)
akissa/sachannelupdate
sachannelupdate/base.py
update_dns
python
def update_dns(config, record, sa_version): "Update the DNS record" try: domain = config.get('domain_name', 'sa.baruwa.com.') dns_key = config.get('domain_key') dns_ip = config.get('domain_ip', '127.0.0.1') keyring = tsigkeyring.from_text({domain: dns_key}) transaction = update.Update( domain, keyring=keyring, keyalgorithm=tsig.HMAC_SHA512) txtrecord = '%s.%s' % (sa_version, domain) transaction.replace(txtrecord, 120, 'txt', record) query.tcp(transaction, dns_ip) return True except DNSException, msg: raise SaChannelUpdateDNSError(msg)
Update the DNS record
train
https://github.com/akissa/sachannelupdate/blob/a1c3c3d86b874f9c92c2407e2608963165d3ae98/sachannelupdate/base.py#L119-L135
null
# -*- coding: utf-8 -*- # vim: ai ts=4 sts=4 et sw=4 # sachannelupdate - Utility for pushing updates to Spamassassin update channels # Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Affero General Public License as published # by the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Affero General Public License for more details. # # You should have received a copy of the GNU Affero General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. """ sachannelupdate: Utility for pushing updates to Spamassassin update channels Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> """ import os import tarfile import datetime from Queue import Queue from hashlib import sha1 from datetime import datetime from gnupg import GPG from dns.exception import DNSException from dns import tsig, query, tsigkeyring, update from sachannelupdate.utils import info from sachannelupdate.exceptions import SaChannelUpdateConfigError \ as CfgError, SaChannelUpdateDNSError, SaChannelUpdateError from sachannelupdate.transports import get_sftp_conn, get_remote_path BLOCKSIZE = 65536 HASHTMPL = """%s %s\n""" def create_file(name, content): "Generic to write file" with open(name, 'w') as writefile: writefile.write(content) def getfiles(qfiles, dirname, names): """Get rule files in a directory""" for name in names: fullname = os.path.join(dirname, name) if os.path.isfile(fullname) and \ fullname.endswith('.cf') or \ fullname.endswith('.post'): qfiles.put(fullname) def deploy_file(source, dest): """Deploy a file""" date = datetime.utcnow().strftime('%Y-%m-%d') shandle = open(source) with open(dest, 'w') as handle: for line in shandle: if line == '# Updated: %date%\n': newline = '# Updated: %s\n' % date else: newline = line handle.write(newline) handle.flush() shandle.close() def package(dest, tardir, p_version): """Package files""" os.chdir(dest) p_filename = '%s.tar.gz' % p_version p_path = os.path.join(tardir, p_filename) tar = tarfile.open(p_path, mode='w:gz') for cf_file in os.listdir('.'): if os.path.isfile(cf_file): tar.add(cf_file) tar.close() def process(dest, rulefiles): """process rules""" deploy = False while not rulefiles.empty(): rulefile = rulefiles.get() base = os.path.basename(rulefile) dest = os.path.join(dest, base) if os.path.exists(dest): # check if older oldtime = os.stat(rulefile).st_mtime newtime = os.stat(dest).st_mtime if oldtime > newtime: deploy = True deploy_file(rulefile, dest) else: deploy = True deploy_file(rulefile, dest) return deploy def get_counter(counterfile): """Get the counter value""" try: version_num = open(counterfile).read() version_num = int(version_num) + 1 except (ValueError, IOError): version_num = 1 create_file(counterfile, "%d" % version_num) except BaseException as msg: raise SaChannelUpdateError(msg) return version_num def sign(config, s_filename): """sign the package""" gpg_home = config.get('gpg_dir', '/var/lib/sachannelupdate/gnupg') gpg_pass = config.get('gpg_passphrase') gpg_keyid = config.get('gpg_keyid') gpg = GPG(gnupghome=gpg_home) try: plaintext = open(s_filename, 'rb') signature = gpg.sign_file( plaintext, keyid=gpg_keyid, passphrase=gpg_pass, detach=True) with open('%s.asc' % s_filename, 'wb') as handle: handle.write(str(signature)) finally: if 'plaintext' in locals(): plaintext.close() def hash_file(tar_filename): """hash the file""" hasher = sha1() with open(tar_filename, 'rb') as afile: buf = afile.read(BLOCKSIZE) while len(buf) > 0: hasher.update(buf) buf = afile.read(BLOCKSIZE) data = HASHTMPL % (hasher.hexdigest(), os.path.basename(tar_filename)) create_file('%s.sha1' % tar_filename, data) def upload(config, remote_loc, u_filename): """Upload the files""" rcode = False try: sftp, transport = get_sftp_conn(config) remote_dir = get_remote_path(remote_loc) for part in ['sha1', 'asc']: local_file = '%s.%s' % (u_filename, part) remote_file = os.path.join(remote_dir, local_file) sftp.put(local_file, remote_file) sftp.put(remote_dir, os.path.join(remote_dir, u_filename)) rcode = True except BaseException: pass finally: if 'transport' in locals(): transport.close() return rcode def queue_files(dirpath, queue): """Add files in a directory to a queue""" for root, _, files in os.walk(os.path.abspath(dirpath)): if not files: continue for filename in files: queue.put(os.path.join(root, filename)) def get_cf_files(path, queue): """Get rule files in a directory and put them in a queue""" for root, _, files in os.walk(os.path.abspath(path)): if not files: continue for filename in files: fullname = os.path.join(root, filename) if os.path.isfile(fullname) and fullname.endswith('.cf') or \ fullname.endswith('.post'): queue.put(fullname) def cleanup(dest, tardir, counterfile): """Remove existing rules""" thefiles = Queue() # dest directory files queue_files(dest, thefiles) # tar directory files queue_files(tardir, thefiles) while not thefiles.empty(): d_file = thefiles.get() info("Deleting file: %s" % d_file) os.unlink(d_file) if os.path.exists(counterfile): info("Deleting the counter file %s" % counterfile) os.unlink(counterfile) def check_required(config): """Validate the input""" if config.get('domain_key') is None: raise CfgError("The domain_key option is required") if config.get('remote_loc') is None: raise CfgError("The remote_location option is required") if config.get('gpg_keyid') is None: raise CfgError("The gpg_keyid option is required") def entry(config, delete_files=None): """Main function""" home_dir = config.get('home_dir', '/var/lib/sachannelupdate') dns_ver = config.get('spamassassin_version', '1.4.3') remote_loc = config.get('remote_location') rule_dir = os.path.join(home_dir, 'rules') dest = os.path.join(home_dir, 'deploy') tardir = os.path.join(home_dir, 'archives') counterfile = os.path.join(home_dir, 'db', 'counters') check_required(config) if delete_files: cleanup(dest, tardir, counterfile) return cffiles = Queue() get_cf_files(rule_dir, cffiles) if process(dest, cffiles): version = get_counter(counterfile) filename = '%s.tar.gz' % version path = os.path.join(tardir, filename) package(dest, tardir, version) sign(config, path) hash_file(path) if upload(config, remote_loc, path): if update_dns(config, str(version), dns_ver): create_file(counterfile, "%d" % version)
akissa/sachannelupdate
sachannelupdate/base.py
sign
python
def sign(config, s_filename): gpg_home = config.get('gpg_dir', '/var/lib/sachannelupdate/gnupg') gpg_pass = config.get('gpg_passphrase') gpg_keyid = config.get('gpg_keyid') gpg = GPG(gnupghome=gpg_home) try: plaintext = open(s_filename, 'rb') signature = gpg.sign_file( plaintext, keyid=gpg_keyid, passphrase=gpg_pass, detach=True) with open('%s.asc' % s_filename, 'wb') as handle: handle.write(str(signature)) finally: if 'plaintext' in locals(): plaintext.close()
sign the package
train
https://github.com/akissa/sachannelupdate/blob/a1c3c3d86b874f9c92c2407e2608963165d3ae98/sachannelupdate/base.py#L138-L152
null
# -*- coding: utf-8 -*- # vim: ai ts=4 sts=4 et sw=4 # sachannelupdate - Utility for pushing updates to Spamassassin update channels # Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Affero General Public License as published # by the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Affero General Public License for more details. # # You should have received a copy of the GNU Affero General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. """ sachannelupdate: Utility for pushing updates to Spamassassin update channels Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> """ import os import tarfile import datetime from Queue import Queue from hashlib import sha1 from datetime import datetime from gnupg import GPG from dns.exception import DNSException from dns import tsig, query, tsigkeyring, update from sachannelupdate.utils import info from sachannelupdate.exceptions import SaChannelUpdateConfigError \ as CfgError, SaChannelUpdateDNSError, SaChannelUpdateError from sachannelupdate.transports import get_sftp_conn, get_remote_path BLOCKSIZE = 65536 HASHTMPL = """%s %s\n""" def create_file(name, content): "Generic to write file" with open(name, 'w') as writefile: writefile.write(content) def getfiles(qfiles, dirname, names): """Get rule files in a directory""" for name in names: fullname = os.path.join(dirname, name) if os.path.isfile(fullname) and \ fullname.endswith('.cf') or \ fullname.endswith('.post'): qfiles.put(fullname) def deploy_file(source, dest): """Deploy a file""" date = datetime.utcnow().strftime('%Y-%m-%d') shandle = open(source) with open(dest, 'w') as handle: for line in shandle: if line == '# Updated: %date%\n': newline = '# Updated: %s\n' % date else: newline = line handle.write(newline) handle.flush() shandle.close() def package(dest, tardir, p_version): """Package files""" os.chdir(dest) p_filename = '%s.tar.gz' % p_version p_path = os.path.join(tardir, p_filename) tar = tarfile.open(p_path, mode='w:gz') for cf_file in os.listdir('.'): if os.path.isfile(cf_file): tar.add(cf_file) tar.close() def process(dest, rulefiles): """process rules""" deploy = False while not rulefiles.empty(): rulefile = rulefiles.get() base = os.path.basename(rulefile) dest = os.path.join(dest, base) if os.path.exists(dest): # check if older oldtime = os.stat(rulefile).st_mtime newtime = os.stat(dest).st_mtime if oldtime > newtime: deploy = True deploy_file(rulefile, dest) else: deploy = True deploy_file(rulefile, dest) return deploy def get_counter(counterfile): """Get the counter value""" try: version_num = open(counterfile).read() version_num = int(version_num) + 1 except (ValueError, IOError): version_num = 1 create_file(counterfile, "%d" % version_num) except BaseException as msg: raise SaChannelUpdateError(msg) return version_num def update_dns(config, record, sa_version): "Update the DNS record" try: domain = config.get('domain_name', 'sa.baruwa.com.') dns_key = config.get('domain_key') dns_ip = config.get('domain_ip', '127.0.0.1') keyring = tsigkeyring.from_text({domain: dns_key}) transaction = update.Update( domain, keyring=keyring, keyalgorithm=tsig.HMAC_SHA512) txtrecord = '%s.%s' % (sa_version, domain) transaction.replace(txtrecord, 120, 'txt', record) query.tcp(transaction, dns_ip) return True except DNSException, msg: raise SaChannelUpdateDNSError(msg) def hash_file(tar_filename): """hash the file""" hasher = sha1() with open(tar_filename, 'rb') as afile: buf = afile.read(BLOCKSIZE) while len(buf) > 0: hasher.update(buf) buf = afile.read(BLOCKSIZE) data = HASHTMPL % (hasher.hexdigest(), os.path.basename(tar_filename)) create_file('%s.sha1' % tar_filename, data) def upload(config, remote_loc, u_filename): """Upload the files""" rcode = False try: sftp, transport = get_sftp_conn(config) remote_dir = get_remote_path(remote_loc) for part in ['sha1', 'asc']: local_file = '%s.%s' % (u_filename, part) remote_file = os.path.join(remote_dir, local_file) sftp.put(local_file, remote_file) sftp.put(remote_dir, os.path.join(remote_dir, u_filename)) rcode = True except BaseException: pass finally: if 'transport' in locals(): transport.close() return rcode def queue_files(dirpath, queue): """Add files in a directory to a queue""" for root, _, files in os.walk(os.path.abspath(dirpath)): if not files: continue for filename in files: queue.put(os.path.join(root, filename)) def get_cf_files(path, queue): """Get rule files in a directory and put them in a queue""" for root, _, files in os.walk(os.path.abspath(path)): if not files: continue for filename in files: fullname = os.path.join(root, filename) if os.path.isfile(fullname) and fullname.endswith('.cf') or \ fullname.endswith('.post'): queue.put(fullname) def cleanup(dest, tardir, counterfile): """Remove existing rules""" thefiles = Queue() # dest directory files queue_files(dest, thefiles) # tar directory files queue_files(tardir, thefiles) while not thefiles.empty(): d_file = thefiles.get() info("Deleting file: %s" % d_file) os.unlink(d_file) if os.path.exists(counterfile): info("Deleting the counter file %s" % counterfile) os.unlink(counterfile) def check_required(config): """Validate the input""" if config.get('domain_key') is None: raise CfgError("The domain_key option is required") if config.get('remote_loc') is None: raise CfgError("The remote_location option is required") if config.get('gpg_keyid') is None: raise CfgError("The gpg_keyid option is required") def entry(config, delete_files=None): """Main function""" home_dir = config.get('home_dir', '/var/lib/sachannelupdate') dns_ver = config.get('spamassassin_version', '1.4.3') remote_loc = config.get('remote_location') rule_dir = os.path.join(home_dir, 'rules') dest = os.path.join(home_dir, 'deploy') tardir = os.path.join(home_dir, 'archives') counterfile = os.path.join(home_dir, 'db', 'counters') check_required(config) if delete_files: cleanup(dest, tardir, counterfile) return cffiles = Queue() get_cf_files(rule_dir, cffiles) if process(dest, cffiles): version = get_counter(counterfile) filename = '%s.tar.gz' % version path = os.path.join(tardir, filename) package(dest, tardir, version) sign(config, path) hash_file(path) if upload(config, remote_loc, path): if update_dns(config, str(version), dns_ver): create_file(counterfile, "%d" % version)
akissa/sachannelupdate
sachannelupdate/base.py
hash_file
python
def hash_file(tar_filename): hasher = sha1() with open(tar_filename, 'rb') as afile: buf = afile.read(BLOCKSIZE) while len(buf) > 0: hasher.update(buf) buf = afile.read(BLOCKSIZE) data = HASHTMPL % (hasher.hexdigest(), os.path.basename(tar_filename)) create_file('%s.sha1' % tar_filename, data)
hash the file
train
https://github.com/akissa/sachannelupdate/blob/a1c3c3d86b874f9c92c2407e2608963165d3ae98/sachannelupdate/base.py#L155-L164
[ "def create_file(name, content):\n \"Generic to write file\"\n with open(name, 'w') as writefile:\n writefile.write(content)\n" ]
# -*- coding: utf-8 -*- # vim: ai ts=4 sts=4 et sw=4 # sachannelupdate - Utility for pushing updates to Spamassassin update channels # Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Affero General Public License as published # by the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Affero General Public License for more details. # # You should have received a copy of the GNU Affero General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. """ sachannelupdate: Utility for pushing updates to Spamassassin update channels Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> """ import os import tarfile import datetime from Queue import Queue from hashlib import sha1 from datetime import datetime from gnupg import GPG from dns.exception import DNSException from dns import tsig, query, tsigkeyring, update from sachannelupdate.utils import info from sachannelupdate.exceptions import SaChannelUpdateConfigError \ as CfgError, SaChannelUpdateDNSError, SaChannelUpdateError from sachannelupdate.transports import get_sftp_conn, get_remote_path BLOCKSIZE = 65536 HASHTMPL = """%s %s\n""" def create_file(name, content): "Generic to write file" with open(name, 'w') as writefile: writefile.write(content) def getfiles(qfiles, dirname, names): """Get rule files in a directory""" for name in names: fullname = os.path.join(dirname, name) if os.path.isfile(fullname) and \ fullname.endswith('.cf') or \ fullname.endswith('.post'): qfiles.put(fullname) def deploy_file(source, dest): """Deploy a file""" date = datetime.utcnow().strftime('%Y-%m-%d') shandle = open(source) with open(dest, 'w') as handle: for line in shandle: if line == '# Updated: %date%\n': newline = '# Updated: %s\n' % date else: newline = line handle.write(newline) handle.flush() shandle.close() def package(dest, tardir, p_version): """Package files""" os.chdir(dest) p_filename = '%s.tar.gz' % p_version p_path = os.path.join(tardir, p_filename) tar = tarfile.open(p_path, mode='w:gz') for cf_file in os.listdir('.'): if os.path.isfile(cf_file): tar.add(cf_file) tar.close() def process(dest, rulefiles): """process rules""" deploy = False while not rulefiles.empty(): rulefile = rulefiles.get() base = os.path.basename(rulefile) dest = os.path.join(dest, base) if os.path.exists(dest): # check if older oldtime = os.stat(rulefile).st_mtime newtime = os.stat(dest).st_mtime if oldtime > newtime: deploy = True deploy_file(rulefile, dest) else: deploy = True deploy_file(rulefile, dest) return deploy def get_counter(counterfile): """Get the counter value""" try: version_num = open(counterfile).read() version_num = int(version_num) + 1 except (ValueError, IOError): version_num = 1 create_file(counterfile, "%d" % version_num) except BaseException as msg: raise SaChannelUpdateError(msg) return version_num def update_dns(config, record, sa_version): "Update the DNS record" try: domain = config.get('domain_name', 'sa.baruwa.com.') dns_key = config.get('domain_key') dns_ip = config.get('domain_ip', '127.0.0.1') keyring = tsigkeyring.from_text({domain: dns_key}) transaction = update.Update( domain, keyring=keyring, keyalgorithm=tsig.HMAC_SHA512) txtrecord = '%s.%s' % (sa_version, domain) transaction.replace(txtrecord, 120, 'txt', record) query.tcp(transaction, dns_ip) return True except DNSException, msg: raise SaChannelUpdateDNSError(msg) def sign(config, s_filename): """sign the package""" gpg_home = config.get('gpg_dir', '/var/lib/sachannelupdate/gnupg') gpg_pass = config.get('gpg_passphrase') gpg_keyid = config.get('gpg_keyid') gpg = GPG(gnupghome=gpg_home) try: plaintext = open(s_filename, 'rb') signature = gpg.sign_file( plaintext, keyid=gpg_keyid, passphrase=gpg_pass, detach=True) with open('%s.asc' % s_filename, 'wb') as handle: handle.write(str(signature)) finally: if 'plaintext' in locals(): plaintext.close() def upload(config, remote_loc, u_filename): """Upload the files""" rcode = False try: sftp, transport = get_sftp_conn(config) remote_dir = get_remote_path(remote_loc) for part in ['sha1', 'asc']: local_file = '%s.%s' % (u_filename, part) remote_file = os.path.join(remote_dir, local_file) sftp.put(local_file, remote_file) sftp.put(remote_dir, os.path.join(remote_dir, u_filename)) rcode = True except BaseException: pass finally: if 'transport' in locals(): transport.close() return rcode def queue_files(dirpath, queue): """Add files in a directory to a queue""" for root, _, files in os.walk(os.path.abspath(dirpath)): if not files: continue for filename in files: queue.put(os.path.join(root, filename)) def get_cf_files(path, queue): """Get rule files in a directory and put them in a queue""" for root, _, files in os.walk(os.path.abspath(path)): if not files: continue for filename in files: fullname = os.path.join(root, filename) if os.path.isfile(fullname) and fullname.endswith('.cf') or \ fullname.endswith('.post'): queue.put(fullname) def cleanup(dest, tardir, counterfile): """Remove existing rules""" thefiles = Queue() # dest directory files queue_files(dest, thefiles) # tar directory files queue_files(tardir, thefiles) while not thefiles.empty(): d_file = thefiles.get() info("Deleting file: %s" % d_file) os.unlink(d_file) if os.path.exists(counterfile): info("Deleting the counter file %s" % counterfile) os.unlink(counterfile) def check_required(config): """Validate the input""" if config.get('domain_key') is None: raise CfgError("The domain_key option is required") if config.get('remote_loc') is None: raise CfgError("The remote_location option is required") if config.get('gpg_keyid') is None: raise CfgError("The gpg_keyid option is required") def entry(config, delete_files=None): """Main function""" home_dir = config.get('home_dir', '/var/lib/sachannelupdate') dns_ver = config.get('spamassassin_version', '1.4.3') remote_loc = config.get('remote_location') rule_dir = os.path.join(home_dir, 'rules') dest = os.path.join(home_dir, 'deploy') tardir = os.path.join(home_dir, 'archives') counterfile = os.path.join(home_dir, 'db', 'counters') check_required(config) if delete_files: cleanup(dest, tardir, counterfile) return cffiles = Queue() get_cf_files(rule_dir, cffiles) if process(dest, cffiles): version = get_counter(counterfile) filename = '%s.tar.gz' % version path = os.path.join(tardir, filename) package(dest, tardir, version) sign(config, path) hash_file(path) if upload(config, remote_loc, path): if update_dns(config, str(version), dns_ver): create_file(counterfile, "%d" % version)
akissa/sachannelupdate
sachannelupdate/base.py
upload
python
def upload(config, remote_loc, u_filename): rcode = False try: sftp, transport = get_sftp_conn(config) remote_dir = get_remote_path(remote_loc) for part in ['sha1', 'asc']: local_file = '%s.%s' % (u_filename, part) remote_file = os.path.join(remote_dir, local_file) sftp.put(local_file, remote_file) sftp.put(remote_dir, os.path.join(remote_dir, u_filename)) rcode = True except BaseException: pass finally: if 'transport' in locals(): transport.close() return rcode
Upload the files
train
https://github.com/akissa/sachannelupdate/blob/a1c3c3d86b874f9c92c2407e2608963165d3ae98/sachannelupdate/base.py#L167-L184
[ "def get_sftp_conn(config):\n \"\"\"Make a SFTP connection, returns sftp client and connection objects\"\"\"\n remote = config.get('remote_location')\n parts = urlparse(remote)\n\n if ':' in parts.netloc:\n hostname, port = parts.netloc.split(':')\n else:\n hostname = parts.netloc\n port = 22\n port = int(port)\n\n username = config.get('remote_username') or getuser()\n luser = get_local_user(username)\n sshdir = get_ssh_dir(config, luser)\n hostkey = get_host_keys(hostname, sshdir)\n\n try:\n sftp = None\n keys = get_ssh_keys(sshdir)\n transport = Transport((hostname, port))\n while not keys.empty():\n try:\n key = PKey.from_private_key_file(keys.get())\n transport.connect(\n hostkey=hostkey,\n username=username,\n password=None,\n pkey=key)\n sftp = SFTPClient.from_transport(transport)\n break\n except (PasswordRequiredException, SSHException):\n pass\n if sftp is None:\n raise SaChannelUpdateTransportError(\"SFTP connection failed\")\n return sftp, transport\n except BaseException as msg:\n raise SaChannelUpdateTransportError(msg)\n", "def get_remote_path(remote_location):\n \"\"\"Get the remote path from the remote location\"\"\"\n parts = urlparse(remote_location)\n return parts.path\n" ]
# -*- coding: utf-8 -*- # vim: ai ts=4 sts=4 et sw=4 # sachannelupdate - Utility for pushing updates to Spamassassin update channels # Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Affero General Public License as published # by the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Affero General Public License for more details. # # You should have received a copy of the GNU Affero General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. """ sachannelupdate: Utility for pushing updates to Spamassassin update channels Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> """ import os import tarfile import datetime from Queue import Queue from hashlib import sha1 from datetime import datetime from gnupg import GPG from dns.exception import DNSException from dns import tsig, query, tsigkeyring, update from sachannelupdate.utils import info from sachannelupdate.exceptions import SaChannelUpdateConfigError \ as CfgError, SaChannelUpdateDNSError, SaChannelUpdateError from sachannelupdate.transports import get_sftp_conn, get_remote_path BLOCKSIZE = 65536 HASHTMPL = """%s %s\n""" def create_file(name, content): "Generic to write file" with open(name, 'w') as writefile: writefile.write(content) def getfiles(qfiles, dirname, names): """Get rule files in a directory""" for name in names: fullname = os.path.join(dirname, name) if os.path.isfile(fullname) and \ fullname.endswith('.cf') or \ fullname.endswith('.post'): qfiles.put(fullname) def deploy_file(source, dest): """Deploy a file""" date = datetime.utcnow().strftime('%Y-%m-%d') shandle = open(source) with open(dest, 'w') as handle: for line in shandle: if line == '# Updated: %date%\n': newline = '# Updated: %s\n' % date else: newline = line handle.write(newline) handle.flush() shandle.close() def package(dest, tardir, p_version): """Package files""" os.chdir(dest) p_filename = '%s.tar.gz' % p_version p_path = os.path.join(tardir, p_filename) tar = tarfile.open(p_path, mode='w:gz') for cf_file in os.listdir('.'): if os.path.isfile(cf_file): tar.add(cf_file) tar.close() def process(dest, rulefiles): """process rules""" deploy = False while not rulefiles.empty(): rulefile = rulefiles.get() base = os.path.basename(rulefile) dest = os.path.join(dest, base) if os.path.exists(dest): # check if older oldtime = os.stat(rulefile).st_mtime newtime = os.stat(dest).st_mtime if oldtime > newtime: deploy = True deploy_file(rulefile, dest) else: deploy = True deploy_file(rulefile, dest) return deploy def get_counter(counterfile): """Get the counter value""" try: version_num = open(counterfile).read() version_num = int(version_num) + 1 except (ValueError, IOError): version_num = 1 create_file(counterfile, "%d" % version_num) except BaseException as msg: raise SaChannelUpdateError(msg) return version_num def update_dns(config, record, sa_version): "Update the DNS record" try: domain = config.get('domain_name', 'sa.baruwa.com.') dns_key = config.get('domain_key') dns_ip = config.get('domain_ip', '127.0.0.1') keyring = tsigkeyring.from_text({domain: dns_key}) transaction = update.Update( domain, keyring=keyring, keyalgorithm=tsig.HMAC_SHA512) txtrecord = '%s.%s' % (sa_version, domain) transaction.replace(txtrecord, 120, 'txt', record) query.tcp(transaction, dns_ip) return True except DNSException, msg: raise SaChannelUpdateDNSError(msg) def sign(config, s_filename): """sign the package""" gpg_home = config.get('gpg_dir', '/var/lib/sachannelupdate/gnupg') gpg_pass = config.get('gpg_passphrase') gpg_keyid = config.get('gpg_keyid') gpg = GPG(gnupghome=gpg_home) try: plaintext = open(s_filename, 'rb') signature = gpg.sign_file( plaintext, keyid=gpg_keyid, passphrase=gpg_pass, detach=True) with open('%s.asc' % s_filename, 'wb') as handle: handle.write(str(signature)) finally: if 'plaintext' in locals(): plaintext.close() def hash_file(tar_filename): """hash the file""" hasher = sha1() with open(tar_filename, 'rb') as afile: buf = afile.read(BLOCKSIZE) while len(buf) > 0: hasher.update(buf) buf = afile.read(BLOCKSIZE) data = HASHTMPL % (hasher.hexdigest(), os.path.basename(tar_filename)) create_file('%s.sha1' % tar_filename, data) def queue_files(dirpath, queue): """Add files in a directory to a queue""" for root, _, files in os.walk(os.path.abspath(dirpath)): if not files: continue for filename in files: queue.put(os.path.join(root, filename)) def get_cf_files(path, queue): """Get rule files in a directory and put them in a queue""" for root, _, files in os.walk(os.path.abspath(path)): if not files: continue for filename in files: fullname = os.path.join(root, filename) if os.path.isfile(fullname) and fullname.endswith('.cf') or \ fullname.endswith('.post'): queue.put(fullname) def cleanup(dest, tardir, counterfile): """Remove existing rules""" thefiles = Queue() # dest directory files queue_files(dest, thefiles) # tar directory files queue_files(tardir, thefiles) while not thefiles.empty(): d_file = thefiles.get() info("Deleting file: %s" % d_file) os.unlink(d_file) if os.path.exists(counterfile): info("Deleting the counter file %s" % counterfile) os.unlink(counterfile) def check_required(config): """Validate the input""" if config.get('domain_key') is None: raise CfgError("The domain_key option is required") if config.get('remote_loc') is None: raise CfgError("The remote_location option is required") if config.get('gpg_keyid') is None: raise CfgError("The gpg_keyid option is required") def entry(config, delete_files=None): """Main function""" home_dir = config.get('home_dir', '/var/lib/sachannelupdate') dns_ver = config.get('spamassassin_version', '1.4.3') remote_loc = config.get('remote_location') rule_dir = os.path.join(home_dir, 'rules') dest = os.path.join(home_dir, 'deploy') tardir = os.path.join(home_dir, 'archives') counterfile = os.path.join(home_dir, 'db', 'counters') check_required(config) if delete_files: cleanup(dest, tardir, counterfile) return cffiles = Queue() get_cf_files(rule_dir, cffiles) if process(dest, cffiles): version = get_counter(counterfile) filename = '%s.tar.gz' % version path = os.path.join(tardir, filename) package(dest, tardir, version) sign(config, path) hash_file(path) if upload(config, remote_loc, path): if update_dns(config, str(version), dns_ver): create_file(counterfile, "%d" % version)
akissa/sachannelupdate
sachannelupdate/base.py
queue_files
python
def queue_files(dirpath, queue): for root, _, files in os.walk(os.path.abspath(dirpath)): if not files: continue for filename in files: queue.put(os.path.join(root, filename))
Add files in a directory to a queue
train
https://github.com/akissa/sachannelupdate/blob/a1c3c3d86b874f9c92c2407e2608963165d3ae98/sachannelupdate/base.py#L187-L193
null
# -*- coding: utf-8 -*- # vim: ai ts=4 sts=4 et sw=4 # sachannelupdate - Utility for pushing updates to Spamassassin update channels # Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Affero General Public License as published # by the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Affero General Public License for more details. # # You should have received a copy of the GNU Affero General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. """ sachannelupdate: Utility for pushing updates to Spamassassin update channels Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> """ import os import tarfile import datetime from Queue import Queue from hashlib import sha1 from datetime import datetime from gnupg import GPG from dns.exception import DNSException from dns import tsig, query, tsigkeyring, update from sachannelupdate.utils import info from sachannelupdate.exceptions import SaChannelUpdateConfigError \ as CfgError, SaChannelUpdateDNSError, SaChannelUpdateError from sachannelupdate.transports import get_sftp_conn, get_remote_path BLOCKSIZE = 65536 HASHTMPL = """%s %s\n""" def create_file(name, content): "Generic to write file" with open(name, 'w') as writefile: writefile.write(content) def getfiles(qfiles, dirname, names): """Get rule files in a directory""" for name in names: fullname = os.path.join(dirname, name) if os.path.isfile(fullname) and \ fullname.endswith('.cf') or \ fullname.endswith('.post'): qfiles.put(fullname) def deploy_file(source, dest): """Deploy a file""" date = datetime.utcnow().strftime('%Y-%m-%d') shandle = open(source) with open(dest, 'w') as handle: for line in shandle: if line == '# Updated: %date%\n': newline = '# Updated: %s\n' % date else: newline = line handle.write(newline) handle.flush() shandle.close() def package(dest, tardir, p_version): """Package files""" os.chdir(dest) p_filename = '%s.tar.gz' % p_version p_path = os.path.join(tardir, p_filename) tar = tarfile.open(p_path, mode='w:gz') for cf_file in os.listdir('.'): if os.path.isfile(cf_file): tar.add(cf_file) tar.close() def process(dest, rulefiles): """process rules""" deploy = False while not rulefiles.empty(): rulefile = rulefiles.get() base = os.path.basename(rulefile) dest = os.path.join(dest, base) if os.path.exists(dest): # check if older oldtime = os.stat(rulefile).st_mtime newtime = os.stat(dest).st_mtime if oldtime > newtime: deploy = True deploy_file(rulefile, dest) else: deploy = True deploy_file(rulefile, dest) return deploy def get_counter(counterfile): """Get the counter value""" try: version_num = open(counterfile).read() version_num = int(version_num) + 1 except (ValueError, IOError): version_num = 1 create_file(counterfile, "%d" % version_num) except BaseException as msg: raise SaChannelUpdateError(msg) return version_num def update_dns(config, record, sa_version): "Update the DNS record" try: domain = config.get('domain_name', 'sa.baruwa.com.') dns_key = config.get('domain_key') dns_ip = config.get('domain_ip', '127.0.0.1') keyring = tsigkeyring.from_text({domain: dns_key}) transaction = update.Update( domain, keyring=keyring, keyalgorithm=tsig.HMAC_SHA512) txtrecord = '%s.%s' % (sa_version, domain) transaction.replace(txtrecord, 120, 'txt', record) query.tcp(transaction, dns_ip) return True except DNSException, msg: raise SaChannelUpdateDNSError(msg) def sign(config, s_filename): """sign the package""" gpg_home = config.get('gpg_dir', '/var/lib/sachannelupdate/gnupg') gpg_pass = config.get('gpg_passphrase') gpg_keyid = config.get('gpg_keyid') gpg = GPG(gnupghome=gpg_home) try: plaintext = open(s_filename, 'rb') signature = gpg.sign_file( plaintext, keyid=gpg_keyid, passphrase=gpg_pass, detach=True) with open('%s.asc' % s_filename, 'wb') as handle: handle.write(str(signature)) finally: if 'plaintext' in locals(): plaintext.close() def hash_file(tar_filename): """hash the file""" hasher = sha1() with open(tar_filename, 'rb') as afile: buf = afile.read(BLOCKSIZE) while len(buf) > 0: hasher.update(buf) buf = afile.read(BLOCKSIZE) data = HASHTMPL % (hasher.hexdigest(), os.path.basename(tar_filename)) create_file('%s.sha1' % tar_filename, data) def upload(config, remote_loc, u_filename): """Upload the files""" rcode = False try: sftp, transport = get_sftp_conn(config) remote_dir = get_remote_path(remote_loc) for part in ['sha1', 'asc']: local_file = '%s.%s' % (u_filename, part) remote_file = os.path.join(remote_dir, local_file) sftp.put(local_file, remote_file) sftp.put(remote_dir, os.path.join(remote_dir, u_filename)) rcode = True except BaseException: pass finally: if 'transport' in locals(): transport.close() return rcode def get_cf_files(path, queue): """Get rule files in a directory and put them in a queue""" for root, _, files in os.walk(os.path.abspath(path)): if not files: continue for filename in files: fullname = os.path.join(root, filename) if os.path.isfile(fullname) and fullname.endswith('.cf') or \ fullname.endswith('.post'): queue.put(fullname) def cleanup(dest, tardir, counterfile): """Remove existing rules""" thefiles = Queue() # dest directory files queue_files(dest, thefiles) # tar directory files queue_files(tardir, thefiles) while not thefiles.empty(): d_file = thefiles.get() info("Deleting file: %s" % d_file) os.unlink(d_file) if os.path.exists(counterfile): info("Deleting the counter file %s" % counterfile) os.unlink(counterfile) def check_required(config): """Validate the input""" if config.get('domain_key') is None: raise CfgError("The domain_key option is required") if config.get('remote_loc') is None: raise CfgError("The remote_location option is required") if config.get('gpg_keyid') is None: raise CfgError("The gpg_keyid option is required") def entry(config, delete_files=None): """Main function""" home_dir = config.get('home_dir', '/var/lib/sachannelupdate') dns_ver = config.get('spamassassin_version', '1.4.3') remote_loc = config.get('remote_location') rule_dir = os.path.join(home_dir, 'rules') dest = os.path.join(home_dir, 'deploy') tardir = os.path.join(home_dir, 'archives') counterfile = os.path.join(home_dir, 'db', 'counters') check_required(config) if delete_files: cleanup(dest, tardir, counterfile) return cffiles = Queue() get_cf_files(rule_dir, cffiles) if process(dest, cffiles): version = get_counter(counterfile) filename = '%s.tar.gz' % version path = os.path.join(tardir, filename) package(dest, tardir, version) sign(config, path) hash_file(path) if upload(config, remote_loc, path): if update_dns(config, str(version), dns_ver): create_file(counterfile, "%d" % version)
akissa/sachannelupdate
sachannelupdate/base.py
get_cf_files
python
def get_cf_files(path, queue): for root, _, files in os.walk(os.path.abspath(path)): if not files: continue for filename in files: fullname = os.path.join(root, filename) if os.path.isfile(fullname) and fullname.endswith('.cf') or \ fullname.endswith('.post'): queue.put(fullname)
Get rule files in a directory and put them in a queue
train
https://github.com/akissa/sachannelupdate/blob/a1c3c3d86b874f9c92c2407e2608963165d3ae98/sachannelupdate/base.py#L196-L205
null
# -*- coding: utf-8 -*- # vim: ai ts=4 sts=4 et sw=4 # sachannelupdate - Utility for pushing updates to Spamassassin update channels # Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Affero General Public License as published # by the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Affero General Public License for more details. # # You should have received a copy of the GNU Affero General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. """ sachannelupdate: Utility for pushing updates to Spamassassin update channels Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> """ import os import tarfile import datetime from Queue import Queue from hashlib import sha1 from datetime import datetime from gnupg import GPG from dns.exception import DNSException from dns import tsig, query, tsigkeyring, update from sachannelupdate.utils import info from sachannelupdate.exceptions import SaChannelUpdateConfigError \ as CfgError, SaChannelUpdateDNSError, SaChannelUpdateError from sachannelupdate.transports import get_sftp_conn, get_remote_path BLOCKSIZE = 65536 HASHTMPL = """%s %s\n""" def create_file(name, content): "Generic to write file" with open(name, 'w') as writefile: writefile.write(content) def getfiles(qfiles, dirname, names): """Get rule files in a directory""" for name in names: fullname = os.path.join(dirname, name) if os.path.isfile(fullname) and \ fullname.endswith('.cf') or \ fullname.endswith('.post'): qfiles.put(fullname) def deploy_file(source, dest): """Deploy a file""" date = datetime.utcnow().strftime('%Y-%m-%d') shandle = open(source) with open(dest, 'w') as handle: for line in shandle: if line == '# Updated: %date%\n': newline = '# Updated: %s\n' % date else: newline = line handle.write(newline) handle.flush() shandle.close() def package(dest, tardir, p_version): """Package files""" os.chdir(dest) p_filename = '%s.tar.gz' % p_version p_path = os.path.join(tardir, p_filename) tar = tarfile.open(p_path, mode='w:gz') for cf_file in os.listdir('.'): if os.path.isfile(cf_file): tar.add(cf_file) tar.close() def process(dest, rulefiles): """process rules""" deploy = False while not rulefiles.empty(): rulefile = rulefiles.get() base = os.path.basename(rulefile) dest = os.path.join(dest, base) if os.path.exists(dest): # check if older oldtime = os.stat(rulefile).st_mtime newtime = os.stat(dest).st_mtime if oldtime > newtime: deploy = True deploy_file(rulefile, dest) else: deploy = True deploy_file(rulefile, dest) return deploy def get_counter(counterfile): """Get the counter value""" try: version_num = open(counterfile).read() version_num = int(version_num) + 1 except (ValueError, IOError): version_num = 1 create_file(counterfile, "%d" % version_num) except BaseException as msg: raise SaChannelUpdateError(msg) return version_num def update_dns(config, record, sa_version): "Update the DNS record" try: domain = config.get('domain_name', 'sa.baruwa.com.') dns_key = config.get('domain_key') dns_ip = config.get('domain_ip', '127.0.0.1') keyring = tsigkeyring.from_text({domain: dns_key}) transaction = update.Update( domain, keyring=keyring, keyalgorithm=tsig.HMAC_SHA512) txtrecord = '%s.%s' % (sa_version, domain) transaction.replace(txtrecord, 120, 'txt', record) query.tcp(transaction, dns_ip) return True except DNSException, msg: raise SaChannelUpdateDNSError(msg) def sign(config, s_filename): """sign the package""" gpg_home = config.get('gpg_dir', '/var/lib/sachannelupdate/gnupg') gpg_pass = config.get('gpg_passphrase') gpg_keyid = config.get('gpg_keyid') gpg = GPG(gnupghome=gpg_home) try: plaintext = open(s_filename, 'rb') signature = gpg.sign_file( plaintext, keyid=gpg_keyid, passphrase=gpg_pass, detach=True) with open('%s.asc' % s_filename, 'wb') as handle: handle.write(str(signature)) finally: if 'plaintext' in locals(): plaintext.close() def hash_file(tar_filename): """hash the file""" hasher = sha1() with open(tar_filename, 'rb') as afile: buf = afile.read(BLOCKSIZE) while len(buf) > 0: hasher.update(buf) buf = afile.read(BLOCKSIZE) data = HASHTMPL % (hasher.hexdigest(), os.path.basename(tar_filename)) create_file('%s.sha1' % tar_filename, data) def upload(config, remote_loc, u_filename): """Upload the files""" rcode = False try: sftp, transport = get_sftp_conn(config) remote_dir = get_remote_path(remote_loc) for part in ['sha1', 'asc']: local_file = '%s.%s' % (u_filename, part) remote_file = os.path.join(remote_dir, local_file) sftp.put(local_file, remote_file) sftp.put(remote_dir, os.path.join(remote_dir, u_filename)) rcode = True except BaseException: pass finally: if 'transport' in locals(): transport.close() return rcode def queue_files(dirpath, queue): """Add files in a directory to a queue""" for root, _, files in os.walk(os.path.abspath(dirpath)): if not files: continue for filename in files: queue.put(os.path.join(root, filename)) def cleanup(dest, tardir, counterfile): """Remove existing rules""" thefiles = Queue() # dest directory files queue_files(dest, thefiles) # tar directory files queue_files(tardir, thefiles) while not thefiles.empty(): d_file = thefiles.get() info("Deleting file: %s" % d_file) os.unlink(d_file) if os.path.exists(counterfile): info("Deleting the counter file %s" % counterfile) os.unlink(counterfile) def check_required(config): """Validate the input""" if config.get('domain_key') is None: raise CfgError("The domain_key option is required") if config.get('remote_loc') is None: raise CfgError("The remote_location option is required") if config.get('gpg_keyid') is None: raise CfgError("The gpg_keyid option is required") def entry(config, delete_files=None): """Main function""" home_dir = config.get('home_dir', '/var/lib/sachannelupdate') dns_ver = config.get('spamassassin_version', '1.4.3') remote_loc = config.get('remote_location') rule_dir = os.path.join(home_dir, 'rules') dest = os.path.join(home_dir, 'deploy') tardir = os.path.join(home_dir, 'archives') counterfile = os.path.join(home_dir, 'db', 'counters') check_required(config) if delete_files: cleanup(dest, tardir, counterfile) return cffiles = Queue() get_cf_files(rule_dir, cffiles) if process(dest, cffiles): version = get_counter(counterfile) filename = '%s.tar.gz' % version path = os.path.join(tardir, filename) package(dest, tardir, version) sign(config, path) hash_file(path) if upload(config, remote_loc, path): if update_dns(config, str(version), dns_ver): create_file(counterfile, "%d" % version)
akissa/sachannelupdate
sachannelupdate/base.py
cleanup
python
def cleanup(dest, tardir, counterfile): thefiles = Queue() # dest directory files queue_files(dest, thefiles) # tar directory files queue_files(tardir, thefiles) while not thefiles.empty(): d_file = thefiles.get() info("Deleting file: %s" % d_file) os.unlink(d_file) if os.path.exists(counterfile): info("Deleting the counter file %s" % counterfile) os.unlink(counterfile)
Remove existing rules
train
https://github.com/akissa/sachannelupdate/blob/a1c3c3d86b874f9c92c2407e2608963165d3ae98/sachannelupdate/base.py#L208-L221
[ "def info(msg):\n \"\"\"print to stdout\"\"\"\n print(msg, file=sys.stdout)\n", "def queue_files(dirpath, queue):\n \"\"\"Add files in a directory to a queue\"\"\"\n for root, _, files in os.walk(os.path.abspath(dirpath)):\n if not files:\n continue\n for filename in files:\n queue.put(os.path.join(root, filename))\n" ]
# -*- coding: utf-8 -*- # vim: ai ts=4 sts=4 et sw=4 # sachannelupdate - Utility for pushing updates to Spamassassin update channels # Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Affero General Public License as published # by the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Affero General Public License for more details. # # You should have received a copy of the GNU Affero General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. """ sachannelupdate: Utility for pushing updates to Spamassassin update channels Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> """ import os import tarfile import datetime from Queue import Queue from hashlib import sha1 from datetime import datetime from gnupg import GPG from dns.exception import DNSException from dns import tsig, query, tsigkeyring, update from sachannelupdate.utils import info from sachannelupdate.exceptions import SaChannelUpdateConfigError \ as CfgError, SaChannelUpdateDNSError, SaChannelUpdateError from sachannelupdate.transports import get_sftp_conn, get_remote_path BLOCKSIZE = 65536 HASHTMPL = """%s %s\n""" def create_file(name, content): "Generic to write file" with open(name, 'w') as writefile: writefile.write(content) def getfiles(qfiles, dirname, names): """Get rule files in a directory""" for name in names: fullname = os.path.join(dirname, name) if os.path.isfile(fullname) and \ fullname.endswith('.cf') or \ fullname.endswith('.post'): qfiles.put(fullname) def deploy_file(source, dest): """Deploy a file""" date = datetime.utcnow().strftime('%Y-%m-%d') shandle = open(source) with open(dest, 'w') as handle: for line in shandle: if line == '# Updated: %date%\n': newline = '# Updated: %s\n' % date else: newline = line handle.write(newline) handle.flush() shandle.close() def package(dest, tardir, p_version): """Package files""" os.chdir(dest) p_filename = '%s.tar.gz' % p_version p_path = os.path.join(tardir, p_filename) tar = tarfile.open(p_path, mode='w:gz') for cf_file in os.listdir('.'): if os.path.isfile(cf_file): tar.add(cf_file) tar.close() def process(dest, rulefiles): """process rules""" deploy = False while not rulefiles.empty(): rulefile = rulefiles.get() base = os.path.basename(rulefile) dest = os.path.join(dest, base) if os.path.exists(dest): # check if older oldtime = os.stat(rulefile).st_mtime newtime = os.stat(dest).st_mtime if oldtime > newtime: deploy = True deploy_file(rulefile, dest) else: deploy = True deploy_file(rulefile, dest) return deploy def get_counter(counterfile): """Get the counter value""" try: version_num = open(counterfile).read() version_num = int(version_num) + 1 except (ValueError, IOError): version_num = 1 create_file(counterfile, "%d" % version_num) except BaseException as msg: raise SaChannelUpdateError(msg) return version_num def update_dns(config, record, sa_version): "Update the DNS record" try: domain = config.get('domain_name', 'sa.baruwa.com.') dns_key = config.get('domain_key') dns_ip = config.get('domain_ip', '127.0.0.1') keyring = tsigkeyring.from_text({domain: dns_key}) transaction = update.Update( domain, keyring=keyring, keyalgorithm=tsig.HMAC_SHA512) txtrecord = '%s.%s' % (sa_version, domain) transaction.replace(txtrecord, 120, 'txt', record) query.tcp(transaction, dns_ip) return True except DNSException, msg: raise SaChannelUpdateDNSError(msg) def sign(config, s_filename): """sign the package""" gpg_home = config.get('gpg_dir', '/var/lib/sachannelupdate/gnupg') gpg_pass = config.get('gpg_passphrase') gpg_keyid = config.get('gpg_keyid') gpg = GPG(gnupghome=gpg_home) try: plaintext = open(s_filename, 'rb') signature = gpg.sign_file( plaintext, keyid=gpg_keyid, passphrase=gpg_pass, detach=True) with open('%s.asc' % s_filename, 'wb') as handle: handle.write(str(signature)) finally: if 'plaintext' in locals(): plaintext.close() def hash_file(tar_filename): """hash the file""" hasher = sha1() with open(tar_filename, 'rb') as afile: buf = afile.read(BLOCKSIZE) while len(buf) > 0: hasher.update(buf) buf = afile.read(BLOCKSIZE) data = HASHTMPL % (hasher.hexdigest(), os.path.basename(tar_filename)) create_file('%s.sha1' % tar_filename, data) def upload(config, remote_loc, u_filename): """Upload the files""" rcode = False try: sftp, transport = get_sftp_conn(config) remote_dir = get_remote_path(remote_loc) for part in ['sha1', 'asc']: local_file = '%s.%s' % (u_filename, part) remote_file = os.path.join(remote_dir, local_file) sftp.put(local_file, remote_file) sftp.put(remote_dir, os.path.join(remote_dir, u_filename)) rcode = True except BaseException: pass finally: if 'transport' in locals(): transport.close() return rcode def queue_files(dirpath, queue): """Add files in a directory to a queue""" for root, _, files in os.walk(os.path.abspath(dirpath)): if not files: continue for filename in files: queue.put(os.path.join(root, filename)) def get_cf_files(path, queue): """Get rule files in a directory and put them in a queue""" for root, _, files in os.walk(os.path.abspath(path)): if not files: continue for filename in files: fullname = os.path.join(root, filename) if os.path.isfile(fullname) and fullname.endswith('.cf') or \ fullname.endswith('.post'): queue.put(fullname) def check_required(config): """Validate the input""" if config.get('domain_key') is None: raise CfgError("The domain_key option is required") if config.get('remote_loc') is None: raise CfgError("The remote_location option is required") if config.get('gpg_keyid') is None: raise CfgError("The gpg_keyid option is required") def entry(config, delete_files=None): """Main function""" home_dir = config.get('home_dir', '/var/lib/sachannelupdate') dns_ver = config.get('spamassassin_version', '1.4.3') remote_loc = config.get('remote_location') rule_dir = os.path.join(home_dir, 'rules') dest = os.path.join(home_dir, 'deploy') tardir = os.path.join(home_dir, 'archives') counterfile = os.path.join(home_dir, 'db', 'counters') check_required(config) if delete_files: cleanup(dest, tardir, counterfile) return cffiles = Queue() get_cf_files(rule_dir, cffiles) if process(dest, cffiles): version = get_counter(counterfile) filename = '%s.tar.gz' % version path = os.path.join(tardir, filename) package(dest, tardir, version) sign(config, path) hash_file(path) if upload(config, remote_loc, path): if update_dns(config, str(version), dns_ver): create_file(counterfile, "%d" % version)
akissa/sachannelupdate
sachannelupdate/base.py
check_required
python
def check_required(config): if config.get('domain_key') is None: raise CfgError("The domain_key option is required") if config.get('remote_loc') is None: raise CfgError("The remote_location option is required") if config.get('gpg_keyid') is None: raise CfgError("The gpg_keyid option is required")
Validate the input
train
https://github.com/akissa/sachannelupdate/blob/a1c3c3d86b874f9c92c2407e2608963165d3ae98/sachannelupdate/base.py#L224-L231
null
# -*- coding: utf-8 -*- # vim: ai ts=4 sts=4 et sw=4 # sachannelupdate - Utility for pushing updates to Spamassassin update channels # Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Affero General Public License as published # by the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Affero General Public License for more details. # # You should have received a copy of the GNU Affero General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. """ sachannelupdate: Utility for pushing updates to Spamassassin update channels Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> """ import os import tarfile import datetime from Queue import Queue from hashlib import sha1 from datetime import datetime from gnupg import GPG from dns.exception import DNSException from dns import tsig, query, tsigkeyring, update from sachannelupdate.utils import info from sachannelupdate.exceptions import SaChannelUpdateConfigError \ as CfgError, SaChannelUpdateDNSError, SaChannelUpdateError from sachannelupdate.transports import get_sftp_conn, get_remote_path BLOCKSIZE = 65536 HASHTMPL = """%s %s\n""" def create_file(name, content): "Generic to write file" with open(name, 'w') as writefile: writefile.write(content) def getfiles(qfiles, dirname, names): """Get rule files in a directory""" for name in names: fullname = os.path.join(dirname, name) if os.path.isfile(fullname) and \ fullname.endswith('.cf') or \ fullname.endswith('.post'): qfiles.put(fullname) def deploy_file(source, dest): """Deploy a file""" date = datetime.utcnow().strftime('%Y-%m-%d') shandle = open(source) with open(dest, 'w') as handle: for line in shandle: if line == '# Updated: %date%\n': newline = '# Updated: %s\n' % date else: newline = line handle.write(newline) handle.flush() shandle.close() def package(dest, tardir, p_version): """Package files""" os.chdir(dest) p_filename = '%s.tar.gz' % p_version p_path = os.path.join(tardir, p_filename) tar = tarfile.open(p_path, mode='w:gz') for cf_file in os.listdir('.'): if os.path.isfile(cf_file): tar.add(cf_file) tar.close() def process(dest, rulefiles): """process rules""" deploy = False while not rulefiles.empty(): rulefile = rulefiles.get() base = os.path.basename(rulefile) dest = os.path.join(dest, base) if os.path.exists(dest): # check if older oldtime = os.stat(rulefile).st_mtime newtime = os.stat(dest).st_mtime if oldtime > newtime: deploy = True deploy_file(rulefile, dest) else: deploy = True deploy_file(rulefile, dest) return deploy def get_counter(counterfile): """Get the counter value""" try: version_num = open(counterfile).read() version_num = int(version_num) + 1 except (ValueError, IOError): version_num = 1 create_file(counterfile, "%d" % version_num) except BaseException as msg: raise SaChannelUpdateError(msg) return version_num def update_dns(config, record, sa_version): "Update the DNS record" try: domain = config.get('domain_name', 'sa.baruwa.com.') dns_key = config.get('domain_key') dns_ip = config.get('domain_ip', '127.0.0.1') keyring = tsigkeyring.from_text({domain: dns_key}) transaction = update.Update( domain, keyring=keyring, keyalgorithm=tsig.HMAC_SHA512) txtrecord = '%s.%s' % (sa_version, domain) transaction.replace(txtrecord, 120, 'txt', record) query.tcp(transaction, dns_ip) return True except DNSException, msg: raise SaChannelUpdateDNSError(msg) def sign(config, s_filename): """sign the package""" gpg_home = config.get('gpg_dir', '/var/lib/sachannelupdate/gnupg') gpg_pass = config.get('gpg_passphrase') gpg_keyid = config.get('gpg_keyid') gpg = GPG(gnupghome=gpg_home) try: plaintext = open(s_filename, 'rb') signature = gpg.sign_file( plaintext, keyid=gpg_keyid, passphrase=gpg_pass, detach=True) with open('%s.asc' % s_filename, 'wb') as handle: handle.write(str(signature)) finally: if 'plaintext' in locals(): plaintext.close() def hash_file(tar_filename): """hash the file""" hasher = sha1() with open(tar_filename, 'rb') as afile: buf = afile.read(BLOCKSIZE) while len(buf) > 0: hasher.update(buf) buf = afile.read(BLOCKSIZE) data = HASHTMPL % (hasher.hexdigest(), os.path.basename(tar_filename)) create_file('%s.sha1' % tar_filename, data) def upload(config, remote_loc, u_filename): """Upload the files""" rcode = False try: sftp, transport = get_sftp_conn(config) remote_dir = get_remote_path(remote_loc) for part in ['sha1', 'asc']: local_file = '%s.%s' % (u_filename, part) remote_file = os.path.join(remote_dir, local_file) sftp.put(local_file, remote_file) sftp.put(remote_dir, os.path.join(remote_dir, u_filename)) rcode = True except BaseException: pass finally: if 'transport' in locals(): transport.close() return rcode def queue_files(dirpath, queue): """Add files in a directory to a queue""" for root, _, files in os.walk(os.path.abspath(dirpath)): if not files: continue for filename in files: queue.put(os.path.join(root, filename)) def get_cf_files(path, queue): """Get rule files in a directory and put them in a queue""" for root, _, files in os.walk(os.path.abspath(path)): if not files: continue for filename in files: fullname = os.path.join(root, filename) if os.path.isfile(fullname) and fullname.endswith('.cf') or \ fullname.endswith('.post'): queue.put(fullname) def cleanup(dest, tardir, counterfile): """Remove existing rules""" thefiles = Queue() # dest directory files queue_files(dest, thefiles) # tar directory files queue_files(tardir, thefiles) while not thefiles.empty(): d_file = thefiles.get() info("Deleting file: %s" % d_file) os.unlink(d_file) if os.path.exists(counterfile): info("Deleting the counter file %s" % counterfile) os.unlink(counterfile) def entry(config, delete_files=None): """Main function""" home_dir = config.get('home_dir', '/var/lib/sachannelupdate') dns_ver = config.get('spamassassin_version', '1.4.3') remote_loc = config.get('remote_location') rule_dir = os.path.join(home_dir, 'rules') dest = os.path.join(home_dir, 'deploy') tardir = os.path.join(home_dir, 'archives') counterfile = os.path.join(home_dir, 'db', 'counters') check_required(config) if delete_files: cleanup(dest, tardir, counterfile) return cffiles = Queue() get_cf_files(rule_dir, cffiles) if process(dest, cffiles): version = get_counter(counterfile) filename = '%s.tar.gz' % version path = os.path.join(tardir, filename) package(dest, tardir, version) sign(config, path) hash_file(path) if upload(config, remote_loc, path): if update_dns(config, str(version), dns_ver): create_file(counterfile, "%d" % version)
akissa/sachannelupdate
sachannelupdate/base.py
entry
python
def entry(config, delete_files=None): home_dir = config.get('home_dir', '/var/lib/sachannelupdate') dns_ver = config.get('spamassassin_version', '1.4.3') remote_loc = config.get('remote_location') rule_dir = os.path.join(home_dir, 'rules') dest = os.path.join(home_dir, 'deploy') tardir = os.path.join(home_dir, 'archives') counterfile = os.path.join(home_dir, 'db', 'counters') check_required(config) if delete_files: cleanup(dest, tardir, counterfile) return cffiles = Queue() get_cf_files(rule_dir, cffiles) if process(dest, cffiles): version = get_counter(counterfile) filename = '%s.tar.gz' % version path = os.path.join(tardir, filename) package(dest, tardir, version) sign(config, path) hash_file(path) if upload(config, remote_loc, path): if update_dns(config, str(version), dns_ver): create_file(counterfile, "%d" % version)
Main function
train
https://github.com/akissa/sachannelupdate/blob/a1c3c3d86b874f9c92c2407e2608963165d3ae98/sachannelupdate/base.py#L234-L262
[ "def process(dest, rulefiles):\n \"\"\"process rules\"\"\"\n deploy = False\n while not rulefiles.empty():\n rulefile = rulefiles.get()\n base = os.path.basename(rulefile)\n dest = os.path.join(dest, base)\n if os.path.exists(dest):\n # check if older\n oldtime = os.stat(rulefile).st_mtime\n newtime = os.stat(dest).st_mtime\n if oldtime > newtime:\n deploy = True\n deploy_file(rulefile, dest)\n else:\n deploy = True\n deploy_file(rulefile, dest)\n return deploy\n", "def cleanup(dest, tardir, counterfile):\n \"\"\"Remove existing rules\"\"\"\n thefiles = Queue()\n # dest directory files\n queue_files(dest, thefiles)\n # tar directory files\n queue_files(tardir, thefiles)\n while not thefiles.empty():\n d_file = thefiles.get()\n info(\"Deleting file: %s\" % d_file)\n os.unlink(d_file)\n if os.path.exists(counterfile):\n info(\"Deleting the counter file %s\" % counterfile)\n os.unlink(counterfile)\n", "def upload(config, remote_loc, u_filename):\n \"\"\"Upload the files\"\"\"\n rcode = False\n try:\n sftp, transport = get_sftp_conn(config)\n remote_dir = get_remote_path(remote_loc)\n for part in ['sha1', 'asc']:\n local_file = '%s.%s' % (u_filename, part)\n remote_file = os.path.join(remote_dir, local_file)\n sftp.put(local_file, remote_file)\n sftp.put(remote_dir, os.path.join(remote_dir, u_filename))\n rcode = True\n except BaseException:\n pass\n finally:\n if 'transport' in locals():\n transport.close()\n return rcode\n", "def package(dest, tardir, p_version):\n \"\"\"Package files\"\"\"\n os.chdir(dest)\n p_filename = '%s.tar.gz' % p_version\n p_path = os.path.join(tardir, p_filename)\n tar = tarfile.open(p_path, mode='w:gz')\n for cf_file in os.listdir('.'):\n if os.path.isfile(cf_file):\n tar.add(cf_file)\n tar.close()\n", "def sign(config, s_filename):\n \"\"\"sign the package\"\"\"\n gpg_home = config.get('gpg_dir', '/var/lib/sachannelupdate/gnupg')\n gpg_pass = config.get('gpg_passphrase')\n gpg_keyid = config.get('gpg_keyid')\n gpg = GPG(gnupghome=gpg_home)\n try:\n plaintext = open(s_filename, 'rb')\n signature = gpg.sign_file(\n plaintext, keyid=gpg_keyid, passphrase=gpg_pass, detach=True)\n with open('%s.asc' % s_filename, 'wb') as handle:\n handle.write(str(signature))\n finally:\n if 'plaintext' in locals():\n plaintext.close()\n", "def create_file(name, content):\n \"Generic to write file\"\n with open(name, 'w') as writefile:\n writefile.write(content)\n", "def get_counter(counterfile):\n \"\"\"Get the counter value\"\"\"\n try:\n version_num = open(counterfile).read()\n version_num = int(version_num) + 1\n except (ValueError, IOError):\n version_num = 1\n create_file(counterfile, \"%d\" % version_num)\n except BaseException as msg:\n raise SaChannelUpdateError(msg)\n return version_num\n", "def update_dns(config, record, sa_version):\n \"Update the DNS record\"\n try:\n domain = config.get('domain_name', 'sa.baruwa.com.')\n dns_key = config.get('domain_key')\n dns_ip = config.get('domain_ip', '127.0.0.1')\n keyring = tsigkeyring.from_text({domain: dns_key})\n transaction = update.Update(\n domain,\n keyring=keyring,\n keyalgorithm=tsig.HMAC_SHA512)\n txtrecord = '%s.%s' % (sa_version, domain)\n transaction.replace(txtrecord, 120, 'txt', record)\n query.tcp(transaction, dns_ip)\n return True\n except DNSException, msg:\n raise SaChannelUpdateDNSError(msg)\n", "def hash_file(tar_filename):\n \"\"\"hash the file\"\"\"\n hasher = sha1()\n with open(tar_filename, 'rb') as afile:\n buf = afile.read(BLOCKSIZE)\n while len(buf) > 0:\n hasher.update(buf)\n buf = afile.read(BLOCKSIZE)\n data = HASHTMPL % (hasher.hexdigest(), os.path.basename(tar_filename))\n create_file('%s.sha1' % tar_filename, data)\n", "def get_cf_files(path, queue):\n \"\"\"Get rule files in a directory and put them in a queue\"\"\"\n for root, _, files in os.walk(os.path.abspath(path)):\n if not files:\n continue\n for filename in files:\n fullname = os.path.join(root, filename)\n if os.path.isfile(fullname) and fullname.endswith('.cf') or \\\n fullname.endswith('.post'):\n queue.put(fullname)\n", "def check_required(config):\n \"\"\"Validate the input\"\"\"\n if config.get('domain_key') is None:\n raise CfgError(\"The domain_key option is required\")\n if config.get('remote_loc') is None:\n raise CfgError(\"The remote_location option is required\")\n if config.get('gpg_keyid') is None:\n raise CfgError(\"The gpg_keyid option is required\")\n" ]
# -*- coding: utf-8 -*- # vim: ai ts=4 sts=4 et sw=4 # sachannelupdate - Utility for pushing updates to Spamassassin update channels # Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Affero General Public License as published # by the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Affero General Public License for more details. # # You should have received a copy of the GNU Affero General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. """ sachannelupdate: Utility for pushing updates to Spamassassin update channels Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> """ import os import tarfile import datetime from Queue import Queue from hashlib import sha1 from datetime import datetime from gnupg import GPG from dns.exception import DNSException from dns import tsig, query, tsigkeyring, update from sachannelupdate.utils import info from sachannelupdate.exceptions import SaChannelUpdateConfigError \ as CfgError, SaChannelUpdateDNSError, SaChannelUpdateError from sachannelupdate.transports import get_sftp_conn, get_remote_path BLOCKSIZE = 65536 HASHTMPL = """%s %s\n""" def create_file(name, content): "Generic to write file" with open(name, 'w') as writefile: writefile.write(content) def getfiles(qfiles, dirname, names): """Get rule files in a directory""" for name in names: fullname = os.path.join(dirname, name) if os.path.isfile(fullname) and \ fullname.endswith('.cf') or \ fullname.endswith('.post'): qfiles.put(fullname) def deploy_file(source, dest): """Deploy a file""" date = datetime.utcnow().strftime('%Y-%m-%d') shandle = open(source) with open(dest, 'w') as handle: for line in shandle: if line == '# Updated: %date%\n': newline = '# Updated: %s\n' % date else: newline = line handle.write(newline) handle.flush() shandle.close() def package(dest, tardir, p_version): """Package files""" os.chdir(dest) p_filename = '%s.tar.gz' % p_version p_path = os.path.join(tardir, p_filename) tar = tarfile.open(p_path, mode='w:gz') for cf_file in os.listdir('.'): if os.path.isfile(cf_file): tar.add(cf_file) tar.close() def process(dest, rulefiles): """process rules""" deploy = False while not rulefiles.empty(): rulefile = rulefiles.get() base = os.path.basename(rulefile) dest = os.path.join(dest, base) if os.path.exists(dest): # check if older oldtime = os.stat(rulefile).st_mtime newtime = os.stat(dest).st_mtime if oldtime > newtime: deploy = True deploy_file(rulefile, dest) else: deploy = True deploy_file(rulefile, dest) return deploy def get_counter(counterfile): """Get the counter value""" try: version_num = open(counterfile).read() version_num = int(version_num) + 1 except (ValueError, IOError): version_num = 1 create_file(counterfile, "%d" % version_num) except BaseException as msg: raise SaChannelUpdateError(msg) return version_num def update_dns(config, record, sa_version): "Update the DNS record" try: domain = config.get('domain_name', 'sa.baruwa.com.') dns_key = config.get('domain_key') dns_ip = config.get('domain_ip', '127.0.0.1') keyring = tsigkeyring.from_text({domain: dns_key}) transaction = update.Update( domain, keyring=keyring, keyalgorithm=tsig.HMAC_SHA512) txtrecord = '%s.%s' % (sa_version, domain) transaction.replace(txtrecord, 120, 'txt', record) query.tcp(transaction, dns_ip) return True except DNSException, msg: raise SaChannelUpdateDNSError(msg) def sign(config, s_filename): """sign the package""" gpg_home = config.get('gpg_dir', '/var/lib/sachannelupdate/gnupg') gpg_pass = config.get('gpg_passphrase') gpg_keyid = config.get('gpg_keyid') gpg = GPG(gnupghome=gpg_home) try: plaintext = open(s_filename, 'rb') signature = gpg.sign_file( plaintext, keyid=gpg_keyid, passphrase=gpg_pass, detach=True) with open('%s.asc' % s_filename, 'wb') as handle: handle.write(str(signature)) finally: if 'plaintext' in locals(): plaintext.close() def hash_file(tar_filename): """hash the file""" hasher = sha1() with open(tar_filename, 'rb') as afile: buf = afile.read(BLOCKSIZE) while len(buf) > 0: hasher.update(buf) buf = afile.read(BLOCKSIZE) data = HASHTMPL % (hasher.hexdigest(), os.path.basename(tar_filename)) create_file('%s.sha1' % tar_filename, data) def upload(config, remote_loc, u_filename): """Upload the files""" rcode = False try: sftp, transport = get_sftp_conn(config) remote_dir = get_remote_path(remote_loc) for part in ['sha1', 'asc']: local_file = '%s.%s' % (u_filename, part) remote_file = os.path.join(remote_dir, local_file) sftp.put(local_file, remote_file) sftp.put(remote_dir, os.path.join(remote_dir, u_filename)) rcode = True except BaseException: pass finally: if 'transport' in locals(): transport.close() return rcode def queue_files(dirpath, queue): """Add files in a directory to a queue""" for root, _, files in os.walk(os.path.abspath(dirpath)): if not files: continue for filename in files: queue.put(os.path.join(root, filename)) def get_cf_files(path, queue): """Get rule files in a directory and put them in a queue""" for root, _, files in os.walk(os.path.abspath(path)): if not files: continue for filename in files: fullname = os.path.join(root, filename) if os.path.isfile(fullname) and fullname.endswith('.cf') or \ fullname.endswith('.post'): queue.put(fullname) def cleanup(dest, tardir, counterfile): """Remove existing rules""" thefiles = Queue() # dest directory files queue_files(dest, thefiles) # tar directory files queue_files(tardir, thefiles) while not thefiles.empty(): d_file = thefiles.get() info("Deleting file: %s" % d_file) os.unlink(d_file) if os.path.exists(counterfile): info("Deleting the counter file %s" % counterfile) os.unlink(counterfile) def check_required(config): """Validate the input""" if config.get('domain_key') is None: raise CfgError("The domain_key option is required") if config.get('remote_loc') is None: raise CfgError("The remote_location option is required") if config.get('gpg_keyid') is None: raise CfgError("The gpg_keyid option is required")
akissa/sachannelupdate
sachannelupdate/cli.py
main
python
def main(): parser = OptionParser() parser.add_option( '-c', '--config', help='configuration file', dest='filename', type='str', default='/etc/sachannelupdate/sachannelupdate.ini') parser.add_option( '-d', '--delete', help='Deletes existing rules', dest='cleanup', action="store_true", default=False,) options, _ = parser.parse_args() if not os.path.isfile(options.filename): raise SaChannelUpdateConfigError( "The configuration file: %s does not exist" % options.filename) config = ConfigParser() config.read(options.filename) try: # pylint: disable=protected-access entry(config._sections['settings'], options.cleanup) except BaseException as msg: error(msg)
Main function
train
https://github.com/akissa/sachannelupdate/blob/a1c3c3d86b874f9c92c2407e2608963165d3ae98/sachannelupdate/cli.py#L29-L55
[ "def error(msg):\n \"\"\"print to stderr\"\"\"\n print(msg, file=sys.stderr)\n", "def entry(config, delete_files=None):\n \"\"\"Main function\"\"\"\n home_dir = config.get('home_dir', '/var/lib/sachannelupdate')\n dns_ver = config.get('spamassassin_version', '1.4.3')\n remote_loc = config.get('remote_location')\n rule_dir = os.path.join(home_dir, 'rules')\n dest = os.path.join(home_dir, 'deploy')\n tardir = os.path.join(home_dir, 'archives')\n counterfile = os.path.join(home_dir, 'db', 'counters')\n\n check_required(config)\n\n if delete_files:\n cleanup(dest, tardir, counterfile)\n return\n\n cffiles = Queue()\n get_cf_files(rule_dir, cffiles)\n\n if process(dest, cffiles):\n version = get_counter(counterfile)\n filename = '%s.tar.gz' % version\n path = os.path.join(tardir, filename)\n package(dest, tardir, version)\n sign(config, path)\n hash_file(path)\n if upload(config, remote_loc, path):\n if update_dns(config, str(version), dns_ver):\n create_file(counterfile, \"%d\" % version)\n" ]
# -*- coding: utf-8 -*- # vim: ai ts=4 sts=4 et sw=4 # sachannelupdate - Utility for pushing updates to Spamassassin update channels # Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Affero General Public License as published # by the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Affero General Public License for more details. # # You should have received a copy of the GNU Affero General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. """ sachannelupdate: CLI functions """ import os from optparse import OptionParser from ConfigParser import ConfigParser from sachannelupdate import entry, error, SaChannelUpdateConfigError
akissa/sachannelupdate
sachannelupdate/transports.py
get_key_files
python
def get_key_files(kfiles, dirname, names): for name in names: fullname = os.path.join(dirname, name) if os.path.isfile(fullname) and \ fullname.endswith('_rsa') or \ fullname.endswith('_dsa'): kfiles.put(fullname)
Return key files
train
https://github.com/akissa/sachannelupdate/blob/a1c3c3d86b874f9c92c2407e2608963165d3ae98/sachannelupdate/transports.py#L35-L42
null
# -*- coding: utf-8 -*- # vim: ai ts=4 sts=4 et sw=4 # sachannelupdate - Utility for pushing updates to Spamassassin update channels # Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Affero General Public License as published # by the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Affero General Public License for more details. # # You should have received a copy of the GNU Affero General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. """ sachannelupdate: Transports """ import os from Queue import Queue from pwd import getpwnam from getpass import getuser from urlparse import urlparse from paramiko.util import load_host_keys from paramiko import Transport, SFTPClient, PKey, PasswordRequiredException, \ SSHException from sachannelupdate.exceptions import SaChannelUpdateTransportError def get_ssh_keys(sshdir): """Get SSH keys""" keys = Queue() for root, _, files in os.walk(os.path.abspath(sshdir)): if not files: continue for filename in files: fullname = os.path.join(root, filename) if (os.path.isfile(fullname) and fullname.endswith('_rsa') or fullname.endswith('_dsa')): keys.put(fullname) return keys def get_remote_path(remote_location): """Get the remote path from the remote location""" parts = urlparse(remote_location) return parts.path def get_ssh_dir(config, username): """Get the users ssh dir""" sshdir = config.get('ssh_config_dir') if not sshdir: sshdir = os.path.expanduser('~/.ssh') if not os.path.isdir(sshdir): pwentry = getpwnam(username) sshdir = os.path.join(pwentry.pw_dir, '.ssh') if not os.path.isdir(sshdir): sshdir = None return sshdir def get_local_user(username): """Get the local username""" try: _ = getpwnam(username) luser = username except KeyError: luser = getuser() return luser def get_host_keys(hostname, sshdir): """get host key""" hostkey = None try: host_keys = load_host_keys(os.path.join(sshdir, 'known_hosts')) except IOError: host_keys = {} if hostname in host_keys: hostkeytype = host_keys[hostname].keys()[0] hostkey = host_keys[hostname][hostkeytype] return hostkey def get_sftp_conn(config): """Make a SFTP connection, returns sftp client and connection objects""" remote = config.get('remote_location') parts = urlparse(remote) if ':' in parts.netloc: hostname, port = parts.netloc.split(':') else: hostname = parts.netloc port = 22 port = int(port) username = config.get('remote_username') or getuser() luser = get_local_user(username) sshdir = get_ssh_dir(config, luser) hostkey = get_host_keys(hostname, sshdir) try: sftp = None keys = get_ssh_keys(sshdir) transport = Transport((hostname, port)) while not keys.empty(): try: key = PKey.from_private_key_file(keys.get()) transport.connect( hostkey=hostkey, username=username, password=None, pkey=key) sftp = SFTPClient.from_transport(transport) break except (PasswordRequiredException, SSHException): pass if sftp is None: raise SaChannelUpdateTransportError("SFTP connection failed") return sftp, transport except BaseException as msg: raise SaChannelUpdateTransportError(msg)
akissa/sachannelupdate
sachannelupdate/transports.py
get_ssh_keys
python
def get_ssh_keys(sshdir): keys = Queue() for root, _, files in os.walk(os.path.abspath(sshdir)): if not files: continue for filename in files: fullname = os.path.join(root, filename) if (os.path.isfile(fullname) and fullname.endswith('_rsa') or fullname.endswith('_dsa')): keys.put(fullname) return keys
Get SSH keys
train
https://github.com/akissa/sachannelupdate/blob/a1c3c3d86b874f9c92c2407e2608963165d3ae98/sachannelupdate/transports.py#L45-L56
null
# -*- coding: utf-8 -*- # vim: ai ts=4 sts=4 et sw=4 # sachannelupdate - Utility for pushing updates to Spamassassin update channels # Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Affero General Public License as published # by the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Affero General Public License for more details. # # You should have received a copy of the GNU Affero General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. """ sachannelupdate: Transports """ import os from Queue import Queue from pwd import getpwnam from getpass import getuser from urlparse import urlparse from paramiko.util import load_host_keys from paramiko import Transport, SFTPClient, PKey, PasswordRequiredException, \ SSHException from sachannelupdate.exceptions import SaChannelUpdateTransportError def get_key_files(kfiles, dirname, names): """Return key files""" for name in names: fullname = os.path.join(dirname, name) if os.path.isfile(fullname) and \ fullname.endswith('_rsa') or \ fullname.endswith('_dsa'): kfiles.put(fullname) def get_remote_path(remote_location): """Get the remote path from the remote location""" parts = urlparse(remote_location) return parts.path def get_ssh_dir(config, username): """Get the users ssh dir""" sshdir = config.get('ssh_config_dir') if not sshdir: sshdir = os.path.expanduser('~/.ssh') if not os.path.isdir(sshdir): pwentry = getpwnam(username) sshdir = os.path.join(pwentry.pw_dir, '.ssh') if not os.path.isdir(sshdir): sshdir = None return sshdir def get_local_user(username): """Get the local username""" try: _ = getpwnam(username) luser = username except KeyError: luser = getuser() return luser def get_host_keys(hostname, sshdir): """get host key""" hostkey = None try: host_keys = load_host_keys(os.path.join(sshdir, 'known_hosts')) except IOError: host_keys = {} if hostname in host_keys: hostkeytype = host_keys[hostname].keys()[0] hostkey = host_keys[hostname][hostkeytype] return hostkey def get_sftp_conn(config): """Make a SFTP connection, returns sftp client and connection objects""" remote = config.get('remote_location') parts = urlparse(remote) if ':' in parts.netloc: hostname, port = parts.netloc.split(':') else: hostname = parts.netloc port = 22 port = int(port) username = config.get('remote_username') or getuser() luser = get_local_user(username) sshdir = get_ssh_dir(config, luser) hostkey = get_host_keys(hostname, sshdir) try: sftp = None keys = get_ssh_keys(sshdir) transport = Transport((hostname, port)) while not keys.empty(): try: key = PKey.from_private_key_file(keys.get()) transport.connect( hostkey=hostkey, username=username, password=None, pkey=key) sftp = SFTPClient.from_transport(transport) break except (PasswordRequiredException, SSHException): pass if sftp is None: raise SaChannelUpdateTransportError("SFTP connection failed") return sftp, transport except BaseException as msg: raise SaChannelUpdateTransportError(msg)
akissa/sachannelupdate
sachannelupdate/transports.py
get_ssh_dir
python
def get_ssh_dir(config, username): sshdir = config.get('ssh_config_dir') if not sshdir: sshdir = os.path.expanduser('~/.ssh') if not os.path.isdir(sshdir): pwentry = getpwnam(username) sshdir = os.path.join(pwentry.pw_dir, '.ssh') if not os.path.isdir(sshdir): sshdir = None return sshdir
Get the users ssh dir
train
https://github.com/akissa/sachannelupdate/blob/a1c3c3d86b874f9c92c2407e2608963165d3ae98/sachannelupdate/transports.py#L65-L75
null
# -*- coding: utf-8 -*- # vim: ai ts=4 sts=4 et sw=4 # sachannelupdate - Utility for pushing updates to Spamassassin update channels # Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Affero General Public License as published # by the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Affero General Public License for more details. # # You should have received a copy of the GNU Affero General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. """ sachannelupdate: Transports """ import os from Queue import Queue from pwd import getpwnam from getpass import getuser from urlparse import urlparse from paramiko.util import load_host_keys from paramiko import Transport, SFTPClient, PKey, PasswordRequiredException, \ SSHException from sachannelupdate.exceptions import SaChannelUpdateTransportError def get_key_files(kfiles, dirname, names): """Return key files""" for name in names: fullname = os.path.join(dirname, name) if os.path.isfile(fullname) and \ fullname.endswith('_rsa') or \ fullname.endswith('_dsa'): kfiles.put(fullname) def get_ssh_keys(sshdir): """Get SSH keys""" keys = Queue() for root, _, files in os.walk(os.path.abspath(sshdir)): if not files: continue for filename in files: fullname = os.path.join(root, filename) if (os.path.isfile(fullname) and fullname.endswith('_rsa') or fullname.endswith('_dsa')): keys.put(fullname) return keys def get_remote_path(remote_location): """Get the remote path from the remote location""" parts = urlparse(remote_location) return parts.path def get_local_user(username): """Get the local username""" try: _ = getpwnam(username) luser = username except KeyError: luser = getuser() return luser def get_host_keys(hostname, sshdir): """get host key""" hostkey = None try: host_keys = load_host_keys(os.path.join(sshdir, 'known_hosts')) except IOError: host_keys = {} if hostname in host_keys: hostkeytype = host_keys[hostname].keys()[0] hostkey = host_keys[hostname][hostkeytype] return hostkey def get_sftp_conn(config): """Make a SFTP connection, returns sftp client and connection objects""" remote = config.get('remote_location') parts = urlparse(remote) if ':' in parts.netloc: hostname, port = parts.netloc.split(':') else: hostname = parts.netloc port = 22 port = int(port) username = config.get('remote_username') or getuser() luser = get_local_user(username) sshdir = get_ssh_dir(config, luser) hostkey = get_host_keys(hostname, sshdir) try: sftp = None keys = get_ssh_keys(sshdir) transport = Transport((hostname, port)) while not keys.empty(): try: key = PKey.from_private_key_file(keys.get()) transport.connect( hostkey=hostkey, username=username, password=None, pkey=key) sftp = SFTPClient.from_transport(transport) break except (PasswordRequiredException, SSHException): pass if sftp is None: raise SaChannelUpdateTransportError("SFTP connection failed") return sftp, transport except BaseException as msg: raise SaChannelUpdateTransportError(msg)
akissa/sachannelupdate
sachannelupdate/transports.py
get_local_user
python
def get_local_user(username): try: _ = getpwnam(username) luser = username except KeyError: luser = getuser() return luser
Get the local username
train
https://github.com/akissa/sachannelupdate/blob/a1c3c3d86b874f9c92c2407e2608963165d3ae98/sachannelupdate/transports.py#L78-L85
null
# -*- coding: utf-8 -*- # vim: ai ts=4 sts=4 et sw=4 # sachannelupdate - Utility for pushing updates to Spamassassin update channels # Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Affero General Public License as published # by the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Affero General Public License for more details. # # You should have received a copy of the GNU Affero General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. """ sachannelupdate: Transports """ import os from Queue import Queue from pwd import getpwnam from getpass import getuser from urlparse import urlparse from paramiko.util import load_host_keys from paramiko import Transport, SFTPClient, PKey, PasswordRequiredException, \ SSHException from sachannelupdate.exceptions import SaChannelUpdateTransportError def get_key_files(kfiles, dirname, names): """Return key files""" for name in names: fullname = os.path.join(dirname, name) if os.path.isfile(fullname) and \ fullname.endswith('_rsa') or \ fullname.endswith('_dsa'): kfiles.put(fullname) def get_ssh_keys(sshdir): """Get SSH keys""" keys = Queue() for root, _, files in os.walk(os.path.abspath(sshdir)): if not files: continue for filename in files: fullname = os.path.join(root, filename) if (os.path.isfile(fullname) and fullname.endswith('_rsa') or fullname.endswith('_dsa')): keys.put(fullname) return keys def get_remote_path(remote_location): """Get the remote path from the remote location""" parts = urlparse(remote_location) return parts.path def get_ssh_dir(config, username): """Get the users ssh dir""" sshdir = config.get('ssh_config_dir') if not sshdir: sshdir = os.path.expanduser('~/.ssh') if not os.path.isdir(sshdir): pwentry = getpwnam(username) sshdir = os.path.join(pwentry.pw_dir, '.ssh') if not os.path.isdir(sshdir): sshdir = None return sshdir def get_host_keys(hostname, sshdir): """get host key""" hostkey = None try: host_keys = load_host_keys(os.path.join(sshdir, 'known_hosts')) except IOError: host_keys = {} if hostname in host_keys: hostkeytype = host_keys[hostname].keys()[0] hostkey = host_keys[hostname][hostkeytype] return hostkey def get_sftp_conn(config): """Make a SFTP connection, returns sftp client and connection objects""" remote = config.get('remote_location') parts = urlparse(remote) if ':' in parts.netloc: hostname, port = parts.netloc.split(':') else: hostname = parts.netloc port = 22 port = int(port) username = config.get('remote_username') or getuser() luser = get_local_user(username) sshdir = get_ssh_dir(config, luser) hostkey = get_host_keys(hostname, sshdir) try: sftp = None keys = get_ssh_keys(sshdir) transport = Transport((hostname, port)) while not keys.empty(): try: key = PKey.from_private_key_file(keys.get()) transport.connect( hostkey=hostkey, username=username, password=None, pkey=key) sftp = SFTPClient.from_transport(transport) break except (PasswordRequiredException, SSHException): pass if sftp is None: raise SaChannelUpdateTransportError("SFTP connection failed") return sftp, transport except BaseException as msg: raise SaChannelUpdateTransportError(msg)
akissa/sachannelupdate
sachannelupdate/transports.py
get_host_keys
python
def get_host_keys(hostname, sshdir): hostkey = None try: host_keys = load_host_keys(os.path.join(sshdir, 'known_hosts')) except IOError: host_keys = {} if hostname in host_keys: hostkeytype = host_keys[hostname].keys()[0] hostkey = host_keys[hostname][hostkeytype] return hostkey
get host key
train
https://github.com/akissa/sachannelupdate/blob/a1c3c3d86b874f9c92c2407e2608963165d3ae98/sachannelupdate/transports.py#L88-L101
null
# -*- coding: utf-8 -*- # vim: ai ts=4 sts=4 et sw=4 # sachannelupdate - Utility for pushing updates to Spamassassin update channels # Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Affero General Public License as published # by the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Affero General Public License for more details. # # You should have received a copy of the GNU Affero General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. """ sachannelupdate: Transports """ import os from Queue import Queue from pwd import getpwnam from getpass import getuser from urlparse import urlparse from paramiko.util import load_host_keys from paramiko import Transport, SFTPClient, PKey, PasswordRequiredException, \ SSHException from sachannelupdate.exceptions import SaChannelUpdateTransportError def get_key_files(kfiles, dirname, names): """Return key files""" for name in names: fullname = os.path.join(dirname, name) if os.path.isfile(fullname) and \ fullname.endswith('_rsa') or \ fullname.endswith('_dsa'): kfiles.put(fullname) def get_ssh_keys(sshdir): """Get SSH keys""" keys = Queue() for root, _, files in os.walk(os.path.abspath(sshdir)): if not files: continue for filename in files: fullname = os.path.join(root, filename) if (os.path.isfile(fullname) and fullname.endswith('_rsa') or fullname.endswith('_dsa')): keys.put(fullname) return keys def get_remote_path(remote_location): """Get the remote path from the remote location""" parts = urlparse(remote_location) return parts.path def get_ssh_dir(config, username): """Get the users ssh dir""" sshdir = config.get('ssh_config_dir') if not sshdir: sshdir = os.path.expanduser('~/.ssh') if not os.path.isdir(sshdir): pwentry = getpwnam(username) sshdir = os.path.join(pwentry.pw_dir, '.ssh') if not os.path.isdir(sshdir): sshdir = None return sshdir def get_local_user(username): """Get the local username""" try: _ = getpwnam(username) luser = username except KeyError: luser = getuser() return luser def get_sftp_conn(config): """Make a SFTP connection, returns sftp client and connection objects""" remote = config.get('remote_location') parts = urlparse(remote) if ':' in parts.netloc: hostname, port = parts.netloc.split(':') else: hostname = parts.netloc port = 22 port = int(port) username = config.get('remote_username') or getuser() luser = get_local_user(username) sshdir = get_ssh_dir(config, luser) hostkey = get_host_keys(hostname, sshdir) try: sftp = None keys = get_ssh_keys(sshdir) transport = Transport((hostname, port)) while not keys.empty(): try: key = PKey.from_private_key_file(keys.get()) transport.connect( hostkey=hostkey, username=username, password=None, pkey=key) sftp = SFTPClient.from_transport(transport) break except (PasswordRequiredException, SSHException): pass if sftp is None: raise SaChannelUpdateTransportError("SFTP connection failed") return sftp, transport except BaseException as msg: raise SaChannelUpdateTransportError(msg)
akissa/sachannelupdate
sachannelupdate/transports.py
get_sftp_conn
python
def get_sftp_conn(config): remote = config.get('remote_location') parts = urlparse(remote) if ':' in parts.netloc: hostname, port = parts.netloc.split(':') else: hostname = parts.netloc port = 22 port = int(port) username = config.get('remote_username') or getuser() luser = get_local_user(username) sshdir = get_ssh_dir(config, luser) hostkey = get_host_keys(hostname, sshdir) try: sftp = None keys = get_ssh_keys(sshdir) transport = Transport((hostname, port)) while not keys.empty(): try: key = PKey.from_private_key_file(keys.get()) transport.connect( hostkey=hostkey, username=username, password=None, pkey=key) sftp = SFTPClient.from_transport(transport) break except (PasswordRequiredException, SSHException): pass if sftp is None: raise SaChannelUpdateTransportError("SFTP connection failed") return sftp, transport except BaseException as msg: raise SaChannelUpdateTransportError(msg)
Make a SFTP connection, returns sftp client and connection objects
train
https://github.com/akissa/sachannelupdate/blob/a1c3c3d86b874f9c92c2407e2608963165d3ae98/sachannelupdate/transports.py#L104-L141
[ "def get_ssh_keys(sshdir):\n \"\"\"Get SSH keys\"\"\"\n keys = Queue()\n for root, _, files in os.walk(os.path.abspath(sshdir)):\n if not files:\n continue\n for filename in files:\n fullname = os.path.join(root, filename)\n if (os.path.isfile(fullname) and fullname.endswith('_rsa') or\n fullname.endswith('_dsa')):\n keys.put(fullname)\n return keys\n", "def get_ssh_dir(config, username):\n \"\"\"Get the users ssh dir\"\"\"\n sshdir = config.get('ssh_config_dir')\n if not sshdir:\n sshdir = os.path.expanduser('~/.ssh')\n if not os.path.isdir(sshdir):\n pwentry = getpwnam(username)\n sshdir = os.path.join(pwentry.pw_dir, '.ssh')\n if not os.path.isdir(sshdir):\n sshdir = None\n return sshdir\n", "def get_local_user(username):\n \"\"\"Get the local username\"\"\"\n try:\n _ = getpwnam(username)\n luser = username\n except KeyError:\n luser = getuser()\n return luser\n", "def get_host_keys(hostname, sshdir):\n \"\"\"get host key\"\"\"\n hostkey = None\n\n try:\n host_keys = load_host_keys(os.path.join(sshdir, 'known_hosts'))\n except IOError:\n host_keys = {}\n\n if hostname in host_keys:\n hostkeytype = host_keys[hostname].keys()[0]\n hostkey = host_keys[hostname][hostkeytype]\n\n return hostkey\n" ]
# -*- coding: utf-8 -*- # vim: ai ts=4 sts=4 et sw=4 # sachannelupdate - Utility for pushing updates to Spamassassin update channels # Copyright (C) 2015 Andrew Colin Kissa <andrew@topdog.za.net> # # This program is free software: you can redistribute it and/or modify # it under the terms of the GNU Affero General Public License as published # by the Free Software Foundation, either version 3 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU Affero General Public License for more details. # # You should have received a copy of the GNU Affero General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. """ sachannelupdate: Transports """ import os from Queue import Queue from pwd import getpwnam from getpass import getuser from urlparse import urlparse from paramiko.util import load_host_keys from paramiko import Transport, SFTPClient, PKey, PasswordRequiredException, \ SSHException from sachannelupdate.exceptions import SaChannelUpdateTransportError def get_key_files(kfiles, dirname, names): """Return key files""" for name in names: fullname = os.path.join(dirname, name) if os.path.isfile(fullname) and \ fullname.endswith('_rsa') or \ fullname.endswith('_dsa'): kfiles.put(fullname) def get_ssh_keys(sshdir): """Get SSH keys""" keys = Queue() for root, _, files in os.walk(os.path.abspath(sshdir)): if not files: continue for filename in files: fullname = os.path.join(root, filename) if (os.path.isfile(fullname) and fullname.endswith('_rsa') or fullname.endswith('_dsa')): keys.put(fullname) return keys def get_remote_path(remote_location): """Get the remote path from the remote location""" parts = urlparse(remote_location) return parts.path def get_ssh_dir(config, username): """Get the users ssh dir""" sshdir = config.get('ssh_config_dir') if not sshdir: sshdir = os.path.expanduser('~/.ssh') if not os.path.isdir(sshdir): pwentry = getpwnam(username) sshdir = os.path.join(pwentry.pw_dir, '.ssh') if not os.path.isdir(sshdir): sshdir = None return sshdir def get_local_user(username): """Get the local username""" try: _ = getpwnam(username) luser = username except KeyError: luser = getuser() return luser def get_host_keys(hostname, sshdir): """get host key""" hostkey = None try: host_keys = load_host_keys(os.path.join(sshdir, 'known_hosts')) except IOError: host_keys = {} if hostname in host_keys: hostkeytype = host_keys[hostname].keys()[0] hostkey = host_keys[hostname][hostkeytype] return hostkey
synw/goerr
goerr/messages.py
Msg.fatal
python
def fatal(self, i: int=None) -> str: head = "[" + colors.red("\033[1mfatal error") + "]" if i is not None: head = str(i) + " " + head return head
Returns a fatal error message
train
https://github.com/synw/goerr/blob/08b3809d6715bffe26899a769d96fa5de8573faf/goerr/messages.py#L9-L16
[ "def red(self, *msg):\n color = '\\033[91m'\n return self._msg(color, *msg)\n" ]
class Msg(): """ Class to handle the messages """ def error(self, i: int=None) -> str: """ Returns an error message """ head = "[" + colors.red("error") + "]" if i is not None: head = str(i) + " " + head return head def warning(self, i: int=None) -> str: """ Returns a warning message """ head = "[" + colors.purple("\033[1mwarning") + "]" if i is not None: head = str(i) + " " + head return head def info(self, i: int=None) -> str: """ Returns an info message """ head = "[" + colors.blue("info") + "]" if i is not None: head = str(i) + " " + head return head def via(self, i: int=None) -> str: """ Returns an via message """ head = "[" + colors.green("via") + "]" if i is not None: head = str(i) + " " + head return head def debug(self, i: int=None) -> str: """ Returns a debug message """ head = "[" + colors.yellow("debug") + "]" if i is not None: head = str(i) + " " + head return head
synw/goerr
goerr/messages.py
Msg.error
python
def error(self, i: int=None) -> str: head = "[" + colors.red("error") + "]" if i is not None: head = str(i) + " " + head return head
Returns an error message
train
https://github.com/synw/goerr/blob/08b3809d6715bffe26899a769d96fa5de8573faf/goerr/messages.py#L18-L25
[ "def red(self, *msg):\n color = '\\033[91m'\n return self._msg(color, *msg)\n" ]
class Msg(): """ Class to handle the messages """ def fatal(self, i: int=None) -> str: """ Returns a fatal error message """ head = "[" + colors.red("\033[1mfatal error") + "]" if i is not None: head = str(i) + " " + head return head def warning(self, i: int=None) -> str: """ Returns a warning message """ head = "[" + colors.purple("\033[1mwarning") + "]" if i is not None: head = str(i) + " " + head return head def info(self, i: int=None) -> str: """ Returns an info message """ head = "[" + colors.blue("info") + "]" if i is not None: head = str(i) + " " + head return head def via(self, i: int=None) -> str: """ Returns an via message """ head = "[" + colors.green("via") + "]" if i is not None: head = str(i) + " " + head return head def debug(self, i: int=None) -> str: """ Returns a debug message """ head = "[" + colors.yellow("debug") + "]" if i is not None: head = str(i) + " " + head return head
synw/goerr
goerr/messages.py
Msg.warning
python
def warning(self, i: int=None) -> str: head = "[" + colors.purple("\033[1mwarning") + "]" if i is not None: head = str(i) + " " + head return head
Returns a warning message
train
https://github.com/synw/goerr/blob/08b3809d6715bffe26899a769d96fa5de8573faf/goerr/messages.py#L27-L34
[ "def purple(self, *msg):\n color = '\\033[95m'\n return self._msg(color, *msg)\n" ]
class Msg(): """ Class to handle the messages """ def fatal(self, i: int=None) -> str: """ Returns a fatal error message """ head = "[" + colors.red("\033[1mfatal error") + "]" if i is not None: head = str(i) + " " + head return head def error(self, i: int=None) -> str: """ Returns an error message """ head = "[" + colors.red("error") + "]" if i is not None: head = str(i) + " " + head return head def info(self, i: int=None) -> str: """ Returns an info message """ head = "[" + colors.blue("info") + "]" if i is not None: head = str(i) + " " + head return head def via(self, i: int=None) -> str: """ Returns an via message """ head = "[" + colors.green("via") + "]" if i is not None: head = str(i) + " " + head return head def debug(self, i: int=None) -> str: """ Returns a debug message """ head = "[" + colors.yellow("debug") + "]" if i is not None: head = str(i) + " " + head return head
synw/goerr
goerr/messages.py
Msg.info
python
def info(self, i: int=None) -> str: head = "[" + colors.blue("info") + "]" if i is not None: head = str(i) + " " + head return head
Returns an info message
train
https://github.com/synw/goerr/blob/08b3809d6715bffe26899a769d96fa5de8573faf/goerr/messages.py#L36-L43
[ "def blue(self, *msg):\n color = '\\033[94m'\n return self._msg(color, *msg)\n" ]
class Msg(): """ Class to handle the messages """ def fatal(self, i: int=None) -> str: """ Returns a fatal error message """ head = "[" + colors.red("\033[1mfatal error") + "]" if i is not None: head = str(i) + " " + head return head def error(self, i: int=None) -> str: """ Returns an error message """ head = "[" + colors.red("error") + "]" if i is not None: head = str(i) + " " + head return head def warning(self, i: int=None) -> str: """ Returns a warning message """ head = "[" + colors.purple("\033[1mwarning") + "]" if i is not None: head = str(i) + " " + head return head def via(self, i: int=None) -> str: """ Returns an via message """ head = "[" + colors.green("via") + "]" if i is not None: head = str(i) + " " + head return head def debug(self, i: int=None) -> str: """ Returns a debug message """ head = "[" + colors.yellow("debug") + "]" if i is not None: head = str(i) + " " + head return head
synw/goerr
goerr/messages.py
Msg.via
python
def via(self, i: int=None) -> str: head = "[" + colors.green("via") + "]" if i is not None: head = str(i) + " " + head return head
Returns an via message
train
https://github.com/synw/goerr/blob/08b3809d6715bffe26899a769d96fa5de8573faf/goerr/messages.py#L45-L52
[ "def green(self, *msg):\n color = '\\033[92m'\n return self._msg(color, *msg)\n" ]
class Msg(): """ Class to handle the messages """ def fatal(self, i: int=None) -> str: """ Returns a fatal error message """ head = "[" + colors.red("\033[1mfatal error") + "]" if i is not None: head = str(i) + " " + head return head def error(self, i: int=None) -> str: """ Returns an error message """ head = "[" + colors.red("error") + "]" if i is not None: head = str(i) + " " + head return head def warning(self, i: int=None) -> str: """ Returns a warning message """ head = "[" + colors.purple("\033[1mwarning") + "]" if i is not None: head = str(i) + " " + head return head def info(self, i: int=None) -> str: """ Returns an info message """ head = "[" + colors.blue("info") + "]" if i is not None: head = str(i) + " " + head return head def debug(self, i: int=None) -> str: """ Returns a debug message """ head = "[" + colors.yellow("debug") + "]" if i is not None: head = str(i) + " " + head return head
synw/goerr
goerr/messages.py
Msg.debug
python
def debug(self, i: int=None) -> str: head = "[" + colors.yellow("debug") + "]" if i is not None: head = str(i) + " " + head return head
Returns a debug message
train
https://github.com/synw/goerr/blob/08b3809d6715bffe26899a769d96fa5de8573faf/goerr/messages.py#L54-L61
[ "def yellow(self, *msg):\n color = '\\033[93m'\n return self._msg(color, *msg)\n" ]
class Msg(): """ Class to handle the messages """ def fatal(self, i: int=None) -> str: """ Returns a fatal error message """ head = "[" + colors.red("\033[1mfatal error") + "]" if i is not None: head = str(i) + " " + head return head def error(self, i: int=None) -> str: """ Returns an error message """ head = "[" + colors.red("error") + "]" if i is not None: head = str(i) + " " + head return head def warning(self, i: int=None) -> str: """ Returns a warning message """ head = "[" + colors.purple("\033[1mwarning") + "]" if i is not None: head = str(i) + " " + head return head def info(self, i: int=None) -> str: """ Returns an info message """ head = "[" + colors.blue("info") + "]" if i is not None: head = str(i) + " " + head return head def via(self, i: int=None) -> str: """ Returns an via message """ head = "[" + colors.green("via") + "]" if i is not None: head = str(i) + " " + head return head
synw/goerr
goerr/__init__.py
Err.panic
python
def panic(self, *args): self._err("fatal", *args) if self.test_errs_mode is False: # pragma: no cover sys.exit(1)
Creates a fatal error and exit
train
https://github.com/synw/goerr/blob/08b3809d6715bffe26899a769d96fa5de8573faf/goerr/__init__.py#L59-L65
[ "def _err(self, errclass: str=\"error\", *args) -> \"Err\":\n \"\"\"\n Creates an error\n \"\"\"\n error = self._new_err(errclass, *args)\n if self.log_errs is True:\n sep = \" \"\n if self.log_format == \"csv\":\n sep = \",\"\n msg = str(datetime.now()) + sep + \\\n self._errmsg(error, msgformat=self.log_format)\n self.logger.error(msg)\n print(self._errmsg(error))\n self._add(error)\n return error\n" ]
class Err(): """ Errors manager """ errors = [] # type: List[Err] trace_errs = False # type: bool errs_traceback = True # type: bool logger = None # type: logging.Logger log_errs = False # type: bool log_format = "csv" # type: str log_path = "errors.log" # type: str test_errs_mode = False # type: bool def __init__(self, function: str=None, date: datetime=datetime.now(), msg: str=None, errtype: str=None, errclass: str=None, line: int=None, file: str=None, code: str=None, tb: str=None, ex: Exception=None, caller: str=None, caller_msg: str=None): """ Datastructure of an error """ self.date = date # type: datetime.datetime self.function = function # type: str self.msg = msg # type: str self.errtype = errtype # type: str self.errclass = errclass # type: str self.line = line # type: int self.file = file # type: str self.code = code # type: str self.tb = tb # type: str self.ex = ex # type: Exception self.caller = caller # type: str self.caller_msg = caller_msg # type: str self.new = self.err def __repr__(self): msg = "<goerror.Err object: " + str(self.errclass) + " error>" return msg def __str__(self): return self.msg def err(self, *args): """ Creates an error """ error = self._err("error", *args) return error # pragma: no cover def warning(self, *args) -> "Err": """ Creates a warning message """ error = self._create_err("warning", *args) print(self._errmsg(error)) return error def info(self, *args) -> "Err": """ Creates an info message """ error = self._create_err("info", *args) print(self._errmsg(error)) return error def debug(self, *args) -> "Err": """ Creates a debug message """ error = self._create_err("debug", *args) print(self._errmsg(error)) return error def _create_err(self, errclass: str, *args) -> "Err": """ Create an error """ error = self._new_err(errclass, *args) self._add(error) return error def _err(self, errclass: str="error", *args) -> "Err": """ Creates an error """ error = self._new_err(errclass, *args) if self.log_errs is True: sep = " " if self.log_format == "csv": sep = "," msg = str(datetime.now()) + sep + \ self._errmsg(error, msgformat=self.log_format) self.logger.error(msg) print(self._errmsg(error)) self._add(error) return error def _new_err(self, errclass: str, *args) -> 'Err': """ Error constructor """ # get the message or exception ex, msg = self._get_args(*args) # construct the error # handle exception ftb = None # type: str function = None # type: str errtype = None # type: str file = None # type: str line = None # type: int code = None # type: str ex_msg = None # type: str caller = None # type: str caller_msg = None # type: str st = inspect.stack() if ex is not None: # get info from exception errobj, ex_msg, tb = sys.exc_info() tb = traceback.extract_tb(tb) file, line, function, code = tb[-1] # if called from an external lib if len(tb) > 1: file, line, caller, code = tb[0] else: call_stack = [] for c in st: call_stack.append(c[3]) caller = self._get_caller(call_stack, function) internals = [ "err", "_new_err", "fatal", "warning", "debug", "info", "<module>"] if caller == function or caller in internals: caller = None # handle messages if msg is not None: caller_msg = msg msg = str(ex_msg) else: msg = str(ex_msg) ftb = traceback.format_exc() errtype = errobj.__name__ if function is None: # for el in st: # print(el) function = st[3][3] if function == "<module>": function = None # init error object date = datetime.now() error = Err( function, date, msg, errtype, errclass, line, file, code, ftb, ex, caller, caller_msg) return error def _headline(self, error, i: int) -> str: """ Format the error message's headline """ msgs = Msg() # get the error title if error.errclass == "fatal": msg = msgs.fatal(i) elif error.errclass == "warning": msg = msgs.warning(i) elif error.errclass == "info": msg = msgs.info(i) elif error.errclass == "debug": msg = msgs.debug(i) elif error.errclass == "via": msg = msgs.via(i) else: msg = msgs.error(i) # function name if error.function is not None: msg += " from " + colors.bold(error.function) if error.caller is not None: msg += " called from " + colors.bold(error.caller) if error.caller_msg is not None: msg += "\n" + error.caller_msg if error.function is not None and error.msg is not None: msg += ": " else: msg = msg + " " if error.errtype is not None: msg += error.errtype + " : " if error.msg is not None: msg += error.msg return msg def _errmsg(self, error: "Err", tb: bool=False, i: int=None, msgformat: str="terminal") -> str: """ Get the error message """ if msgformat == "terminal": msg = self._headline(error, i) if error.ex is not None: msg += "\n" + "line " + colors.bold(str(error.line)) msg += ": " + colors.yellow(error.code) msg += "\n" + str(error.file) if self.errs_traceback is True or tb is True: if error.tb is not None: msg += "\n" + error.tb elif msgformat == "csv": sep = "," msg = error.msg + sep msg += str(error.line) + sep + error.code + sep msg += str(error.file) elif msgformat == "text": sep = "," msg = error.msg if error.ex is not None: msg += sep + str(error.line) + sep + error.code + sep msg += str(error.file) + sep if self.errs_traceback is True or tb is True: if error.tb is not None: msg += sep + error.tb elif msgformat == "dict": msg = {"date": datetime.now()} if error.ex is not None: msg["msg"] = error.msg msg["line"] = error.line msg["code"] = error.code msg["file"] = error.file if self.errs_traceback is True or tb is True: if error.tb is not None: msg["traceback"] = error.tb return msg def to_dict(self): """ Returns a dictionnary with the error elements """ return self._errmsg(self, msgformat="dict") def _print_errs(self): """ Prints the errors trace with tracebacks """ i = 0 for error in self.errors: print(self._errmsg(error, tb=True, i=i)) # for spacing if self.errs_traceback is False: print() i += 1 def _add(self, error: "Err"): """ Adds an error to the trace if required """ if self.trace_errs is True: self.errors.append(error) def _get_caller(self, callers: List[str], function: str) -> str: """ Get the caller function from the provided function """ is_next = False for c in callers: if is_next is True: return c if function == c: is_next = True def _get_args(self, *args) -> (Exception, str): """ Returns exception and message from the provided arguments """ ex = None msg = None for arg in args: if isinstance(arg, str): msg = arg elif isinstance(arg, Exception): ex = arg return ex, msg
synw/goerr
goerr/__init__.py
Err.warning
python
def warning(self, *args) -> "Err": error = self._create_err("warning", *args) print(self._errmsg(error)) return error
Creates a warning message
train
https://github.com/synw/goerr/blob/08b3809d6715bffe26899a769d96fa5de8573faf/goerr/__init__.py#L67-L73
[ "def _create_err(self, errclass: str, *args) -> \"Err\":\n \"\"\"\n Create an error\n \"\"\"\n error = self._new_err(errclass, *args)\n self._add(error)\n return error\n", "def _errmsg(self, error: \"Err\", tb: bool=False, i: int=None,\n msgformat: str=\"terminal\") -> str:\n \"\"\"\n Get the error message\n \"\"\"\n if msgformat == \"terminal\":\n msg = self._headline(error, i)\n if error.ex is not None:\n msg += \"\\n\" + \"line \" + colors.bold(str(error.line))\n msg += \": \" + colors.yellow(error.code)\n msg += \"\\n\" + str(error.file)\n if self.errs_traceback is True or tb is True:\n if error.tb is not None:\n msg += \"\\n\" + error.tb\n elif msgformat == \"csv\":\n sep = \",\"\n msg = error.msg + sep\n msg += str(error.line) + sep + error.code + sep\n msg += str(error.file)\n elif msgformat == \"text\":\n sep = \",\"\n msg = error.msg\n if error.ex is not None:\n msg += sep + str(error.line) + sep + error.code + sep\n msg += str(error.file) + sep \n if self.errs_traceback is True or tb is True:\n if error.tb is not None:\n msg += sep + error.tb\n elif msgformat == \"dict\":\n msg = {\"date\": datetime.now()}\n if error.ex is not None:\n msg[\"msg\"] = error.msg\n msg[\"line\"] = error.line\n msg[\"code\"] = error.code\n msg[\"file\"] = error.file \n if self.errs_traceback is True or tb is True:\n if error.tb is not None:\n msg[\"traceback\"] = error.tb\n return msg\n" ]
class Err(): """ Errors manager """ errors = [] # type: List[Err] trace_errs = False # type: bool errs_traceback = True # type: bool logger = None # type: logging.Logger log_errs = False # type: bool log_format = "csv" # type: str log_path = "errors.log" # type: str test_errs_mode = False # type: bool def __init__(self, function: str=None, date: datetime=datetime.now(), msg: str=None, errtype: str=None, errclass: str=None, line: int=None, file: str=None, code: str=None, tb: str=None, ex: Exception=None, caller: str=None, caller_msg: str=None): """ Datastructure of an error """ self.date = date # type: datetime.datetime self.function = function # type: str self.msg = msg # type: str self.errtype = errtype # type: str self.errclass = errclass # type: str self.line = line # type: int self.file = file # type: str self.code = code # type: str self.tb = tb # type: str self.ex = ex # type: Exception self.caller = caller # type: str self.caller_msg = caller_msg # type: str self.new = self.err def __repr__(self): msg = "<goerror.Err object: " + str(self.errclass) + " error>" return msg def __str__(self): return self.msg def err(self, *args): """ Creates an error """ error = self._err("error", *args) return error def panic(self, *args): """ Creates a fatal error and exit """ self._err("fatal", *args) if self.test_errs_mode is False: # pragma: no cover sys.exit(1) # pragma: no cover def info(self, *args) -> "Err": """ Creates an info message """ error = self._create_err("info", *args) print(self._errmsg(error)) return error def debug(self, *args) -> "Err": """ Creates a debug message """ error = self._create_err("debug", *args) print(self._errmsg(error)) return error def _create_err(self, errclass: str, *args) -> "Err": """ Create an error """ error = self._new_err(errclass, *args) self._add(error) return error def _err(self, errclass: str="error", *args) -> "Err": """ Creates an error """ error = self._new_err(errclass, *args) if self.log_errs is True: sep = " " if self.log_format == "csv": sep = "," msg = str(datetime.now()) + sep + \ self._errmsg(error, msgformat=self.log_format) self.logger.error(msg) print(self._errmsg(error)) self._add(error) return error def _new_err(self, errclass: str, *args) -> 'Err': """ Error constructor """ # get the message or exception ex, msg = self._get_args(*args) # construct the error # handle exception ftb = None # type: str function = None # type: str errtype = None # type: str file = None # type: str line = None # type: int code = None # type: str ex_msg = None # type: str caller = None # type: str caller_msg = None # type: str st = inspect.stack() if ex is not None: # get info from exception errobj, ex_msg, tb = sys.exc_info() tb = traceback.extract_tb(tb) file, line, function, code = tb[-1] # if called from an external lib if len(tb) > 1: file, line, caller, code = tb[0] else: call_stack = [] for c in st: call_stack.append(c[3]) caller = self._get_caller(call_stack, function) internals = [ "err", "_new_err", "fatal", "warning", "debug", "info", "<module>"] if caller == function or caller in internals: caller = None # handle messages if msg is not None: caller_msg = msg msg = str(ex_msg) else: msg = str(ex_msg) ftb = traceback.format_exc() errtype = errobj.__name__ if function is None: # for el in st: # print(el) function = st[3][3] if function == "<module>": function = None # init error object date = datetime.now() error = Err( function, date, msg, errtype, errclass, line, file, code, ftb, ex, caller, caller_msg) return error def _headline(self, error, i: int) -> str: """ Format the error message's headline """ msgs = Msg() # get the error title if error.errclass == "fatal": msg = msgs.fatal(i) elif error.errclass == "warning": msg = msgs.warning(i) elif error.errclass == "info": msg = msgs.info(i) elif error.errclass == "debug": msg = msgs.debug(i) elif error.errclass == "via": msg = msgs.via(i) else: msg = msgs.error(i) # function name if error.function is not None: msg += " from " + colors.bold(error.function) if error.caller is not None: msg += " called from " + colors.bold(error.caller) if error.caller_msg is not None: msg += "\n" + error.caller_msg if error.function is not None and error.msg is not None: msg += ": " else: msg = msg + " " if error.errtype is not None: msg += error.errtype + " : " if error.msg is not None: msg += error.msg return msg def _errmsg(self, error: "Err", tb: bool=False, i: int=None, msgformat: str="terminal") -> str: """ Get the error message """ if msgformat == "terminal": msg = self._headline(error, i) if error.ex is not None: msg += "\n" + "line " + colors.bold(str(error.line)) msg += ": " + colors.yellow(error.code) msg += "\n" + str(error.file) if self.errs_traceback is True or tb is True: if error.tb is not None: msg += "\n" + error.tb elif msgformat == "csv": sep = "," msg = error.msg + sep msg += str(error.line) + sep + error.code + sep msg += str(error.file) elif msgformat == "text": sep = "," msg = error.msg if error.ex is not None: msg += sep + str(error.line) + sep + error.code + sep msg += str(error.file) + sep if self.errs_traceback is True or tb is True: if error.tb is not None: msg += sep + error.tb elif msgformat == "dict": msg = {"date": datetime.now()} if error.ex is not None: msg["msg"] = error.msg msg["line"] = error.line msg["code"] = error.code msg["file"] = error.file if self.errs_traceback is True or tb is True: if error.tb is not None: msg["traceback"] = error.tb return msg def to_dict(self): """ Returns a dictionnary with the error elements """ return self._errmsg(self, msgformat="dict") def _print_errs(self): """ Prints the errors trace with tracebacks """ i = 0 for error in self.errors: print(self._errmsg(error, tb=True, i=i)) # for spacing if self.errs_traceback is False: print() i += 1 def _add(self, error: "Err"): """ Adds an error to the trace if required """ if self.trace_errs is True: self.errors.append(error) def _get_caller(self, callers: List[str], function: str) -> str: """ Get the caller function from the provided function """ is_next = False for c in callers: if is_next is True: return c if function == c: is_next = True def _get_args(self, *args) -> (Exception, str): """ Returns exception and message from the provided arguments """ ex = None msg = None for arg in args: if isinstance(arg, str): msg = arg elif isinstance(arg, Exception): ex = arg return ex, msg
synw/goerr
goerr/__init__.py
Err.info
python
def info(self, *args) -> "Err": error = self._create_err("info", *args) print(self._errmsg(error)) return error
Creates an info message
train
https://github.com/synw/goerr/blob/08b3809d6715bffe26899a769d96fa5de8573faf/goerr/__init__.py#L75-L81
[ "def _create_err(self, errclass: str, *args) -> \"Err\":\n \"\"\"\n Create an error\n \"\"\"\n error = self._new_err(errclass, *args)\n self._add(error)\n return error\n", "def _errmsg(self, error: \"Err\", tb: bool=False, i: int=None,\n msgformat: str=\"terminal\") -> str:\n \"\"\"\n Get the error message\n \"\"\"\n if msgformat == \"terminal\":\n msg = self._headline(error, i)\n if error.ex is not None:\n msg += \"\\n\" + \"line \" + colors.bold(str(error.line))\n msg += \": \" + colors.yellow(error.code)\n msg += \"\\n\" + str(error.file)\n if self.errs_traceback is True or tb is True:\n if error.tb is not None:\n msg += \"\\n\" + error.tb\n elif msgformat == \"csv\":\n sep = \",\"\n msg = error.msg + sep\n msg += str(error.line) + sep + error.code + sep\n msg += str(error.file)\n elif msgformat == \"text\":\n sep = \",\"\n msg = error.msg\n if error.ex is not None:\n msg += sep + str(error.line) + sep + error.code + sep\n msg += str(error.file) + sep \n if self.errs_traceback is True or tb is True:\n if error.tb is not None:\n msg += sep + error.tb\n elif msgformat == \"dict\":\n msg = {\"date\": datetime.now()}\n if error.ex is not None:\n msg[\"msg\"] = error.msg\n msg[\"line\"] = error.line\n msg[\"code\"] = error.code\n msg[\"file\"] = error.file \n if self.errs_traceback is True or tb is True:\n if error.tb is not None:\n msg[\"traceback\"] = error.tb\n return msg\n" ]
class Err(): """ Errors manager """ errors = [] # type: List[Err] trace_errs = False # type: bool errs_traceback = True # type: bool logger = None # type: logging.Logger log_errs = False # type: bool log_format = "csv" # type: str log_path = "errors.log" # type: str test_errs_mode = False # type: bool def __init__(self, function: str=None, date: datetime=datetime.now(), msg: str=None, errtype: str=None, errclass: str=None, line: int=None, file: str=None, code: str=None, tb: str=None, ex: Exception=None, caller: str=None, caller_msg: str=None): """ Datastructure of an error """ self.date = date # type: datetime.datetime self.function = function # type: str self.msg = msg # type: str self.errtype = errtype # type: str self.errclass = errclass # type: str self.line = line # type: int self.file = file # type: str self.code = code # type: str self.tb = tb # type: str self.ex = ex # type: Exception self.caller = caller # type: str self.caller_msg = caller_msg # type: str self.new = self.err def __repr__(self): msg = "<goerror.Err object: " + str(self.errclass) + " error>" return msg def __str__(self): return self.msg def err(self, *args): """ Creates an error """ error = self._err("error", *args) return error def panic(self, *args): """ Creates a fatal error and exit """ self._err("fatal", *args) if self.test_errs_mode is False: # pragma: no cover sys.exit(1) # pragma: no cover def warning(self, *args) -> "Err": """ Creates a warning message """ error = self._create_err("warning", *args) print(self._errmsg(error)) return error def debug(self, *args) -> "Err": """ Creates a debug message """ error = self._create_err("debug", *args) print(self._errmsg(error)) return error def _create_err(self, errclass: str, *args) -> "Err": """ Create an error """ error = self._new_err(errclass, *args) self._add(error) return error def _err(self, errclass: str="error", *args) -> "Err": """ Creates an error """ error = self._new_err(errclass, *args) if self.log_errs is True: sep = " " if self.log_format == "csv": sep = "," msg = str(datetime.now()) + sep + \ self._errmsg(error, msgformat=self.log_format) self.logger.error(msg) print(self._errmsg(error)) self._add(error) return error def _new_err(self, errclass: str, *args) -> 'Err': """ Error constructor """ # get the message or exception ex, msg = self._get_args(*args) # construct the error # handle exception ftb = None # type: str function = None # type: str errtype = None # type: str file = None # type: str line = None # type: int code = None # type: str ex_msg = None # type: str caller = None # type: str caller_msg = None # type: str st = inspect.stack() if ex is not None: # get info from exception errobj, ex_msg, tb = sys.exc_info() tb = traceback.extract_tb(tb) file, line, function, code = tb[-1] # if called from an external lib if len(tb) > 1: file, line, caller, code = tb[0] else: call_stack = [] for c in st: call_stack.append(c[3]) caller = self._get_caller(call_stack, function) internals = [ "err", "_new_err", "fatal", "warning", "debug", "info", "<module>"] if caller == function or caller in internals: caller = None # handle messages if msg is not None: caller_msg = msg msg = str(ex_msg) else: msg = str(ex_msg) ftb = traceback.format_exc() errtype = errobj.__name__ if function is None: # for el in st: # print(el) function = st[3][3] if function == "<module>": function = None # init error object date = datetime.now() error = Err( function, date, msg, errtype, errclass, line, file, code, ftb, ex, caller, caller_msg) return error def _headline(self, error, i: int) -> str: """ Format the error message's headline """ msgs = Msg() # get the error title if error.errclass == "fatal": msg = msgs.fatal(i) elif error.errclass == "warning": msg = msgs.warning(i) elif error.errclass == "info": msg = msgs.info(i) elif error.errclass == "debug": msg = msgs.debug(i) elif error.errclass == "via": msg = msgs.via(i) else: msg = msgs.error(i) # function name if error.function is not None: msg += " from " + colors.bold(error.function) if error.caller is not None: msg += " called from " + colors.bold(error.caller) if error.caller_msg is not None: msg += "\n" + error.caller_msg if error.function is not None and error.msg is not None: msg += ": " else: msg = msg + " " if error.errtype is not None: msg += error.errtype + " : " if error.msg is not None: msg += error.msg return msg def _errmsg(self, error: "Err", tb: bool=False, i: int=None, msgformat: str="terminal") -> str: """ Get the error message """ if msgformat == "terminal": msg = self._headline(error, i) if error.ex is not None: msg += "\n" + "line " + colors.bold(str(error.line)) msg += ": " + colors.yellow(error.code) msg += "\n" + str(error.file) if self.errs_traceback is True or tb is True: if error.tb is not None: msg += "\n" + error.tb elif msgformat == "csv": sep = "," msg = error.msg + sep msg += str(error.line) + sep + error.code + sep msg += str(error.file) elif msgformat == "text": sep = "," msg = error.msg if error.ex is not None: msg += sep + str(error.line) + sep + error.code + sep msg += str(error.file) + sep if self.errs_traceback is True or tb is True: if error.tb is not None: msg += sep + error.tb elif msgformat == "dict": msg = {"date": datetime.now()} if error.ex is not None: msg["msg"] = error.msg msg["line"] = error.line msg["code"] = error.code msg["file"] = error.file if self.errs_traceback is True or tb is True: if error.tb is not None: msg["traceback"] = error.tb return msg def to_dict(self): """ Returns a dictionnary with the error elements """ return self._errmsg(self, msgformat="dict") def _print_errs(self): """ Prints the errors trace with tracebacks """ i = 0 for error in self.errors: print(self._errmsg(error, tb=True, i=i)) # for spacing if self.errs_traceback is False: print() i += 1 def _add(self, error: "Err"): """ Adds an error to the trace if required """ if self.trace_errs is True: self.errors.append(error) def _get_caller(self, callers: List[str], function: str) -> str: """ Get the caller function from the provided function """ is_next = False for c in callers: if is_next is True: return c if function == c: is_next = True def _get_args(self, *args) -> (Exception, str): """ Returns exception and message from the provided arguments """ ex = None msg = None for arg in args: if isinstance(arg, str): msg = arg elif isinstance(arg, Exception): ex = arg return ex, msg
synw/goerr
goerr/__init__.py
Err.debug
python
def debug(self, *args) -> "Err": error = self._create_err("debug", *args) print(self._errmsg(error)) return error
Creates a debug message
train
https://github.com/synw/goerr/blob/08b3809d6715bffe26899a769d96fa5de8573faf/goerr/__init__.py#L83-L89
[ "def _create_err(self, errclass: str, *args) -> \"Err\":\n \"\"\"\n Create an error\n \"\"\"\n error = self._new_err(errclass, *args)\n self._add(error)\n return error\n", "def _errmsg(self, error: \"Err\", tb: bool=False, i: int=None,\n msgformat: str=\"terminal\") -> str:\n \"\"\"\n Get the error message\n \"\"\"\n if msgformat == \"terminal\":\n msg = self._headline(error, i)\n if error.ex is not None:\n msg += \"\\n\" + \"line \" + colors.bold(str(error.line))\n msg += \": \" + colors.yellow(error.code)\n msg += \"\\n\" + str(error.file)\n if self.errs_traceback is True or tb is True:\n if error.tb is not None:\n msg += \"\\n\" + error.tb\n elif msgformat == \"csv\":\n sep = \",\"\n msg = error.msg + sep\n msg += str(error.line) + sep + error.code + sep\n msg += str(error.file)\n elif msgformat == \"text\":\n sep = \",\"\n msg = error.msg\n if error.ex is not None:\n msg += sep + str(error.line) + sep + error.code + sep\n msg += str(error.file) + sep \n if self.errs_traceback is True or tb is True:\n if error.tb is not None:\n msg += sep + error.tb\n elif msgformat == \"dict\":\n msg = {\"date\": datetime.now()}\n if error.ex is not None:\n msg[\"msg\"] = error.msg\n msg[\"line\"] = error.line\n msg[\"code\"] = error.code\n msg[\"file\"] = error.file \n if self.errs_traceback is True or tb is True:\n if error.tb is not None:\n msg[\"traceback\"] = error.tb\n return msg\n" ]
class Err(): """ Errors manager """ errors = [] # type: List[Err] trace_errs = False # type: bool errs_traceback = True # type: bool logger = None # type: logging.Logger log_errs = False # type: bool log_format = "csv" # type: str log_path = "errors.log" # type: str test_errs_mode = False # type: bool def __init__(self, function: str=None, date: datetime=datetime.now(), msg: str=None, errtype: str=None, errclass: str=None, line: int=None, file: str=None, code: str=None, tb: str=None, ex: Exception=None, caller: str=None, caller_msg: str=None): """ Datastructure of an error """ self.date = date # type: datetime.datetime self.function = function # type: str self.msg = msg # type: str self.errtype = errtype # type: str self.errclass = errclass # type: str self.line = line # type: int self.file = file # type: str self.code = code # type: str self.tb = tb # type: str self.ex = ex # type: Exception self.caller = caller # type: str self.caller_msg = caller_msg # type: str self.new = self.err def __repr__(self): msg = "<goerror.Err object: " + str(self.errclass) + " error>" return msg def __str__(self): return self.msg def err(self, *args): """ Creates an error """ error = self._err("error", *args) return error def panic(self, *args): """ Creates a fatal error and exit """ self._err("fatal", *args) if self.test_errs_mode is False: # pragma: no cover sys.exit(1) # pragma: no cover def warning(self, *args) -> "Err": """ Creates a warning message """ error = self._create_err("warning", *args) print(self._errmsg(error)) return error def info(self, *args) -> "Err": """ Creates an info message """ error = self._create_err("info", *args) print(self._errmsg(error)) return error def _create_err(self, errclass: str, *args) -> "Err": """ Create an error """ error = self._new_err(errclass, *args) self._add(error) return error def _err(self, errclass: str="error", *args) -> "Err": """ Creates an error """ error = self._new_err(errclass, *args) if self.log_errs is True: sep = " " if self.log_format == "csv": sep = "," msg = str(datetime.now()) + sep + \ self._errmsg(error, msgformat=self.log_format) self.logger.error(msg) print(self._errmsg(error)) self._add(error) return error def _new_err(self, errclass: str, *args) -> 'Err': """ Error constructor """ # get the message or exception ex, msg = self._get_args(*args) # construct the error # handle exception ftb = None # type: str function = None # type: str errtype = None # type: str file = None # type: str line = None # type: int code = None # type: str ex_msg = None # type: str caller = None # type: str caller_msg = None # type: str st = inspect.stack() if ex is not None: # get info from exception errobj, ex_msg, tb = sys.exc_info() tb = traceback.extract_tb(tb) file, line, function, code = tb[-1] # if called from an external lib if len(tb) > 1: file, line, caller, code = tb[0] else: call_stack = [] for c in st: call_stack.append(c[3]) caller = self._get_caller(call_stack, function) internals = [ "err", "_new_err", "fatal", "warning", "debug", "info", "<module>"] if caller == function or caller in internals: caller = None # handle messages if msg is not None: caller_msg = msg msg = str(ex_msg) else: msg = str(ex_msg) ftb = traceback.format_exc() errtype = errobj.__name__ if function is None: # for el in st: # print(el) function = st[3][3] if function == "<module>": function = None # init error object date = datetime.now() error = Err( function, date, msg, errtype, errclass, line, file, code, ftb, ex, caller, caller_msg) return error def _headline(self, error, i: int) -> str: """ Format the error message's headline """ msgs = Msg() # get the error title if error.errclass == "fatal": msg = msgs.fatal(i) elif error.errclass == "warning": msg = msgs.warning(i) elif error.errclass == "info": msg = msgs.info(i) elif error.errclass == "debug": msg = msgs.debug(i) elif error.errclass == "via": msg = msgs.via(i) else: msg = msgs.error(i) # function name if error.function is not None: msg += " from " + colors.bold(error.function) if error.caller is not None: msg += " called from " + colors.bold(error.caller) if error.caller_msg is not None: msg += "\n" + error.caller_msg if error.function is not None and error.msg is not None: msg += ": " else: msg = msg + " " if error.errtype is not None: msg += error.errtype + " : " if error.msg is not None: msg += error.msg return msg def _errmsg(self, error: "Err", tb: bool=False, i: int=None, msgformat: str="terminal") -> str: """ Get the error message """ if msgformat == "terminal": msg = self._headline(error, i) if error.ex is not None: msg += "\n" + "line " + colors.bold(str(error.line)) msg += ": " + colors.yellow(error.code) msg += "\n" + str(error.file) if self.errs_traceback is True or tb is True: if error.tb is not None: msg += "\n" + error.tb elif msgformat == "csv": sep = "," msg = error.msg + sep msg += str(error.line) + sep + error.code + sep msg += str(error.file) elif msgformat == "text": sep = "," msg = error.msg if error.ex is not None: msg += sep + str(error.line) + sep + error.code + sep msg += str(error.file) + sep if self.errs_traceback is True or tb is True: if error.tb is not None: msg += sep + error.tb elif msgformat == "dict": msg = {"date": datetime.now()} if error.ex is not None: msg["msg"] = error.msg msg["line"] = error.line msg["code"] = error.code msg["file"] = error.file if self.errs_traceback is True or tb is True: if error.tb is not None: msg["traceback"] = error.tb return msg def to_dict(self): """ Returns a dictionnary with the error elements """ return self._errmsg(self, msgformat="dict") def _print_errs(self): """ Prints the errors trace with tracebacks """ i = 0 for error in self.errors: print(self._errmsg(error, tb=True, i=i)) # for spacing if self.errs_traceback is False: print() i += 1 def _add(self, error: "Err"): """ Adds an error to the trace if required """ if self.trace_errs is True: self.errors.append(error) def _get_caller(self, callers: List[str], function: str) -> str: """ Get the caller function from the provided function """ is_next = False for c in callers: if is_next is True: return c if function == c: is_next = True def _get_args(self, *args) -> (Exception, str): """ Returns exception and message from the provided arguments """ ex = None msg = None for arg in args: if isinstance(arg, str): msg = arg elif isinstance(arg, Exception): ex = arg return ex, msg
synw/goerr
goerr/__init__.py
Err._create_err
python
def _create_err(self, errclass: str, *args) -> "Err": error = self._new_err(errclass, *args) self._add(error) return error
Create an error
train
https://github.com/synw/goerr/blob/08b3809d6715bffe26899a769d96fa5de8573faf/goerr/__init__.py#L91-L97
[ "def _new_err(self, errclass: str, *args) -> 'Err':\n \"\"\"\n Error constructor\n \"\"\"\n # get the message or exception\n ex, msg = self._get_args(*args)\n # construct the error\n # handle exception\n ftb = None # type: str\n function = None # type: str\n errtype = None # type: str\n file = None # type: str\n line = None # type: int\n code = None # type: str\n ex_msg = None # type: str\n caller = None # type: str\n caller_msg = None # type: str\n\n st = inspect.stack()\n\n if ex is not None:\n # get info from exception\n errobj, ex_msg, tb = sys.exc_info()\n tb = traceback.extract_tb(tb)\n file, line, function, code = tb[-1]\n # if called from an external lib\n if len(tb) > 1:\n file, line, caller, code = tb[0]\n else:\n call_stack = []\n for c in st:\n call_stack.append(c[3])\n caller = self._get_caller(call_stack, function)\n\n internals = [\n \"err\",\n \"_new_err\",\n \"fatal\",\n \"warning\",\n \"debug\",\n \"info\",\n \"<module>\"] \n if caller == function or caller in internals:\n caller = None\n # handle messages\n if msg is not None:\n caller_msg = msg\n msg = str(ex_msg)\n else:\n msg = str(ex_msg)\n ftb = traceback.format_exc()\n errtype = errobj.__name__\n if function is None:\n # for el in st:\n # print(el)\n function = st[3][3]\n if function == \"<module>\":\n function = None\n # init error object\n date = datetime.now()\n error = Err(\n function,\n date,\n msg,\n errtype,\n errclass,\n line,\n file,\n code,\n ftb,\n ex,\n caller,\n caller_msg)\n return error\n", "def _add(self, error: \"Err\"):\n \"\"\"\n Adds an error to the trace if required\n \"\"\"\n if self.trace_errs is True:\n self.errors.append(error)\n" ]
class Err(): """ Errors manager """ errors = [] # type: List[Err] trace_errs = False # type: bool errs_traceback = True # type: bool logger = None # type: logging.Logger log_errs = False # type: bool log_format = "csv" # type: str log_path = "errors.log" # type: str test_errs_mode = False # type: bool def __init__(self, function: str=None, date: datetime=datetime.now(), msg: str=None, errtype: str=None, errclass: str=None, line: int=None, file: str=None, code: str=None, tb: str=None, ex: Exception=None, caller: str=None, caller_msg: str=None): """ Datastructure of an error """ self.date = date # type: datetime.datetime self.function = function # type: str self.msg = msg # type: str self.errtype = errtype # type: str self.errclass = errclass # type: str self.line = line # type: int self.file = file # type: str self.code = code # type: str self.tb = tb # type: str self.ex = ex # type: Exception self.caller = caller # type: str self.caller_msg = caller_msg # type: str self.new = self.err def __repr__(self): msg = "<goerror.Err object: " + str(self.errclass) + " error>" return msg def __str__(self): return self.msg def err(self, *args): """ Creates an error """ error = self._err("error", *args) return error def panic(self, *args): """ Creates a fatal error and exit """ self._err("fatal", *args) if self.test_errs_mode is False: # pragma: no cover sys.exit(1) # pragma: no cover def warning(self, *args) -> "Err": """ Creates a warning message """ error = self._create_err("warning", *args) print(self._errmsg(error)) return error def info(self, *args) -> "Err": """ Creates an info message """ error = self._create_err("info", *args) print(self._errmsg(error)) return error def debug(self, *args) -> "Err": """ Creates a debug message """ error = self._create_err("debug", *args) print(self._errmsg(error)) return error def _err(self, errclass: str="error", *args) -> "Err": """ Creates an error """ error = self._new_err(errclass, *args) if self.log_errs is True: sep = " " if self.log_format == "csv": sep = "," msg = str(datetime.now()) + sep + \ self._errmsg(error, msgformat=self.log_format) self.logger.error(msg) print(self._errmsg(error)) self._add(error) return error def _new_err(self, errclass: str, *args) -> 'Err': """ Error constructor """ # get the message or exception ex, msg = self._get_args(*args) # construct the error # handle exception ftb = None # type: str function = None # type: str errtype = None # type: str file = None # type: str line = None # type: int code = None # type: str ex_msg = None # type: str caller = None # type: str caller_msg = None # type: str st = inspect.stack() if ex is not None: # get info from exception errobj, ex_msg, tb = sys.exc_info() tb = traceback.extract_tb(tb) file, line, function, code = tb[-1] # if called from an external lib if len(tb) > 1: file, line, caller, code = tb[0] else: call_stack = [] for c in st: call_stack.append(c[3]) caller = self._get_caller(call_stack, function) internals = [ "err", "_new_err", "fatal", "warning", "debug", "info", "<module>"] if caller == function or caller in internals: caller = None # handle messages if msg is not None: caller_msg = msg msg = str(ex_msg) else: msg = str(ex_msg) ftb = traceback.format_exc() errtype = errobj.__name__ if function is None: # for el in st: # print(el) function = st[3][3] if function == "<module>": function = None # init error object date = datetime.now() error = Err( function, date, msg, errtype, errclass, line, file, code, ftb, ex, caller, caller_msg) return error def _headline(self, error, i: int) -> str: """ Format the error message's headline """ msgs = Msg() # get the error title if error.errclass == "fatal": msg = msgs.fatal(i) elif error.errclass == "warning": msg = msgs.warning(i) elif error.errclass == "info": msg = msgs.info(i) elif error.errclass == "debug": msg = msgs.debug(i) elif error.errclass == "via": msg = msgs.via(i) else: msg = msgs.error(i) # function name if error.function is not None: msg += " from " + colors.bold(error.function) if error.caller is not None: msg += " called from " + colors.bold(error.caller) if error.caller_msg is not None: msg += "\n" + error.caller_msg if error.function is not None and error.msg is not None: msg += ": " else: msg = msg + " " if error.errtype is not None: msg += error.errtype + " : " if error.msg is not None: msg += error.msg return msg def _errmsg(self, error: "Err", tb: bool=False, i: int=None, msgformat: str="terminal") -> str: """ Get the error message """ if msgformat == "terminal": msg = self._headline(error, i) if error.ex is not None: msg += "\n" + "line " + colors.bold(str(error.line)) msg += ": " + colors.yellow(error.code) msg += "\n" + str(error.file) if self.errs_traceback is True or tb is True: if error.tb is not None: msg += "\n" + error.tb elif msgformat == "csv": sep = "," msg = error.msg + sep msg += str(error.line) + sep + error.code + sep msg += str(error.file) elif msgformat == "text": sep = "," msg = error.msg if error.ex is not None: msg += sep + str(error.line) + sep + error.code + sep msg += str(error.file) + sep if self.errs_traceback is True or tb is True: if error.tb is not None: msg += sep + error.tb elif msgformat == "dict": msg = {"date": datetime.now()} if error.ex is not None: msg["msg"] = error.msg msg["line"] = error.line msg["code"] = error.code msg["file"] = error.file if self.errs_traceback is True or tb is True: if error.tb is not None: msg["traceback"] = error.tb return msg def to_dict(self): """ Returns a dictionnary with the error elements """ return self._errmsg(self, msgformat="dict") def _print_errs(self): """ Prints the errors trace with tracebacks """ i = 0 for error in self.errors: print(self._errmsg(error, tb=True, i=i)) # for spacing if self.errs_traceback is False: print() i += 1 def _add(self, error: "Err"): """ Adds an error to the trace if required """ if self.trace_errs is True: self.errors.append(error) def _get_caller(self, callers: List[str], function: str) -> str: """ Get the caller function from the provided function """ is_next = False for c in callers: if is_next is True: return c if function == c: is_next = True def _get_args(self, *args) -> (Exception, str): """ Returns exception and message from the provided arguments """ ex = None msg = None for arg in args: if isinstance(arg, str): msg = arg elif isinstance(arg, Exception): ex = arg return ex, msg
synw/goerr
goerr/__init__.py
Err._err
python
def _err(self, errclass: str="error", *args) -> "Err": error = self._new_err(errclass, *args) if self.log_errs is True: sep = " " if self.log_format == "csv": sep = "," msg = str(datetime.now()) + sep + \ self._errmsg(error, msgformat=self.log_format) self.logger.error(msg) print(self._errmsg(error)) self._add(error) return error
Creates an error
train
https://github.com/synw/goerr/blob/08b3809d6715bffe26899a769d96fa5de8573faf/goerr/__init__.py#L99-L113
[ "def _new_err(self, errclass: str, *args) -> 'Err':\n \"\"\"\n Error constructor\n \"\"\"\n # get the message or exception\n ex, msg = self._get_args(*args)\n # construct the error\n # handle exception\n ftb = None # type: str\n function = None # type: str\n errtype = None # type: str\n file = None # type: str\n line = None # type: int\n code = None # type: str\n ex_msg = None # type: str\n caller = None # type: str\n caller_msg = None # type: str\n\n st = inspect.stack()\n\n if ex is not None:\n # get info from exception\n errobj, ex_msg, tb = sys.exc_info()\n tb = traceback.extract_tb(tb)\n file, line, function, code = tb[-1]\n # if called from an external lib\n if len(tb) > 1:\n file, line, caller, code = tb[0]\n else:\n call_stack = []\n for c in st:\n call_stack.append(c[3])\n caller = self._get_caller(call_stack, function)\n\n internals = [\n \"err\",\n \"_new_err\",\n \"fatal\",\n \"warning\",\n \"debug\",\n \"info\",\n \"<module>\"] \n if caller == function or caller in internals:\n caller = None\n # handle messages\n if msg is not None:\n caller_msg = msg\n msg = str(ex_msg)\n else:\n msg = str(ex_msg)\n ftb = traceback.format_exc()\n errtype = errobj.__name__\n if function is None:\n # for el in st:\n # print(el)\n function = st[3][3]\n if function == \"<module>\":\n function = None\n # init error object\n date = datetime.now()\n error = Err(\n function,\n date,\n msg,\n errtype,\n errclass,\n line,\n file,\n code,\n ftb,\n ex,\n caller,\n caller_msg)\n return error\n", "def _errmsg(self, error: \"Err\", tb: bool=False, i: int=None,\n msgformat: str=\"terminal\") -> str:\n \"\"\"\n Get the error message\n \"\"\"\n if msgformat == \"terminal\":\n msg = self._headline(error, i)\n if error.ex is not None:\n msg += \"\\n\" + \"line \" + colors.bold(str(error.line))\n msg += \": \" + colors.yellow(error.code)\n msg += \"\\n\" + str(error.file)\n if self.errs_traceback is True or tb is True:\n if error.tb is not None:\n msg += \"\\n\" + error.tb\n elif msgformat == \"csv\":\n sep = \",\"\n msg = error.msg + sep\n msg += str(error.line) + sep + error.code + sep\n msg += str(error.file)\n elif msgformat == \"text\":\n sep = \",\"\n msg = error.msg\n if error.ex is not None:\n msg += sep + str(error.line) + sep + error.code + sep\n msg += str(error.file) + sep \n if self.errs_traceback is True or tb is True:\n if error.tb is not None:\n msg += sep + error.tb\n elif msgformat == \"dict\":\n msg = {\"date\": datetime.now()}\n if error.ex is not None:\n msg[\"msg\"] = error.msg\n msg[\"line\"] = error.line\n msg[\"code\"] = error.code\n msg[\"file\"] = error.file \n if self.errs_traceback is True or tb is True:\n if error.tb is not None:\n msg[\"traceback\"] = error.tb\n return msg\n", "def _add(self, error: \"Err\"):\n \"\"\"\n Adds an error to the trace if required\n \"\"\"\n if self.trace_errs is True:\n self.errors.append(error)\n" ]
class Err(): """ Errors manager """ errors = [] # type: List[Err] trace_errs = False # type: bool errs_traceback = True # type: bool logger = None # type: logging.Logger log_errs = False # type: bool log_format = "csv" # type: str log_path = "errors.log" # type: str test_errs_mode = False # type: bool def __init__(self, function: str=None, date: datetime=datetime.now(), msg: str=None, errtype: str=None, errclass: str=None, line: int=None, file: str=None, code: str=None, tb: str=None, ex: Exception=None, caller: str=None, caller_msg: str=None): """ Datastructure of an error """ self.date = date # type: datetime.datetime self.function = function # type: str self.msg = msg # type: str self.errtype = errtype # type: str self.errclass = errclass # type: str self.line = line # type: int self.file = file # type: str self.code = code # type: str self.tb = tb # type: str self.ex = ex # type: Exception self.caller = caller # type: str self.caller_msg = caller_msg # type: str self.new = self.err def __repr__(self): msg = "<goerror.Err object: " + str(self.errclass) + " error>" return msg def __str__(self): return self.msg def err(self, *args): """ Creates an error """ error = self._err("error", *args) return error def panic(self, *args): """ Creates a fatal error and exit """ self._err("fatal", *args) if self.test_errs_mode is False: # pragma: no cover sys.exit(1) # pragma: no cover def warning(self, *args) -> "Err": """ Creates a warning message """ error = self._create_err("warning", *args) print(self._errmsg(error)) return error def info(self, *args) -> "Err": """ Creates an info message """ error = self._create_err("info", *args) print(self._errmsg(error)) return error def debug(self, *args) -> "Err": """ Creates a debug message """ error = self._create_err("debug", *args) print(self._errmsg(error)) return error def _create_err(self, errclass: str, *args) -> "Err": """ Create an error """ error = self._new_err(errclass, *args) self._add(error) return error def _new_err(self, errclass: str, *args) -> 'Err': """ Error constructor """ # get the message or exception ex, msg = self._get_args(*args) # construct the error # handle exception ftb = None # type: str function = None # type: str errtype = None # type: str file = None # type: str line = None # type: int code = None # type: str ex_msg = None # type: str caller = None # type: str caller_msg = None # type: str st = inspect.stack() if ex is not None: # get info from exception errobj, ex_msg, tb = sys.exc_info() tb = traceback.extract_tb(tb) file, line, function, code = tb[-1] # if called from an external lib if len(tb) > 1: file, line, caller, code = tb[0] else: call_stack = [] for c in st: call_stack.append(c[3]) caller = self._get_caller(call_stack, function) internals = [ "err", "_new_err", "fatal", "warning", "debug", "info", "<module>"] if caller == function or caller in internals: caller = None # handle messages if msg is not None: caller_msg = msg msg = str(ex_msg) else: msg = str(ex_msg) ftb = traceback.format_exc() errtype = errobj.__name__ if function is None: # for el in st: # print(el) function = st[3][3] if function == "<module>": function = None # init error object date = datetime.now() error = Err( function, date, msg, errtype, errclass, line, file, code, ftb, ex, caller, caller_msg) return error def _headline(self, error, i: int) -> str: """ Format the error message's headline """ msgs = Msg() # get the error title if error.errclass == "fatal": msg = msgs.fatal(i) elif error.errclass == "warning": msg = msgs.warning(i) elif error.errclass == "info": msg = msgs.info(i) elif error.errclass == "debug": msg = msgs.debug(i) elif error.errclass == "via": msg = msgs.via(i) else: msg = msgs.error(i) # function name if error.function is not None: msg += " from " + colors.bold(error.function) if error.caller is not None: msg += " called from " + colors.bold(error.caller) if error.caller_msg is not None: msg += "\n" + error.caller_msg if error.function is not None and error.msg is not None: msg += ": " else: msg = msg + " " if error.errtype is not None: msg += error.errtype + " : " if error.msg is not None: msg += error.msg return msg def _errmsg(self, error: "Err", tb: bool=False, i: int=None, msgformat: str="terminal") -> str: """ Get the error message """ if msgformat == "terminal": msg = self._headline(error, i) if error.ex is not None: msg += "\n" + "line " + colors.bold(str(error.line)) msg += ": " + colors.yellow(error.code) msg += "\n" + str(error.file) if self.errs_traceback is True or tb is True: if error.tb is not None: msg += "\n" + error.tb elif msgformat == "csv": sep = "," msg = error.msg + sep msg += str(error.line) + sep + error.code + sep msg += str(error.file) elif msgformat == "text": sep = "," msg = error.msg if error.ex is not None: msg += sep + str(error.line) + sep + error.code + sep msg += str(error.file) + sep if self.errs_traceback is True or tb is True: if error.tb is not None: msg += sep + error.tb elif msgformat == "dict": msg = {"date": datetime.now()} if error.ex is not None: msg["msg"] = error.msg msg["line"] = error.line msg["code"] = error.code msg["file"] = error.file if self.errs_traceback is True or tb is True: if error.tb is not None: msg["traceback"] = error.tb return msg def to_dict(self): """ Returns a dictionnary with the error elements """ return self._errmsg(self, msgformat="dict") def _print_errs(self): """ Prints the errors trace with tracebacks """ i = 0 for error in self.errors: print(self._errmsg(error, tb=True, i=i)) # for spacing if self.errs_traceback is False: print() i += 1 def _add(self, error: "Err"): """ Adds an error to the trace if required """ if self.trace_errs is True: self.errors.append(error) def _get_caller(self, callers: List[str], function: str) -> str: """ Get the caller function from the provided function """ is_next = False for c in callers: if is_next is True: return c if function == c: is_next = True def _get_args(self, *args) -> (Exception, str): """ Returns exception and message from the provided arguments """ ex = None msg = None for arg in args: if isinstance(arg, str): msg = arg elif isinstance(arg, Exception): ex = arg return ex, msg
synw/goerr
goerr/__init__.py
Err._new_err
python
def _new_err(self, errclass: str, *args) -> 'Err': # get the message or exception ex, msg = self._get_args(*args) # construct the error # handle exception ftb = None # type: str function = None # type: str errtype = None # type: str file = None # type: str line = None # type: int code = None # type: str ex_msg = None # type: str caller = None # type: str caller_msg = None # type: str st = inspect.stack() if ex is not None: # get info from exception errobj, ex_msg, tb = sys.exc_info() tb = traceback.extract_tb(tb) file, line, function, code = tb[-1] # if called from an external lib if len(tb) > 1: file, line, caller, code = tb[0] else: call_stack = [] for c in st: call_stack.append(c[3]) caller = self._get_caller(call_stack, function) internals = [ "err", "_new_err", "fatal", "warning", "debug", "info", "<module>"] if caller == function or caller in internals: caller = None # handle messages if msg is not None: caller_msg = msg msg = str(ex_msg) else: msg = str(ex_msg) ftb = traceback.format_exc() errtype = errobj.__name__ if function is None: # for el in st: # print(el) function = st[3][3] if function == "<module>": function = None # init error object date = datetime.now() error = Err( function, date, msg, errtype, errclass, line, file, code, ftb, ex, caller, caller_msg) return error
Error constructor
train
https://github.com/synw/goerr/blob/08b3809d6715bffe26899a769d96fa5de8573faf/goerr/__init__.py#L115-L188
null
class Err(): """ Errors manager """ errors = [] # type: List[Err] trace_errs = False # type: bool errs_traceback = True # type: bool logger = None # type: logging.Logger log_errs = False # type: bool log_format = "csv" # type: str log_path = "errors.log" # type: str test_errs_mode = False # type: bool def __init__(self, function: str=None, date: datetime=datetime.now(), msg: str=None, errtype: str=None, errclass: str=None, line: int=None, file: str=None, code: str=None, tb: str=None, ex: Exception=None, caller: str=None, caller_msg: str=None): """ Datastructure of an error """ self.date = date # type: datetime.datetime self.function = function # type: str self.msg = msg # type: str self.errtype = errtype # type: str self.errclass = errclass # type: str self.line = line # type: int self.file = file # type: str self.code = code # type: str self.tb = tb # type: str self.ex = ex # type: Exception self.caller = caller # type: str self.caller_msg = caller_msg # type: str self.new = self.err def __repr__(self): msg = "<goerror.Err object: " + str(self.errclass) + " error>" return msg def __str__(self): return self.msg def err(self, *args): """ Creates an error """ error = self._err("error", *args) return error def panic(self, *args): """ Creates a fatal error and exit """ self._err("fatal", *args) if self.test_errs_mode is False: # pragma: no cover sys.exit(1) # pragma: no cover def warning(self, *args) -> "Err": """ Creates a warning message """ error = self._create_err("warning", *args) print(self._errmsg(error)) return error def info(self, *args) -> "Err": """ Creates an info message """ error = self._create_err("info", *args) print(self._errmsg(error)) return error def debug(self, *args) -> "Err": """ Creates a debug message """ error = self._create_err("debug", *args) print(self._errmsg(error)) return error def _create_err(self, errclass: str, *args) -> "Err": """ Create an error """ error = self._new_err(errclass, *args) self._add(error) return error def _err(self, errclass: str="error", *args) -> "Err": """ Creates an error """ error = self._new_err(errclass, *args) if self.log_errs is True: sep = " " if self.log_format == "csv": sep = "," msg = str(datetime.now()) + sep + \ self._errmsg(error, msgformat=self.log_format) self.logger.error(msg) print(self._errmsg(error)) self._add(error) return error def _headline(self, error, i: int) -> str: """ Format the error message's headline """ msgs = Msg() # get the error title if error.errclass == "fatal": msg = msgs.fatal(i) elif error.errclass == "warning": msg = msgs.warning(i) elif error.errclass == "info": msg = msgs.info(i) elif error.errclass == "debug": msg = msgs.debug(i) elif error.errclass == "via": msg = msgs.via(i) else: msg = msgs.error(i) # function name if error.function is not None: msg += " from " + colors.bold(error.function) if error.caller is not None: msg += " called from " + colors.bold(error.caller) if error.caller_msg is not None: msg += "\n" + error.caller_msg if error.function is not None and error.msg is not None: msg += ": " else: msg = msg + " " if error.errtype is not None: msg += error.errtype + " : " if error.msg is not None: msg += error.msg return msg def _errmsg(self, error: "Err", tb: bool=False, i: int=None, msgformat: str="terminal") -> str: """ Get the error message """ if msgformat == "terminal": msg = self._headline(error, i) if error.ex is not None: msg += "\n" + "line " + colors.bold(str(error.line)) msg += ": " + colors.yellow(error.code) msg += "\n" + str(error.file) if self.errs_traceback is True or tb is True: if error.tb is not None: msg += "\n" + error.tb elif msgformat == "csv": sep = "," msg = error.msg + sep msg += str(error.line) + sep + error.code + sep msg += str(error.file) elif msgformat == "text": sep = "," msg = error.msg if error.ex is not None: msg += sep + str(error.line) + sep + error.code + sep msg += str(error.file) + sep if self.errs_traceback is True or tb is True: if error.tb is not None: msg += sep + error.tb elif msgformat == "dict": msg = {"date": datetime.now()} if error.ex is not None: msg["msg"] = error.msg msg["line"] = error.line msg["code"] = error.code msg["file"] = error.file if self.errs_traceback is True or tb is True: if error.tb is not None: msg["traceback"] = error.tb return msg def to_dict(self): """ Returns a dictionnary with the error elements """ return self._errmsg(self, msgformat="dict") def _print_errs(self): """ Prints the errors trace with tracebacks """ i = 0 for error in self.errors: print(self._errmsg(error, tb=True, i=i)) # for spacing if self.errs_traceback is False: print() i += 1 def _add(self, error: "Err"): """ Adds an error to the trace if required """ if self.trace_errs is True: self.errors.append(error) def _get_caller(self, callers: List[str], function: str) -> str: """ Get the caller function from the provided function """ is_next = False for c in callers: if is_next is True: return c if function == c: is_next = True def _get_args(self, *args) -> (Exception, str): """ Returns exception and message from the provided arguments """ ex = None msg = None for arg in args: if isinstance(arg, str): msg = arg elif isinstance(arg, Exception): ex = arg return ex, msg
synw/goerr
goerr/__init__.py
Err._headline
python
def _headline(self, error, i: int) -> str: msgs = Msg() # get the error title if error.errclass == "fatal": msg = msgs.fatal(i) elif error.errclass == "warning": msg = msgs.warning(i) elif error.errclass == "info": msg = msgs.info(i) elif error.errclass == "debug": msg = msgs.debug(i) elif error.errclass == "via": msg = msgs.via(i) else: msg = msgs.error(i) # function name if error.function is not None: msg += " from " + colors.bold(error.function) if error.caller is not None: msg += " called from " + colors.bold(error.caller) if error.caller_msg is not None: msg += "\n" + error.caller_msg if error.function is not None and error.msg is not None: msg += ": " else: msg = msg + " " if error.errtype is not None: msg += error.errtype + " : " if error.msg is not None: msg += error.msg return msg
Format the error message's headline
train
https://github.com/synw/goerr/blob/08b3809d6715bffe26899a769d96fa5de8573faf/goerr/__init__.py#L190-L223
null
class Err(): """ Errors manager """ errors = [] # type: List[Err] trace_errs = False # type: bool errs_traceback = True # type: bool logger = None # type: logging.Logger log_errs = False # type: bool log_format = "csv" # type: str log_path = "errors.log" # type: str test_errs_mode = False # type: bool def __init__(self, function: str=None, date: datetime=datetime.now(), msg: str=None, errtype: str=None, errclass: str=None, line: int=None, file: str=None, code: str=None, tb: str=None, ex: Exception=None, caller: str=None, caller_msg: str=None): """ Datastructure of an error """ self.date = date # type: datetime.datetime self.function = function # type: str self.msg = msg # type: str self.errtype = errtype # type: str self.errclass = errclass # type: str self.line = line # type: int self.file = file # type: str self.code = code # type: str self.tb = tb # type: str self.ex = ex # type: Exception self.caller = caller # type: str self.caller_msg = caller_msg # type: str self.new = self.err def __repr__(self): msg = "<goerror.Err object: " + str(self.errclass) + " error>" return msg def __str__(self): return self.msg def err(self, *args): """ Creates an error """ error = self._err("error", *args) return error def panic(self, *args): """ Creates a fatal error and exit """ self._err("fatal", *args) if self.test_errs_mode is False: # pragma: no cover sys.exit(1) # pragma: no cover def warning(self, *args) -> "Err": """ Creates a warning message """ error = self._create_err("warning", *args) print(self._errmsg(error)) return error def info(self, *args) -> "Err": """ Creates an info message """ error = self._create_err("info", *args) print(self._errmsg(error)) return error def debug(self, *args) -> "Err": """ Creates a debug message """ error = self._create_err("debug", *args) print(self._errmsg(error)) return error def _create_err(self, errclass: str, *args) -> "Err": """ Create an error """ error = self._new_err(errclass, *args) self._add(error) return error def _err(self, errclass: str="error", *args) -> "Err": """ Creates an error """ error = self._new_err(errclass, *args) if self.log_errs is True: sep = " " if self.log_format == "csv": sep = "," msg = str(datetime.now()) + sep + \ self._errmsg(error, msgformat=self.log_format) self.logger.error(msg) print(self._errmsg(error)) self._add(error) return error def _new_err(self, errclass: str, *args) -> 'Err': """ Error constructor """ # get the message or exception ex, msg = self._get_args(*args) # construct the error # handle exception ftb = None # type: str function = None # type: str errtype = None # type: str file = None # type: str line = None # type: int code = None # type: str ex_msg = None # type: str caller = None # type: str caller_msg = None # type: str st = inspect.stack() if ex is not None: # get info from exception errobj, ex_msg, tb = sys.exc_info() tb = traceback.extract_tb(tb) file, line, function, code = tb[-1] # if called from an external lib if len(tb) > 1: file, line, caller, code = tb[0] else: call_stack = [] for c in st: call_stack.append(c[3]) caller = self._get_caller(call_stack, function) internals = [ "err", "_new_err", "fatal", "warning", "debug", "info", "<module>"] if caller == function or caller in internals: caller = None # handle messages if msg is not None: caller_msg = msg msg = str(ex_msg) else: msg = str(ex_msg) ftb = traceback.format_exc() errtype = errobj.__name__ if function is None: # for el in st: # print(el) function = st[3][3] if function == "<module>": function = None # init error object date = datetime.now() error = Err( function, date, msg, errtype, errclass, line, file, code, ftb, ex, caller, caller_msg) return error def _errmsg(self, error: "Err", tb: bool=False, i: int=None, msgformat: str="terminal") -> str: """ Get the error message """ if msgformat == "terminal": msg = self._headline(error, i) if error.ex is not None: msg += "\n" + "line " + colors.bold(str(error.line)) msg += ": " + colors.yellow(error.code) msg += "\n" + str(error.file) if self.errs_traceback is True or tb is True: if error.tb is not None: msg += "\n" + error.tb elif msgformat == "csv": sep = "," msg = error.msg + sep msg += str(error.line) + sep + error.code + sep msg += str(error.file) elif msgformat == "text": sep = "," msg = error.msg if error.ex is not None: msg += sep + str(error.line) + sep + error.code + sep msg += str(error.file) + sep if self.errs_traceback is True or tb is True: if error.tb is not None: msg += sep + error.tb elif msgformat == "dict": msg = {"date": datetime.now()} if error.ex is not None: msg["msg"] = error.msg msg["line"] = error.line msg["code"] = error.code msg["file"] = error.file if self.errs_traceback is True or tb is True: if error.tb is not None: msg["traceback"] = error.tb return msg def to_dict(self): """ Returns a dictionnary with the error elements """ return self._errmsg(self, msgformat="dict") def _print_errs(self): """ Prints the errors trace with tracebacks """ i = 0 for error in self.errors: print(self._errmsg(error, tb=True, i=i)) # for spacing if self.errs_traceback is False: print() i += 1 def _add(self, error: "Err"): """ Adds an error to the trace if required """ if self.trace_errs is True: self.errors.append(error) def _get_caller(self, callers: List[str], function: str) -> str: """ Get the caller function from the provided function """ is_next = False for c in callers: if is_next is True: return c if function == c: is_next = True def _get_args(self, *args) -> (Exception, str): """ Returns exception and message from the provided arguments """ ex = None msg = None for arg in args: if isinstance(arg, str): msg = arg elif isinstance(arg, Exception): ex = arg return ex, msg
synw/goerr
goerr/__init__.py
Err._errmsg
python
def _errmsg(self, error: "Err", tb: bool=False, i: int=None, msgformat: str="terminal") -> str: if msgformat == "terminal": msg = self._headline(error, i) if error.ex is not None: msg += "\n" + "line " + colors.bold(str(error.line)) msg += ": " + colors.yellow(error.code) msg += "\n" + str(error.file) if self.errs_traceback is True or tb is True: if error.tb is not None: msg += "\n" + error.tb elif msgformat == "csv": sep = "," msg = error.msg + sep msg += str(error.line) + sep + error.code + sep msg += str(error.file) elif msgformat == "text": sep = "," msg = error.msg if error.ex is not None: msg += sep + str(error.line) + sep + error.code + sep msg += str(error.file) + sep if self.errs_traceback is True or tb is True: if error.tb is not None: msg += sep + error.tb elif msgformat == "dict": msg = {"date": datetime.now()} if error.ex is not None: msg["msg"] = error.msg msg["line"] = error.line msg["code"] = error.code msg["file"] = error.file if self.errs_traceback is True or tb is True: if error.tb is not None: msg["traceback"] = error.tb return msg
Get the error message
train
https://github.com/synw/goerr/blob/08b3809d6715bffe26899a769d96fa5de8573faf/goerr/__init__.py#L225-L263
[ "def _headline(self, error, i: int) -> str:\n \"\"\"\n Format the error message's headline\n \"\"\"\n msgs = Msg()\n # get the error title\n if error.errclass == \"fatal\":\n msg = msgs.fatal(i)\n elif error.errclass == \"warning\":\n msg = msgs.warning(i)\n elif error.errclass == \"info\":\n msg = msgs.info(i)\n elif error.errclass == \"debug\":\n msg = msgs.debug(i)\n elif error.errclass == \"via\":\n msg = msgs.via(i)\n else:\n msg = msgs.error(i)\n # function name\n if error.function is not None:\n msg += \" from \" + colors.bold(error.function)\n if error.caller is not None:\n msg += \" called from \" + colors.bold(error.caller)\n if error.caller_msg is not None:\n msg += \"\\n\" + error.caller_msg\n if error.function is not None and error.msg is not None:\n msg += \": \"\n else:\n msg = msg + \" \"\n if error.errtype is not None:\n msg += error.errtype + \" : \"\n if error.msg is not None:\n msg += error.msg\n return msg\n", "def yellow(self, *msg):\n color = '\\033[93m'\n return self._msg(color, *msg)\n", "def bold(self, *msg):\n color = '\\033[1m'\n return self._msg(color, *msg)\n" ]
class Err(): """ Errors manager """ errors = [] # type: List[Err] trace_errs = False # type: bool errs_traceback = True # type: bool logger = None # type: logging.Logger log_errs = False # type: bool log_format = "csv" # type: str log_path = "errors.log" # type: str test_errs_mode = False # type: bool def __init__(self, function: str=None, date: datetime=datetime.now(), msg: str=None, errtype: str=None, errclass: str=None, line: int=None, file: str=None, code: str=None, tb: str=None, ex: Exception=None, caller: str=None, caller_msg: str=None): """ Datastructure of an error """ self.date = date # type: datetime.datetime self.function = function # type: str self.msg = msg # type: str self.errtype = errtype # type: str self.errclass = errclass # type: str self.line = line # type: int self.file = file # type: str self.code = code # type: str self.tb = tb # type: str self.ex = ex # type: Exception self.caller = caller # type: str self.caller_msg = caller_msg # type: str self.new = self.err def __repr__(self): msg = "<goerror.Err object: " + str(self.errclass) + " error>" return msg def __str__(self): return self.msg def err(self, *args): """ Creates an error """ error = self._err("error", *args) return error def panic(self, *args): """ Creates a fatal error and exit """ self._err("fatal", *args) if self.test_errs_mode is False: # pragma: no cover sys.exit(1) # pragma: no cover def warning(self, *args) -> "Err": """ Creates a warning message """ error = self._create_err("warning", *args) print(self._errmsg(error)) return error def info(self, *args) -> "Err": """ Creates an info message """ error = self._create_err("info", *args) print(self._errmsg(error)) return error def debug(self, *args) -> "Err": """ Creates a debug message """ error = self._create_err("debug", *args) print(self._errmsg(error)) return error def _create_err(self, errclass: str, *args) -> "Err": """ Create an error """ error = self._new_err(errclass, *args) self._add(error) return error def _err(self, errclass: str="error", *args) -> "Err": """ Creates an error """ error = self._new_err(errclass, *args) if self.log_errs is True: sep = " " if self.log_format == "csv": sep = "," msg = str(datetime.now()) + sep + \ self._errmsg(error, msgformat=self.log_format) self.logger.error(msg) print(self._errmsg(error)) self._add(error) return error def _new_err(self, errclass: str, *args) -> 'Err': """ Error constructor """ # get the message or exception ex, msg = self._get_args(*args) # construct the error # handle exception ftb = None # type: str function = None # type: str errtype = None # type: str file = None # type: str line = None # type: int code = None # type: str ex_msg = None # type: str caller = None # type: str caller_msg = None # type: str st = inspect.stack() if ex is not None: # get info from exception errobj, ex_msg, tb = sys.exc_info() tb = traceback.extract_tb(tb) file, line, function, code = tb[-1] # if called from an external lib if len(tb) > 1: file, line, caller, code = tb[0] else: call_stack = [] for c in st: call_stack.append(c[3]) caller = self._get_caller(call_stack, function) internals = [ "err", "_new_err", "fatal", "warning", "debug", "info", "<module>"] if caller == function or caller in internals: caller = None # handle messages if msg is not None: caller_msg = msg msg = str(ex_msg) else: msg = str(ex_msg) ftb = traceback.format_exc() errtype = errobj.__name__ if function is None: # for el in st: # print(el) function = st[3][3] if function == "<module>": function = None # init error object date = datetime.now() error = Err( function, date, msg, errtype, errclass, line, file, code, ftb, ex, caller, caller_msg) return error def _headline(self, error, i: int) -> str: """ Format the error message's headline """ msgs = Msg() # get the error title if error.errclass == "fatal": msg = msgs.fatal(i) elif error.errclass == "warning": msg = msgs.warning(i) elif error.errclass == "info": msg = msgs.info(i) elif error.errclass == "debug": msg = msgs.debug(i) elif error.errclass == "via": msg = msgs.via(i) else: msg = msgs.error(i) # function name if error.function is not None: msg += " from " + colors.bold(error.function) if error.caller is not None: msg += " called from " + colors.bold(error.caller) if error.caller_msg is not None: msg += "\n" + error.caller_msg if error.function is not None and error.msg is not None: msg += ": " else: msg = msg + " " if error.errtype is not None: msg += error.errtype + " : " if error.msg is not None: msg += error.msg return msg def to_dict(self): """ Returns a dictionnary with the error elements """ return self._errmsg(self, msgformat="dict") def _print_errs(self): """ Prints the errors trace with tracebacks """ i = 0 for error in self.errors: print(self._errmsg(error, tb=True, i=i)) # for spacing if self.errs_traceback is False: print() i += 1 def _add(self, error: "Err"): """ Adds an error to the trace if required """ if self.trace_errs is True: self.errors.append(error) def _get_caller(self, callers: List[str], function: str) -> str: """ Get the caller function from the provided function """ is_next = False for c in callers: if is_next is True: return c if function == c: is_next = True def _get_args(self, *args) -> (Exception, str): """ Returns exception and message from the provided arguments """ ex = None msg = None for arg in args: if isinstance(arg, str): msg = arg elif isinstance(arg, Exception): ex = arg return ex, msg
synw/goerr
goerr/__init__.py
Err._print_errs
python
def _print_errs(self): i = 0 for error in self.errors: print(self._errmsg(error, tb=True, i=i)) # for spacing if self.errs_traceback is False: print() i += 1
Prints the errors trace with tracebacks
train
https://github.com/synw/goerr/blob/08b3809d6715bffe26899a769d96fa5de8573faf/goerr/__init__.py#L271-L281
[ "def _errmsg(self, error: \"Err\", tb: bool=False, i: int=None,\n msgformat: str=\"terminal\") -> str:\n \"\"\"\n Get the error message\n \"\"\"\n if msgformat == \"terminal\":\n msg = self._headline(error, i)\n if error.ex is not None:\n msg += \"\\n\" + \"line \" + colors.bold(str(error.line))\n msg += \": \" + colors.yellow(error.code)\n msg += \"\\n\" + str(error.file)\n if self.errs_traceback is True or tb is True:\n if error.tb is not None:\n msg += \"\\n\" + error.tb\n elif msgformat == \"csv\":\n sep = \",\"\n msg = error.msg + sep\n msg += str(error.line) + sep + error.code + sep\n msg += str(error.file)\n elif msgformat == \"text\":\n sep = \",\"\n msg = error.msg\n if error.ex is not None:\n msg += sep + str(error.line) + sep + error.code + sep\n msg += str(error.file) + sep \n if self.errs_traceback is True or tb is True:\n if error.tb is not None:\n msg += sep + error.tb\n elif msgformat == \"dict\":\n msg = {\"date\": datetime.now()}\n if error.ex is not None:\n msg[\"msg\"] = error.msg\n msg[\"line\"] = error.line\n msg[\"code\"] = error.code\n msg[\"file\"] = error.file \n if self.errs_traceback is True or tb is True:\n if error.tb is not None:\n msg[\"traceback\"] = error.tb\n return msg\n" ]
class Err(): """ Errors manager """ errors = [] # type: List[Err] trace_errs = False # type: bool errs_traceback = True # type: bool logger = None # type: logging.Logger log_errs = False # type: bool log_format = "csv" # type: str log_path = "errors.log" # type: str test_errs_mode = False # type: bool def __init__(self, function: str=None, date: datetime=datetime.now(), msg: str=None, errtype: str=None, errclass: str=None, line: int=None, file: str=None, code: str=None, tb: str=None, ex: Exception=None, caller: str=None, caller_msg: str=None): """ Datastructure of an error """ self.date = date # type: datetime.datetime self.function = function # type: str self.msg = msg # type: str self.errtype = errtype # type: str self.errclass = errclass # type: str self.line = line # type: int self.file = file # type: str self.code = code # type: str self.tb = tb # type: str self.ex = ex # type: Exception self.caller = caller # type: str self.caller_msg = caller_msg # type: str self.new = self.err def __repr__(self): msg = "<goerror.Err object: " + str(self.errclass) + " error>" return msg def __str__(self): return self.msg def err(self, *args): """ Creates an error """ error = self._err("error", *args) return error def panic(self, *args): """ Creates a fatal error and exit """ self._err("fatal", *args) if self.test_errs_mode is False: # pragma: no cover sys.exit(1) # pragma: no cover def warning(self, *args) -> "Err": """ Creates a warning message """ error = self._create_err("warning", *args) print(self._errmsg(error)) return error def info(self, *args) -> "Err": """ Creates an info message """ error = self._create_err("info", *args) print(self._errmsg(error)) return error def debug(self, *args) -> "Err": """ Creates a debug message """ error = self._create_err("debug", *args) print(self._errmsg(error)) return error def _create_err(self, errclass: str, *args) -> "Err": """ Create an error """ error = self._new_err(errclass, *args) self._add(error) return error def _err(self, errclass: str="error", *args) -> "Err": """ Creates an error """ error = self._new_err(errclass, *args) if self.log_errs is True: sep = " " if self.log_format == "csv": sep = "," msg = str(datetime.now()) + sep + \ self._errmsg(error, msgformat=self.log_format) self.logger.error(msg) print(self._errmsg(error)) self._add(error) return error def _new_err(self, errclass: str, *args) -> 'Err': """ Error constructor """ # get the message or exception ex, msg = self._get_args(*args) # construct the error # handle exception ftb = None # type: str function = None # type: str errtype = None # type: str file = None # type: str line = None # type: int code = None # type: str ex_msg = None # type: str caller = None # type: str caller_msg = None # type: str st = inspect.stack() if ex is not None: # get info from exception errobj, ex_msg, tb = sys.exc_info() tb = traceback.extract_tb(tb) file, line, function, code = tb[-1] # if called from an external lib if len(tb) > 1: file, line, caller, code = tb[0] else: call_stack = [] for c in st: call_stack.append(c[3]) caller = self._get_caller(call_stack, function) internals = [ "err", "_new_err", "fatal", "warning", "debug", "info", "<module>"] if caller == function or caller in internals: caller = None # handle messages if msg is not None: caller_msg = msg msg = str(ex_msg) else: msg = str(ex_msg) ftb = traceback.format_exc() errtype = errobj.__name__ if function is None: # for el in st: # print(el) function = st[3][3] if function == "<module>": function = None # init error object date = datetime.now() error = Err( function, date, msg, errtype, errclass, line, file, code, ftb, ex, caller, caller_msg) return error def _headline(self, error, i: int) -> str: """ Format the error message's headline """ msgs = Msg() # get the error title if error.errclass == "fatal": msg = msgs.fatal(i) elif error.errclass == "warning": msg = msgs.warning(i) elif error.errclass == "info": msg = msgs.info(i) elif error.errclass == "debug": msg = msgs.debug(i) elif error.errclass == "via": msg = msgs.via(i) else: msg = msgs.error(i) # function name if error.function is not None: msg += " from " + colors.bold(error.function) if error.caller is not None: msg += " called from " + colors.bold(error.caller) if error.caller_msg is not None: msg += "\n" + error.caller_msg if error.function is not None and error.msg is not None: msg += ": " else: msg = msg + " " if error.errtype is not None: msg += error.errtype + " : " if error.msg is not None: msg += error.msg return msg def _errmsg(self, error: "Err", tb: bool=False, i: int=None, msgformat: str="terminal") -> str: """ Get the error message """ if msgformat == "terminal": msg = self._headline(error, i) if error.ex is not None: msg += "\n" + "line " + colors.bold(str(error.line)) msg += ": " + colors.yellow(error.code) msg += "\n" + str(error.file) if self.errs_traceback is True or tb is True: if error.tb is not None: msg += "\n" + error.tb elif msgformat == "csv": sep = "," msg = error.msg + sep msg += str(error.line) + sep + error.code + sep msg += str(error.file) elif msgformat == "text": sep = "," msg = error.msg if error.ex is not None: msg += sep + str(error.line) + sep + error.code + sep msg += str(error.file) + sep if self.errs_traceback is True or tb is True: if error.tb is not None: msg += sep + error.tb elif msgformat == "dict": msg = {"date": datetime.now()} if error.ex is not None: msg["msg"] = error.msg msg["line"] = error.line msg["code"] = error.code msg["file"] = error.file if self.errs_traceback is True or tb is True: if error.tb is not None: msg["traceback"] = error.tb return msg def to_dict(self): """ Returns a dictionnary with the error elements """ return self._errmsg(self, msgformat="dict") def _add(self, error: "Err"): """ Adds an error to the trace if required """ if self.trace_errs is True: self.errors.append(error) def _get_caller(self, callers: List[str], function: str) -> str: """ Get the caller function from the provided function """ is_next = False for c in callers: if is_next is True: return c if function == c: is_next = True def _get_args(self, *args) -> (Exception, str): """ Returns exception and message from the provided arguments """ ex = None msg = None for arg in args: if isinstance(arg, str): msg = arg elif isinstance(arg, Exception): ex = arg return ex, msg
synw/goerr
goerr/__init__.py
Err._add
python
def _add(self, error: "Err"): if self.trace_errs is True: self.errors.append(error)
Adds an error to the trace if required
train
https://github.com/synw/goerr/blob/08b3809d6715bffe26899a769d96fa5de8573faf/goerr/__init__.py#L283-L288
null
class Err(): """ Errors manager """ errors = [] # type: List[Err] trace_errs = False # type: bool errs_traceback = True # type: bool logger = None # type: logging.Logger log_errs = False # type: bool log_format = "csv" # type: str log_path = "errors.log" # type: str test_errs_mode = False # type: bool def __init__(self, function: str=None, date: datetime=datetime.now(), msg: str=None, errtype: str=None, errclass: str=None, line: int=None, file: str=None, code: str=None, tb: str=None, ex: Exception=None, caller: str=None, caller_msg: str=None): """ Datastructure of an error """ self.date = date # type: datetime.datetime self.function = function # type: str self.msg = msg # type: str self.errtype = errtype # type: str self.errclass = errclass # type: str self.line = line # type: int self.file = file # type: str self.code = code # type: str self.tb = tb # type: str self.ex = ex # type: Exception self.caller = caller # type: str self.caller_msg = caller_msg # type: str self.new = self.err def __repr__(self): msg = "<goerror.Err object: " + str(self.errclass) + " error>" return msg def __str__(self): return self.msg def err(self, *args): """ Creates an error """ error = self._err("error", *args) return error def panic(self, *args): """ Creates a fatal error and exit """ self._err("fatal", *args) if self.test_errs_mode is False: # pragma: no cover sys.exit(1) # pragma: no cover def warning(self, *args) -> "Err": """ Creates a warning message """ error = self._create_err("warning", *args) print(self._errmsg(error)) return error def info(self, *args) -> "Err": """ Creates an info message """ error = self._create_err("info", *args) print(self._errmsg(error)) return error def debug(self, *args) -> "Err": """ Creates a debug message """ error = self._create_err("debug", *args) print(self._errmsg(error)) return error def _create_err(self, errclass: str, *args) -> "Err": """ Create an error """ error = self._new_err(errclass, *args) self._add(error) return error def _err(self, errclass: str="error", *args) -> "Err": """ Creates an error """ error = self._new_err(errclass, *args) if self.log_errs is True: sep = " " if self.log_format == "csv": sep = "," msg = str(datetime.now()) + sep + \ self._errmsg(error, msgformat=self.log_format) self.logger.error(msg) print(self._errmsg(error)) self._add(error) return error def _new_err(self, errclass: str, *args) -> 'Err': """ Error constructor """ # get the message or exception ex, msg = self._get_args(*args) # construct the error # handle exception ftb = None # type: str function = None # type: str errtype = None # type: str file = None # type: str line = None # type: int code = None # type: str ex_msg = None # type: str caller = None # type: str caller_msg = None # type: str st = inspect.stack() if ex is not None: # get info from exception errobj, ex_msg, tb = sys.exc_info() tb = traceback.extract_tb(tb) file, line, function, code = tb[-1] # if called from an external lib if len(tb) > 1: file, line, caller, code = tb[0] else: call_stack = [] for c in st: call_stack.append(c[3]) caller = self._get_caller(call_stack, function) internals = [ "err", "_new_err", "fatal", "warning", "debug", "info", "<module>"] if caller == function or caller in internals: caller = None # handle messages if msg is not None: caller_msg = msg msg = str(ex_msg) else: msg = str(ex_msg) ftb = traceback.format_exc() errtype = errobj.__name__ if function is None: # for el in st: # print(el) function = st[3][3] if function == "<module>": function = None # init error object date = datetime.now() error = Err( function, date, msg, errtype, errclass, line, file, code, ftb, ex, caller, caller_msg) return error def _headline(self, error, i: int) -> str: """ Format the error message's headline """ msgs = Msg() # get the error title if error.errclass == "fatal": msg = msgs.fatal(i) elif error.errclass == "warning": msg = msgs.warning(i) elif error.errclass == "info": msg = msgs.info(i) elif error.errclass == "debug": msg = msgs.debug(i) elif error.errclass == "via": msg = msgs.via(i) else: msg = msgs.error(i) # function name if error.function is not None: msg += " from " + colors.bold(error.function) if error.caller is not None: msg += " called from " + colors.bold(error.caller) if error.caller_msg is not None: msg += "\n" + error.caller_msg if error.function is not None and error.msg is not None: msg += ": " else: msg = msg + " " if error.errtype is not None: msg += error.errtype + " : " if error.msg is not None: msg += error.msg return msg def _errmsg(self, error: "Err", tb: bool=False, i: int=None, msgformat: str="terminal") -> str: """ Get the error message """ if msgformat == "terminal": msg = self._headline(error, i) if error.ex is not None: msg += "\n" + "line " + colors.bold(str(error.line)) msg += ": " + colors.yellow(error.code) msg += "\n" + str(error.file) if self.errs_traceback is True or tb is True: if error.tb is not None: msg += "\n" + error.tb elif msgformat == "csv": sep = "," msg = error.msg + sep msg += str(error.line) + sep + error.code + sep msg += str(error.file) elif msgformat == "text": sep = "," msg = error.msg if error.ex is not None: msg += sep + str(error.line) + sep + error.code + sep msg += str(error.file) + sep if self.errs_traceback is True or tb is True: if error.tb is not None: msg += sep + error.tb elif msgformat == "dict": msg = {"date": datetime.now()} if error.ex is not None: msg["msg"] = error.msg msg["line"] = error.line msg["code"] = error.code msg["file"] = error.file if self.errs_traceback is True or tb is True: if error.tb is not None: msg["traceback"] = error.tb return msg def to_dict(self): """ Returns a dictionnary with the error elements """ return self._errmsg(self, msgformat="dict") def _print_errs(self): """ Prints the errors trace with tracebacks """ i = 0 for error in self.errors: print(self._errmsg(error, tb=True, i=i)) # for spacing if self.errs_traceback is False: print() i += 1 def _get_caller(self, callers: List[str], function: str) -> str: """ Get the caller function from the provided function """ is_next = False for c in callers: if is_next is True: return c if function == c: is_next = True def _get_args(self, *args) -> (Exception, str): """ Returns exception and message from the provided arguments """ ex = None msg = None for arg in args: if isinstance(arg, str): msg = arg elif isinstance(arg, Exception): ex = arg return ex, msg
synw/goerr
goerr/__init__.py
Err._get_caller
python
def _get_caller(self, callers: List[str], function: str) -> str: is_next = False for c in callers: if is_next is True: return c if function == c: is_next = True
Get the caller function from the provided function
train
https://github.com/synw/goerr/blob/08b3809d6715bffe26899a769d96fa5de8573faf/goerr/__init__.py#L290-L299
null
class Err(): """ Errors manager """ errors = [] # type: List[Err] trace_errs = False # type: bool errs_traceback = True # type: bool logger = None # type: logging.Logger log_errs = False # type: bool log_format = "csv" # type: str log_path = "errors.log" # type: str test_errs_mode = False # type: bool def __init__(self, function: str=None, date: datetime=datetime.now(), msg: str=None, errtype: str=None, errclass: str=None, line: int=None, file: str=None, code: str=None, tb: str=None, ex: Exception=None, caller: str=None, caller_msg: str=None): """ Datastructure of an error """ self.date = date # type: datetime.datetime self.function = function # type: str self.msg = msg # type: str self.errtype = errtype # type: str self.errclass = errclass # type: str self.line = line # type: int self.file = file # type: str self.code = code # type: str self.tb = tb # type: str self.ex = ex # type: Exception self.caller = caller # type: str self.caller_msg = caller_msg # type: str self.new = self.err def __repr__(self): msg = "<goerror.Err object: " + str(self.errclass) + " error>" return msg def __str__(self): return self.msg def err(self, *args): """ Creates an error """ error = self._err("error", *args) return error def panic(self, *args): """ Creates a fatal error and exit """ self._err("fatal", *args) if self.test_errs_mode is False: # pragma: no cover sys.exit(1) # pragma: no cover def warning(self, *args) -> "Err": """ Creates a warning message """ error = self._create_err("warning", *args) print(self._errmsg(error)) return error def info(self, *args) -> "Err": """ Creates an info message """ error = self._create_err("info", *args) print(self._errmsg(error)) return error def debug(self, *args) -> "Err": """ Creates a debug message """ error = self._create_err("debug", *args) print(self._errmsg(error)) return error def _create_err(self, errclass: str, *args) -> "Err": """ Create an error """ error = self._new_err(errclass, *args) self._add(error) return error def _err(self, errclass: str="error", *args) -> "Err": """ Creates an error """ error = self._new_err(errclass, *args) if self.log_errs is True: sep = " " if self.log_format == "csv": sep = "," msg = str(datetime.now()) + sep + \ self._errmsg(error, msgformat=self.log_format) self.logger.error(msg) print(self._errmsg(error)) self._add(error) return error def _new_err(self, errclass: str, *args) -> 'Err': """ Error constructor """ # get the message or exception ex, msg = self._get_args(*args) # construct the error # handle exception ftb = None # type: str function = None # type: str errtype = None # type: str file = None # type: str line = None # type: int code = None # type: str ex_msg = None # type: str caller = None # type: str caller_msg = None # type: str st = inspect.stack() if ex is not None: # get info from exception errobj, ex_msg, tb = sys.exc_info() tb = traceback.extract_tb(tb) file, line, function, code = tb[-1] # if called from an external lib if len(tb) > 1: file, line, caller, code = tb[0] else: call_stack = [] for c in st: call_stack.append(c[3]) caller = self._get_caller(call_stack, function) internals = [ "err", "_new_err", "fatal", "warning", "debug", "info", "<module>"] if caller == function or caller in internals: caller = None # handle messages if msg is not None: caller_msg = msg msg = str(ex_msg) else: msg = str(ex_msg) ftb = traceback.format_exc() errtype = errobj.__name__ if function is None: # for el in st: # print(el) function = st[3][3] if function == "<module>": function = None # init error object date = datetime.now() error = Err( function, date, msg, errtype, errclass, line, file, code, ftb, ex, caller, caller_msg) return error def _headline(self, error, i: int) -> str: """ Format the error message's headline """ msgs = Msg() # get the error title if error.errclass == "fatal": msg = msgs.fatal(i) elif error.errclass == "warning": msg = msgs.warning(i) elif error.errclass == "info": msg = msgs.info(i) elif error.errclass == "debug": msg = msgs.debug(i) elif error.errclass == "via": msg = msgs.via(i) else: msg = msgs.error(i) # function name if error.function is not None: msg += " from " + colors.bold(error.function) if error.caller is not None: msg += " called from " + colors.bold(error.caller) if error.caller_msg is not None: msg += "\n" + error.caller_msg if error.function is not None and error.msg is not None: msg += ": " else: msg = msg + " " if error.errtype is not None: msg += error.errtype + " : " if error.msg is not None: msg += error.msg return msg def _errmsg(self, error: "Err", tb: bool=False, i: int=None, msgformat: str="terminal") -> str: """ Get the error message """ if msgformat == "terminal": msg = self._headline(error, i) if error.ex is not None: msg += "\n" + "line " + colors.bold(str(error.line)) msg += ": " + colors.yellow(error.code) msg += "\n" + str(error.file) if self.errs_traceback is True or tb is True: if error.tb is not None: msg += "\n" + error.tb elif msgformat == "csv": sep = "," msg = error.msg + sep msg += str(error.line) + sep + error.code + sep msg += str(error.file) elif msgformat == "text": sep = "," msg = error.msg if error.ex is not None: msg += sep + str(error.line) + sep + error.code + sep msg += str(error.file) + sep if self.errs_traceback is True or tb is True: if error.tb is not None: msg += sep + error.tb elif msgformat == "dict": msg = {"date": datetime.now()} if error.ex is not None: msg["msg"] = error.msg msg["line"] = error.line msg["code"] = error.code msg["file"] = error.file if self.errs_traceback is True or tb is True: if error.tb is not None: msg["traceback"] = error.tb return msg def to_dict(self): """ Returns a dictionnary with the error elements """ return self._errmsg(self, msgformat="dict") def _print_errs(self): """ Prints the errors trace with tracebacks """ i = 0 for error in self.errors: print(self._errmsg(error, tb=True, i=i)) # for spacing if self.errs_traceback is False: print() i += 1 def _add(self, error: "Err"): """ Adds an error to the trace if required """ if self.trace_errs is True: self.errors.append(error) def _get_args(self, *args) -> (Exception, str): """ Returns exception and message from the provided arguments """ ex = None msg = None for arg in args: if isinstance(arg, str): msg = arg elif isinstance(arg, Exception): ex = arg return ex, msg
synw/goerr
goerr/__init__.py
Err._get_args
python
def _get_args(self, *args) -> (Exception, str): ex = None msg = None for arg in args: if isinstance(arg, str): msg = arg elif isinstance(arg, Exception): ex = arg return ex, msg
Returns exception and message from the provided arguments
train
https://github.com/synw/goerr/blob/08b3809d6715bffe26899a769d96fa5de8573faf/goerr/__init__.py#L301-L312
null
class Err(): """ Errors manager """ errors = [] # type: List[Err] trace_errs = False # type: bool errs_traceback = True # type: bool logger = None # type: logging.Logger log_errs = False # type: bool log_format = "csv" # type: str log_path = "errors.log" # type: str test_errs_mode = False # type: bool def __init__(self, function: str=None, date: datetime=datetime.now(), msg: str=None, errtype: str=None, errclass: str=None, line: int=None, file: str=None, code: str=None, tb: str=None, ex: Exception=None, caller: str=None, caller_msg: str=None): """ Datastructure of an error """ self.date = date # type: datetime.datetime self.function = function # type: str self.msg = msg # type: str self.errtype = errtype # type: str self.errclass = errclass # type: str self.line = line # type: int self.file = file # type: str self.code = code # type: str self.tb = tb # type: str self.ex = ex # type: Exception self.caller = caller # type: str self.caller_msg = caller_msg # type: str self.new = self.err def __repr__(self): msg = "<goerror.Err object: " + str(self.errclass) + " error>" return msg def __str__(self): return self.msg def err(self, *args): """ Creates an error """ error = self._err("error", *args) return error def panic(self, *args): """ Creates a fatal error and exit """ self._err("fatal", *args) if self.test_errs_mode is False: # pragma: no cover sys.exit(1) # pragma: no cover def warning(self, *args) -> "Err": """ Creates a warning message """ error = self._create_err("warning", *args) print(self._errmsg(error)) return error def info(self, *args) -> "Err": """ Creates an info message """ error = self._create_err("info", *args) print(self._errmsg(error)) return error def debug(self, *args) -> "Err": """ Creates a debug message """ error = self._create_err("debug", *args) print(self._errmsg(error)) return error def _create_err(self, errclass: str, *args) -> "Err": """ Create an error """ error = self._new_err(errclass, *args) self._add(error) return error def _err(self, errclass: str="error", *args) -> "Err": """ Creates an error """ error = self._new_err(errclass, *args) if self.log_errs is True: sep = " " if self.log_format == "csv": sep = "," msg = str(datetime.now()) + sep + \ self._errmsg(error, msgformat=self.log_format) self.logger.error(msg) print(self._errmsg(error)) self._add(error) return error def _new_err(self, errclass: str, *args) -> 'Err': """ Error constructor """ # get the message or exception ex, msg = self._get_args(*args) # construct the error # handle exception ftb = None # type: str function = None # type: str errtype = None # type: str file = None # type: str line = None # type: int code = None # type: str ex_msg = None # type: str caller = None # type: str caller_msg = None # type: str st = inspect.stack() if ex is not None: # get info from exception errobj, ex_msg, tb = sys.exc_info() tb = traceback.extract_tb(tb) file, line, function, code = tb[-1] # if called from an external lib if len(tb) > 1: file, line, caller, code = tb[0] else: call_stack = [] for c in st: call_stack.append(c[3]) caller = self._get_caller(call_stack, function) internals = [ "err", "_new_err", "fatal", "warning", "debug", "info", "<module>"] if caller == function or caller in internals: caller = None # handle messages if msg is not None: caller_msg = msg msg = str(ex_msg) else: msg = str(ex_msg) ftb = traceback.format_exc() errtype = errobj.__name__ if function is None: # for el in st: # print(el) function = st[3][3] if function == "<module>": function = None # init error object date = datetime.now() error = Err( function, date, msg, errtype, errclass, line, file, code, ftb, ex, caller, caller_msg) return error def _headline(self, error, i: int) -> str: """ Format the error message's headline """ msgs = Msg() # get the error title if error.errclass == "fatal": msg = msgs.fatal(i) elif error.errclass == "warning": msg = msgs.warning(i) elif error.errclass == "info": msg = msgs.info(i) elif error.errclass == "debug": msg = msgs.debug(i) elif error.errclass == "via": msg = msgs.via(i) else: msg = msgs.error(i) # function name if error.function is not None: msg += " from " + colors.bold(error.function) if error.caller is not None: msg += " called from " + colors.bold(error.caller) if error.caller_msg is not None: msg += "\n" + error.caller_msg if error.function is not None and error.msg is not None: msg += ": " else: msg = msg + " " if error.errtype is not None: msg += error.errtype + " : " if error.msg is not None: msg += error.msg return msg def _errmsg(self, error: "Err", tb: bool=False, i: int=None, msgformat: str="terminal") -> str: """ Get the error message """ if msgformat == "terminal": msg = self._headline(error, i) if error.ex is not None: msg += "\n" + "line " + colors.bold(str(error.line)) msg += ": " + colors.yellow(error.code) msg += "\n" + str(error.file) if self.errs_traceback is True or tb is True: if error.tb is not None: msg += "\n" + error.tb elif msgformat == "csv": sep = "," msg = error.msg + sep msg += str(error.line) + sep + error.code + sep msg += str(error.file) elif msgformat == "text": sep = "," msg = error.msg if error.ex is not None: msg += sep + str(error.line) + sep + error.code + sep msg += str(error.file) + sep if self.errs_traceback is True or tb is True: if error.tb is not None: msg += sep + error.tb elif msgformat == "dict": msg = {"date": datetime.now()} if error.ex is not None: msg["msg"] = error.msg msg["line"] = error.line msg["code"] = error.code msg["file"] = error.file if self.errs_traceback is True or tb is True: if error.tb is not None: msg["traceback"] = error.tb return msg def to_dict(self): """ Returns a dictionnary with the error elements """ return self._errmsg(self, msgformat="dict") def _print_errs(self): """ Prints the errors trace with tracebacks """ i = 0 for error in self.errors: print(self._errmsg(error, tb=True, i=i)) # for spacing if self.errs_traceback is False: print() i += 1 def _add(self, error: "Err"): """ Adds an error to the trace if required """ if self.trace_errs is True: self.errors.append(error) def _get_caller(self, callers: List[str], function: str) -> str: """ Get the caller function from the provided function """ is_next = False for c in callers: if is_next is True: return c if function == c: is_next = True
synw/goerr
goerr/__init__.py
Trace.trace
python
def trace(self): if len(self.errors) > 0: numerrs = len(self.errors) print("========= Trace (" + str(numerrs) + ") =========") self._print_errs() self.errors = []
Print the errors trace if there are some errors
train
https://github.com/synw/goerr/blob/08b3809d6715bffe26899a769d96fa5de8573faf/goerr/__init__.py#L331-L339
[ "def _print_errs(self):\n \"\"\"\n Prints the errors trace with tracebacks\n \"\"\"\n i = 0\n for error in self.errors:\n print(self._errmsg(error, tb=True, i=i))\n # for spacing\n if self.errs_traceback is False:\n print()\n i += 1\n" ]
class Trace(Err): """ Tracess manager """ errors = [] # type: List[Err] trace_errs = True # type: bool errs_traceback = False # type: bool def __repr__(self): s = "s" numerrs = len(self.errors) if numerrs == 1: s = "" msg = "<goerror.Trace object: " + str(numerrs) + " error" + s + ">" return msg def via(self, *args): """ Creates an empty error to record in the stack trace """ error = None if len(self.errors) > 0: error = self._err("via", *args) return error def panic(self, *args): """ Creates a fatal error and exit """ self._err(*args) self.trace() if self.test_errs_mode is False: # pragma: no cover sys.exit(1)
synw/goerr
goerr/__init__.py
Trace.via
python
def via(self, *args): error = None if len(self.errors) > 0: error = self._err("via", *args) return error
Creates an empty error to record in the stack trace
train
https://github.com/synw/goerr/blob/08b3809d6715bffe26899a769d96fa5de8573faf/goerr/__init__.py#L341-L349
[ "def _err(self, errclass: str=\"error\", *args) -> \"Err\":\n \"\"\"\n Creates an error\n \"\"\"\n error = self._new_err(errclass, *args)\n if self.log_errs is True:\n sep = \" \"\n if self.log_format == \"csv\":\n sep = \",\"\n msg = str(datetime.now()) + sep + \\\n self._errmsg(error, msgformat=self.log_format)\n self.logger.error(msg)\n print(self._errmsg(error))\n self._add(error)\n return error\n" ]
class Trace(Err): """ Tracess manager """ errors = [] # type: List[Err] trace_errs = True # type: bool errs_traceback = False # type: bool def __repr__(self): s = "s" numerrs = len(self.errors) if numerrs == 1: s = "" msg = "<goerror.Trace object: " + str(numerrs) + " error" + s + ">" return msg def trace(self): """ Print the errors trace if there are some errors """ if len(self.errors) > 0: numerrs = len(self.errors) print("========= Trace (" + str(numerrs) + ") =========") self._print_errs() self.errors = [] def panic(self, *args): """ Creates a fatal error and exit """ self._err(*args) self.trace() if self.test_errs_mode is False: # pragma: no cover sys.exit(1)
jeremylow/pyshk
pyshk/api.py
Api.get_user
python
def get_user(self, user_id=None, user_name=None): if user_id: endpoint = '/api/user_id/{0}'.format(user_id) elif user_name: endpoint = '/api/user_name/{0}'.format(user_name) else: # Return currently authorized user endpoint = '/api/user' data = self._make_request(verb="GET", endpoint=endpoint) try: return User.NewFromJSON(data) except: return data
Get a user object from the API. If no ``user_id`` or ``user_name`` is specified, it will return the User object for the currently authenticated user. Args: user_id (int): User ID of the user for whom you want to get information. [Optional] user_name(str): Username for the user for whom you want to get information. [Optional] Returns: A User object.
train
https://github.com/jeremylow/pyshk/blob/3ab92f6706397cde7a18367266eba9e0f1ada868/pyshk/api.py#L267-L295
[ "def _make_request(self, verb, endpoint=None, data=None, files=None):\n if not self.authenticated:\n raise ApiInstanceUnauthorized\n\n resource_url = self._get_url_endpoint(endpoint)\n\n timestamp = int(time.mktime(datetime.datetime.utcnow().timetuple()))\n nonce = self.get_nonce()\n\n authorization_header = self._make_headers(\n verb=verb,\n endpoint=endpoint,\n nonce=nonce,\n timestamp=timestamp)\n\n if verb == \"GET\":\n req = requests.get(\n resource_url,\n headers={'Authorization': authorization_header},\n verify=False)\n elif verb == \"POST\":\n if data:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n data=data)\n elif files:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n files=files)\n elif data and files:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n files=files,\n data=data)\n else:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header})\n\n if req.status_code == 401:\n raise ApiResponseUnauthorized(req)\n elif req.status_code == 404:\n raise NotFound404(req)\n elif req.status_code == 500:\n raise Exception(req)\n\n if self.testing:\n return req\n\n try:\n return req.json()\n except:\n print('returning req', req._content)\n return req\n", "def NewFromJSON(data):\n \"\"\"\n Create a new User instance from a JSON dict.\n\n Args:\n data (dict): JSON dictionary representing a user.\n\n Returns:\n A User instance.\n \"\"\"\n if data.get('shakes', None):\n shakes = [Shake.NewFromJSON(shk) for shk in data.get('shakes')]\n else:\n shakes = None\n\n return User(\n id=data.get('id', None),\n name=data.get('name', None),\n profile_image_url=data.get('profile_image_url', None),\n about=data.get('about', None),\n website=data.get('website', None),\n shakes=shakes)\n" ]
class Api(object): def __init__(self, consumer_key=None, consumer_secret=None, access_token_key=None, access_token_secret=None, base_url=None, testing=False): if base_url is None: self.base_url = 'http://mlkshk.com' else: self.base_url = base_url self.port = 80 self.authenticated = False self.testing = False if testing: self.testing = True self.consumer_key = consumer_key self.consumer_secret = consumer_secret self.access_token_key = access_token_key self.access_token_secret = access_token_secret auth_list = [consumer_key, consumer_secret, access_token_key, access_token_secret] if all(auth_list): self.authenticated = True # Set headers, client info for requests. # default_headers = {'User-Agent': 'PyShk v0.0.1'} # self.client_args = {} # self.client_args['headers'] = default_headers # Set up auth - TODO: # self.auth = None # if self.access_token_key: # token = { # 'token_type': 'mac', # 'hash_algorithm': 'hmac-sha-1', # 'access_token': self.access_token_key # } # self.auth = OAuth2(self.consumer_key, token=token) # self.client = requests.Session() # self.client.auth = self.auth def get_auth(self, redirect_uri=None): if not redirect_uri: redirect_uri = "http://localhost:8000" authentication_url = ( "https://mlkshk.com/api/authorize" "?response_type=code&client_id={key}&redirect_uri={uri}").format( key=self.consumer_key, uri=redirect_uri) access_token_url = 'https://mlkshk.com/api/token' if not self.testing: webbrowser.open(authentication_url, new=1) authorization_code = input("Enter the code from the redirected URL: ") else: authorization_code = 123456 message = { 'grant_type': "authorization_code", 'code': authorization_code, 'redirect_uri': redirect_uri, 'client_id': self.consumer_key, 'client_secret': self.consumer_secret} data = urlencode(message) req = requests.post(access_token_url, params=data, verify=False) json_resp = req.json() print(""" {full_token} >>> Your access token is: {token} >>> Your access secret is: {secret} """.format(full_token=json_resp, token=json_resp['access_token'], secret=json_resp['secret'])) self.access_token_key = json_resp['access_token'] self.access_token_secret = json_resp['secret'] def _get_url_endpoint(self, endpoint): return self.base_url + endpoint def _make_headers(self, verb=None, endpoint=None, nonce=None, timestamp=None): normalized_string = "{0}\n".format(self.access_token_key) normalized_string += "{0}\n".format(timestamp) normalized_string += "{0}\n".format(nonce) normalized_string += "{0}\n".format(verb) normalized_string += "mlkshk.com\n" normalized_string += "80\n" normalized_string += "{0}\n".format(endpoint) digest = hmac.new( self.access_token_secret.encode('ascii'), normalized_string.encode('ascii'), sha1).digest() if six.PY2: signature = base64.encodestring(digest).strip().decode('utf8') else: signature = base64.encodebytes(digest).strip().decode('utf8') auth_str = ( 'MAC token="{0}", ' 'timestamp="{1}", ' 'nonce="{2}", ' 'signature="{3}"').format( self.access_token_key, str(timestamp), nonce, signature) return auth_str def _make_request(self, verb, endpoint=None, data=None, files=None): if not self.authenticated: raise ApiInstanceUnauthorized resource_url = self._get_url_endpoint(endpoint) timestamp = int(time.mktime(datetime.datetime.utcnow().timetuple())) nonce = self.get_nonce() authorization_header = self._make_headers( verb=verb, endpoint=endpoint, nonce=nonce, timestamp=timestamp) if verb == "GET": req = requests.get( resource_url, headers={'Authorization': authorization_header}, verify=False) elif verb == "POST": if data: req = requests.post( resource_url, headers={'Authorization': authorization_header}, data=data) elif files: req = requests.post( resource_url, headers={'Authorization': authorization_header}, files=files) elif data and files: req = requests.post( resource_url, headers={'Authorization': authorization_header}, files=files, data=data) else: req = requests.post( resource_url, headers={'Authorization': authorization_header}) if req.status_code == 401: raise ApiResponseUnauthorized(req) elif req.status_code == 404: raise NotFound404(req) elif req.status_code == 500: raise Exception(req) if self.testing: return req try: return req.json() except: print('returning req', req._content) return req @staticmethod def get_nonce(): nonce = md5( str(random.SystemRandom().randint(0, 100000000)).encode('utf8') ).hexdigest() return nonce @staticmethod def _get_image_type(image): if imghdr.what(image) == 'jpeg': return 'image/jpeg' elif imghdr.what(image) == 'gif': return 'image/gif' elif imghdr.what(image) == 'png': return 'image/png' def get_favorites(self, before=None, after=None): """ Get a list of the authenticated user's 10 most recent favorites (likes). Args: before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the authenticated user's set of Likes. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the authenticated user's set of Likes. Returns: List of SharedFile objects. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/favorites' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['favorites']] def get_user_shakes(self): """ Get a list of Shake objects for the currently authenticated user. Returns: A list of Shake objects. """ endpoint = '/api/shakes' data = self._make_request(verb="GET", endpoint=endpoint) shakes = [Shake.NewFromJSON(shk) for shk in data['shakes']] return shakes def get_shared_files_from_shake(self, shake_id=None, before=None, after=None): """ Returns a list of SharedFile objects from a particular shake. Args: shake_id (int): Shake from which to get a list of SharedFiles before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the given Shake. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the given Shake. Returns: List (list) of SharedFiles. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/shakes' if shake_id: endpoint += '/{0}'.format(shake_id) if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request(verb="GET", endpoint=endpoint) return [SharedFile.NewFromJSON(f) for f in data['sharedfiles']] def get_shared_file(self, sharekey=None): """ Returns a SharedFile object given by the sharekey. Args: sharekey (str): Sharekey of the SharedFile you want to retrieve. Returns: SharedFile """ if not sharekey: raise Exception("You must specify a sharekey.") endpoint = '/api/sharedfile/{0}'.format(sharekey) data = self._make_request('GET', endpoint) return SharedFile.NewFromJSON(data) def like_shared_file(self, sharekey=None): """ 'Like' a SharedFile. mlkshk doesn't allow you to unlike a sharedfile, so this is ~~permanent~~. Args: sharekey (str): Sharekey for the file you want to 'like'. Returns: Either a SharedFile on success, or an exception on error. """ if not sharekey: raise Exception( "You must specify a sharekey of the file you" "want to 'like'.") endpoint = '/api/sharedfile/{sharekey}/like'.format(sharekey=sharekey) data = self._make_request("POST", endpoint=endpoint, data=None) try: sf = SharedFile.NewFromJSON(data) sf.liked = True return sf except: raise Exception("{0}".format(data['error'])) def save_shared_file(self, sharekey=None): """ Save a SharedFile to your Shake. Args: sharekey (str): Sharekey for the file to save. Returns: SharedFile saved to your shake. """ endpoint = '/api/sharedfile/{sharekey}/save'.format(sharekey=sharekey) data = self._make_request("POST", endpoint=endpoint, data=None) try: sf = SharedFile.NewFromJSON(data) sf.saved = True return sf except: raise Exception("{0}".format(data['error'])) def get_friends_shake(self, before=None, after=None): """ Contrary to the endpoint naming, this resource is for a list of SharedFiles from your friends on mlkshk. Returns: List of SharedFiles. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/friends' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['friend_shake']] def get_incoming_shake(self, before=None, after=None): """ Returns a list of the most recent SharedFiles on mlkshk.com Args: before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the Incoming Shake. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the Incoming Shake. Returns: List of SharedFile objects. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/incoming' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['incoming']] def get_magic_shake(self, before=None, after=None): """ From the API: Returns the 10 most recent files accepted by the 'magic' file selection algorithm. Currently any files with 10 or more likes are magic. Returns: List of SharedFile objects """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/magicfiles' if before: endpoint += '/before/{key}'.format(key=before) elif after: endpoint += '/after/{key}'.format(key=after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['magicfiles']] def get_comments(self, sharekey=None): """ Retrieve comments on a SharedFile Args: sharekey (str): Sharekey for the file from which you want to return the set of comments. Returns: List of Comment objects. """ if not sharekey: raise Exception( "You must specify a sharekey of the file you" "want to 'like'.") endpoint = '/api/sharedfile/{0}/comments'.format(sharekey) data = self._make_request("GET", endpoint=endpoint) return [Comment.NewFromJSON(c) for c in data['comments']] def post_comment(self, sharekey=None, comment=None): """ Post a comment on behalf of the current user to the SharedFile with the given sharekey. Args: sharekey (str): Sharekey of the SharedFile to which you'd like to post a comment. comment (str): Text of the comment to post. Returns: Comment object. """ endpoint = '/api/sharedfile/{0}/comments'.format(sharekey) post_data = {'body': comment} data = self._make_request("POST", endpoint=endpoint, data=post_data) return Comment.NewFromJSON(data) def post_shared_file(self, image_file=None, source_link=None, shake_id=None, title=None, description=None): """ Upload an image. TODO: Don't have a pro account to test (or even write) code to upload a shared filed to a particular shake. Args: image_file (str): path to an image (jpg/gif) on your computer. source_link (str): URL of a source (youtube/vine/etc.) shake_id (int): shake to which to upload the file or source_link [optional] title (str): title of the SharedFile [optional] description (str): description of the SharedFile Returns: SharedFile key. """ if image_file and source_link: raise Exception('You can only specify an image file or ' 'a source link, not both.') if not image_file and not source_link: raise Exception('You must specify an image file or a source link') content_type = self._get_image_type(image_file) if not title: title = os.path.basename(image_file) f = open(image_file, 'rb') endpoint = '/api/upload' files = {'file': (title, f, content_type)} data = self._make_request('POST', endpoint=endpoint, files=files) f.close() return data def update_shared_file(self, sharekey=None, title=None, description=None): """ Update the editable details (just the title and description) of a SharedFile. Args: sharekey (str): Sharekey of the SharedFile to update. title (Optional[str]): Title of the SharedFile. description (Optional[str]): Description of the SharedFile Returns: SharedFile on success, 404 on Sharekey not found, 403 on unauthorized. """ if not sharekey: raise Exception( "You must specify a sharekey for the sharedfile" "you wish to update.") if not (title or description): raise Exception("You must specify a title or description.") post_data = {} if title: post_data['title'] = title if description: post_data['description'] = description endpoint = '/api/sharedfile/{0}'.format(sharekey) data = self._make_request('POST', endpoint=endpoint, data=post_data) return SharedFile.NewFromJSON(data)
jeremylow/pyshk
pyshk/api.py
Api.get_user_shakes
python
def get_user_shakes(self): endpoint = '/api/shakes' data = self._make_request(verb="GET", endpoint=endpoint) shakes = [Shake.NewFromJSON(shk) for shk in data['shakes']] return shakes
Get a list of Shake objects for the currently authenticated user. Returns: A list of Shake objects.
train
https://github.com/jeremylow/pyshk/blob/3ab92f6706397cde7a18367266eba9e0f1ada868/pyshk/api.py#L297-L307
[ "def _make_request(self, verb, endpoint=None, data=None, files=None):\n if not self.authenticated:\n raise ApiInstanceUnauthorized\n\n resource_url = self._get_url_endpoint(endpoint)\n\n timestamp = int(time.mktime(datetime.datetime.utcnow().timetuple()))\n nonce = self.get_nonce()\n\n authorization_header = self._make_headers(\n verb=verb,\n endpoint=endpoint,\n nonce=nonce,\n timestamp=timestamp)\n\n if verb == \"GET\":\n req = requests.get(\n resource_url,\n headers={'Authorization': authorization_header},\n verify=False)\n elif verb == \"POST\":\n if data:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n data=data)\n elif files:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n files=files)\n elif data and files:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n files=files,\n data=data)\n else:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header})\n\n if req.status_code == 401:\n raise ApiResponseUnauthorized(req)\n elif req.status_code == 404:\n raise NotFound404(req)\n elif req.status_code == 500:\n raise Exception(req)\n\n if self.testing:\n return req\n\n try:\n return req.json()\n except:\n print('returning req', req._content)\n return req\n" ]
class Api(object): def __init__(self, consumer_key=None, consumer_secret=None, access_token_key=None, access_token_secret=None, base_url=None, testing=False): if base_url is None: self.base_url = 'http://mlkshk.com' else: self.base_url = base_url self.port = 80 self.authenticated = False self.testing = False if testing: self.testing = True self.consumer_key = consumer_key self.consumer_secret = consumer_secret self.access_token_key = access_token_key self.access_token_secret = access_token_secret auth_list = [consumer_key, consumer_secret, access_token_key, access_token_secret] if all(auth_list): self.authenticated = True # Set headers, client info for requests. # default_headers = {'User-Agent': 'PyShk v0.0.1'} # self.client_args = {} # self.client_args['headers'] = default_headers # Set up auth - TODO: # self.auth = None # if self.access_token_key: # token = { # 'token_type': 'mac', # 'hash_algorithm': 'hmac-sha-1', # 'access_token': self.access_token_key # } # self.auth = OAuth2(self.consumer_key, token=token) # self.client = requests.Session() # self.client.auth = self.auth def get_auth(self, redirect_uri=None): if not redirect_uri: redirect_uri = "http://localhost:8000" authentication_url = ( "https://mlkshk.com/api/authorize" "?response_type=code&client_id={key}&redirect_uri={uri}").format( key=self.consumer_key, uri=redirect_uri) access_token_url = 'https://mlkshk.com/api/token' if not self.testing: webbrowser.open(authentication_url, new=1) authorization_code = input("Enter the code from the redirected URL: ") else: authorization_code = 123456 message = { 'grant_type': "authorization_code", 'code': authorization_code, 'redirect_uri': redirect_uri, 'client_id': self.consumer_key, 'client_secret': self.consumer_secret} data = urlencode(message) req = requests.post(access_token_url, params=data, verify=False) json_resp = req.json() print(""" {full_token} >>> Your access token is: {token} >>> Your access secret is: {secret} """.format(full_token=json_resp, token=json_resp['access_token'], secret=json_resp['secret'])) self.access_token_key = json_resp['access_token'] self.access_token_secret = json_resp['secret'] def _get_url_endpoint(self, endpoint): return self.base_url + endpoint def _make_headers(self, verb=None, endpoint=None, nonce=None, timestamp=None): normalized_string = "{0}\n".format(self.access_token_key) normalized_string += "{0}\n".format(timestamp) normalized_string += "{0}\n".format(nonce) normalized_string += "{0}\n".format(verb) normalized_string += "mlkshk.com\n" normalized_string += "80\n" normalized_string += "{0}\n".format(endpoint) digest = hmac.new( self.access_token_secret.encode('ascii'), normalized_string.encode('ascii'), sha1).digest() if six.PY2: signature = base64.encodestring(digest).strip().decode('utf8') else: signature = base64.encodebytes(digest).strip().decode('utf8') auth_str = ( 'MAC token="{0}", ' 'timestamp="{1}", ' 'nonce="{2}", ' 'signature="{3}"').format( self.access_token_key, str(timestamp), nonce, signature) return auth_str def _make_request(self, verb, endpoint=None, data=None, files=None): if not self.authenticated: raise ApiInstanceUnauthorized resource_url = self._get_url_endpoint(endpoint) timestamp = int(time.mktime(datetime.datetime.utcnow().timetuple())) nonce = self.get_nonce() authorization_header = self._make_headers( verb=verb, endpoint=endpoint, nonce=nonce, timestamp=timestamp) if verb == "GET": req = requests.get( resource_url, headers={'Authorization': authorization_header}, verify=False) elif verb == "POST": if data: req = requests.post( resource_url, headers={'Authorization': authorization_header}, data=data) elif files: req = requests.post( resource_url, headers={'Authorization': authorization_header}, files=files) elif data and files: req = requests.post( resource_url, headers={'Authorization': authorization_header}, files=files, data=data) else: req = requests.post( resource_url, headers={'Authorization': authorization_header}) if req.status_code == 401: raise ApiResponseUnauthorized(req) elif req.status_code == 404: raise NotFound404(req) elif req.status_code == 500: raise Exception(req) if self.testing: return req try: return req.json() except: print('returning req', req._content) return req @staticmethod def get_nonce(): nonce = md5( str(random.SystemRandom().randint(0, 100000000)).encode('utf8') ).hexdigest() return nonce @staticmethod def _get_image_type(image): if imghdr.what(image) == 'jpeg': return 'image/jpeg' elif imghdr.what(image) == 'gif': return 'image/gif' elif imghdr.what(image) == 'png': return 'image/png' def get_favorites(self, before=None, after=None): """ Get a list of the authenticated user's 10 most recent favorites (likes). Args: before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the authenticated user's set of Likes. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the authenticated user's set of Likes. Returns: List of SharedFile objects. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/favorites' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['favorites']] def get_user(self, user_id=None, user_name=None): """ Get a user object from the API. If no ``user_id`` or ``user_name`` is specified, it will return the User object for the currently authenticated user. Args: user_id (int): User ID of the user for whom you want to get information. [Optional] user_name(str): Username for the user for whom you want to get information. [Optional] Returns: A User object. """ if user_id: endpoint = '/api/user_id/{0}'.format(user_id) elif user_name: endpoint = '/api/user_name/{0}'.format(user_name) else: # Return currently authorized user endpoint = '/api/user' data = self._make_request(verb="GET", endpoint=endpoint) try: return User.NewFromJSON(data) except: return data def get_shared_files_from_shake(self, shake_id=None, before=None, after=None): """ Returns a list of SharedFile objects from a particular shake. Args: shake_id (int): Shake from which to get a list of SharedFiles before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the given Shake. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the given Shake. Returns: List (list) of SharedFiles. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/shakes' if shake_id: endpoint += '/{0}'.format(shake_id) if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request(verb="GET", endpoint=endpoint) return [SharedFile.NewFromJSON(f) for f in data['sharedfiles']] def get_shared_file(self, sharekey=None): """ Returns a SharedFile object given by the sharekey. Args: sharekey (str): Sharekey of the SharedFile you want to retrieve. Returns: SharedFile """ if not sharekey: raise Exception("You must specify a sharekey.") endpoint = '/api/sharedfile/{0}'.format(sharekey) data = self._make_request('GET', endpoint) return SharedFile.NewFromJSON(data) def like_shared_file(self, sharekey=None): """ 'Like' a SharedFile. mlkshk doesn't allow you to unlike a sharedfile, so this is ~~permanent~~. Args: sharekey (str): Sharekey for the file you want to 'like'. Returns: Either a SharedFile on success, or an exception on error. """ if not sharekey: raise Exception( "You must specify a sharekey of the file you" "want to 'like'.") endpoint = '/api/sharedfile/{sharekey}/like'.format(sharekey=sharekey) data = self._make_request("POST", endpoint=endpoint, data=None) try: sf = SharedFile.NewFromJSON(data) sf.liked = True return sf except: raise Exception("{0}".format(data['error'])) def save_shared_file(self, sharekey=None): """ Save a SharedFile to your Shake. Args: sharekey (str): Sharekey for the file to save. Returns: SharedFile saved to your shake. """ endpoint = '/api/sharedfile/{sharekey}/save'.format(sharekey=sharekey) data = self._make_request("POST", endpoint=endpoint, data=None) try: sf = SharedFile.NewFromJSON(data) sf.saved = True return sf except: raise Exception("{0}".format(data['error'])) def get_friends_shake(self, before=None, after=None): """ Contrary to the endpoint naming, this resource is for a list of SharedFiles from your friends on mlkshk. Returns: List of SharedFiles. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/friends' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['friend_shake']] def get_incoming_shake(self, before=None, after=None): """ Returns a list of the most recent SharedFiles on mlkshk.com Args: before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the Incoming Shake. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the Incoming Shake. Returns: List of SharedFile objects. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/incoming' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['incoming']] def get_magic_shake(self, before=None, after=None): """ From the API: Returns the 10 most recent files accepted by the 'magic' file selection algorithm. Currently any files with 10 or more likes are magic. Returns: List of SharedFile objects """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/magicfiles' if before: endpoint += '/before/{key}'.format(key=before) elif after: endpoint += '/after/{key}'.format(key=after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['magicfiles']] def get_comments(self, sharekey=None): """ Retrieve comments on a SharedFile Args: sharekey (str): Sharekey for the file from which you want to return the set of comments. Returns: List of Comment objects. """ if not sharekey: raise Exception( "You must specify a sharekey of the file you" "want to 'like'.") endpoint = '/api/sharedfile/{0}/comments'.format(sharekey) data = self._make_request("GET", endpoint=endpoint) return [Comment.NewFromJSON(c) for c in data['comments']] def post_comment(self, sharekey=None, comment=None): """ Post a comment on behalf of the current user to the SharedFile with the given sharekey. Args: sharekey (str): Sharekey of the SharedFile to which you'd like to post a comment. comment (str): Text of the comment to post. Returns: Comment object. """ endpoint = '/api/sharedfile/{0}/comments'.format(sharekey) post_data = {'body': comment} data = self._make_request("POST", endpoint=endpoint, data=post_data) return Comment.NewFromJSON(data) def post_shared_file(self, image_file=None, source_link=None, shake_id=None, title=None, description=None): """ Upload an image. TODO: Don't have a pro account to test (or even write) code to upload a shared filed to a particular shake. Args: image_file (str): path to an image (jpg/gif) on your computer. source_link (str): URL of a source (youtube/vine/etc.) shake_id (int): shake to which to upload the file or source_link [optional] title (str): title of the SharedFile [optional] description (str): description of the SharedFile Returns: SharedFile key. """ if image_file and source_link: raise Exception('You can only specify an image file or ' 'a source link, not both.') if not image_file and not source_link: raise Exception('You must specify an image file or a source link') content_type = self._get_image_type(image_file) if not title: title = os.path.basename(image_file) f = open(image_file, 'rb') endpoint = '/api/upload' files = {'file': (title, f, content_type)} data = self._make_request('POST', endpoint=endpoint, files=files) f.close() return data def update_shared_file(self, sharekey=None, title=None, description=None): """ Update the editable details (just the title and description) of a SharedFile. Args: sharekey (str): Sharekey of the SharedFile to update. title (Optional[str]): Title of the SharedFile. description (Optional[str]): Description of the SharedFile Returns: SharedFile on success, 404 on Sharekey not found, 403 on unauthorized. """ if not sharekey: raise Exception( "You must specify a sharekey for the sharedfile" "you wish to update.") if not (title or description): raise Exception("You must specify a title or description.") post_data = {} if title: post_data['title'] = title if description: post_data['description'] = description endpoint = '/api/sharedfile/{0}'.format(sharekey) data = self._make_request('POST', endpoint=endpoint, data=post_data) return SharedFile.NewFromJSON(data)
jeremylow/pyshk
pyshk/api.py
Api.get_shared_files_from_shake
python
def get_shared_files_from_shake(self, shake_id=None, before=None, after=None): if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/shakes' if shake_id: endpoint += '/{0}'.format(shake_id) if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request(verb="GET", endpoint=endpoint) return [SharedFile.NewFromJSON(f) for f in data['sharedfiles']]
Returns a list of SharedFile objects from a particular shake. Args: shake_id (int): Shake from which to get a list of SharedFiles before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the given Shake. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the given Shake. Returns: List (list) of SharedFiles.
train
https://github.com/jeremylow/pyshk/blob/3ab92f6706397cde7a18367266eba9e0f1ada868/pyshk/api.py#L309-L340
[ "def _make_request(self, verb, endpoint=None, data=None, files=None):\n if not self.authenticated:\n raise ApiInstanceUnauthorized\n\n resource_url = self._get_url_endpoint(endpoint)\n\n timestamp = int(time.mktime(datetime.datetime.utcnow().timetuple()))\n nonce = self.get_nonce()\n\n authorization_header = self._make_headers(\n verb=verb,\n endpoint=endpoint,\n nonce=nonce,\n timestamp=timestamp)\n\n if verb == \"GET\":\n req = requests.get(\n resource_url,\n headers={'Authorization': authorization_header},\n verify=False)\n elif verb == \"POST\":\n if data:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n data=data)\n elif files:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n files=files)\n elif data and files:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n files=files,\n data=data)\n else:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header})\n\n if req.status_code == 401:\n raise ApiResponseUnauthorized(req)\n elif req.status_code == 404:\n raise NotFound404(req)\n elif req.status_code == 500:\n raise Exception(req)\n\n if self.testing:\n return req\n\n try:\n return req.json()\n except:\n print('returning req', req._content)\n return req\n" ]
class Api(object): def __init__(self, consumer_key=None, consumer_secret=None, access_token_key=None, access_token_secret=None, base_url=None, testing=False): if base_url is None: self.base_url = 'http://mlkshk.com' else: self.base_url = base_url self.port = 80 self.authenticated = False self.testing = False if testing: self.testing = True self.consumer_key = consumer_key self.consumer_secret = consumer_secret self.access_token_key = access_token_key self.access_token_secret = access_token_secret auth_list = [consumer_key, consumer_secret, access_token_key, access_token_secret] if all(auth_list): self.authenticated = True # Set headers, client info for requests. # default_headers = {'User-Agent': 'PyShk v0.0.1'} # self.client_args = {} # self.client_args['headers'] = default_headers # Set up auth - TODO: # self.auth = None # if self.access_token_key: # token = { # 'token_type': 'mac', # 'hash_algorithm': 'hmac-sha-1', # 'access_token': self.access_token_key # } # self.auth = OAuth2(self.consumer_key, token=token) # self.client = requests.Session() # self.client.auth = self.auth def get_auth(self, redirect_uri=None): if not redirect_uri: redirect_uri = "http://localhost:8000" authentication_url = ( "https://mlkshk.com/api/authorize" "?response_type=code&client_id={key}&redirect_uri={uri}").format( key=self.consumer_key, uri=redirect_uri) access_token_url = 'https://mlkshk.com/api/token' if not self.testing: webbrowser.open(authentication_url, new=1) authorization_code = input("Enter the code from the redirected URL: ") else: authorization_code = 123456 message = { 'grant_type': "authorization_code", 'code': authorization_code, 'redirect_uri': redirect_uri, 'client_id': self.consumer_key, 'client_secret': self.consumer_secret} data = urlencode(message) req = requests.post(access_token_url, params=data, verify=False) json_resp = req.json() print(""" {full_token} >>> Your access token is: {token} >>> Your access secret is: {secret} """.format(full_token=json_resp, token=json_resp['access_token'], secret=json_resp['secret'])) self.access_token_key = json_resp['access_token'] self.access_token_secret = json_resp['secret'] def _get_url_endpoint(self, endpoint): return self.base_url + endpoint def _make_headers(self, verb=None, endpoint=None, nonce=None, timestamp=None): normalized_string = "{0}\n".format(self.access_token_key) normalized_string += "{0}\n".format(timestamp) normalized_string += "{0}\n".format(nonce) normalized_string += "{0}\n".format(verb) normalized_string += "mlkshk.com\n" normalized_string += "80\n" normalized_string += "{0}\n".format(endpoint) digest = hmac.new( self.access_token_secret.encode('ascii'), normalized_string.encode('ascii'), sha1).digest() if six.PY2: signature = base64.encodestring(digest).strip().decode('utf8') else: signature = base64.encodebytes(digest).strip().decode('utf8') auth_str = ( 'MAC token="{0}", ' 'timestamp="{1}", ' 'nonce="{2}", ' 'signature="{3}"').format( self.access_token_key, str(timestamp), nonce, signature) return auth_str def _make_request(self, verb, endpoint=None, data=None, files=None): if not self.authenticated: raise ApiInstanceUnauthorized resource_url = self._get_url_endpoint(endpoint) timestamp = int(time.mktime(datetime.datetime.utcnow().timetuple())) nonce = self.get_nonce() authorization_header = self._make_headers( verb=verb, endpoint=endpoint, nonce=nonce, timestamp=timestamp) if verb == "GET": req = requests.get( resource_url, headers={'Authorization': authorization_header}, verify=False) elif verb == "POST": if data: req = requests.post( resource_url, headers={'Authorization': authorization_header}, data=data) elif files: req = requests.post( resource_url, headers={'Authorization': authorization_header}, files=files) elif data and files: req = requests.post( resource_url, headers={'Authorization': authorization_header}, files=files, data=data) else: req = requests.post( resource_url, headers={'Authorization': authorization_header}) if req.status_code == 401: raise ApiResponseUnauthorized(req) elif req.status_code == 404: raise NotFound404(req) elif req.status_code == 500: raise Exception(req) if self.testing: return req try: return req.json() except: print('returning req', req._content) return req @staticmethod def get_nonce(): nonce = md5( str(random.SystemRandom().randint(0, 100000000)).encode('utf8') ).hexdigest() return nonce @staticmethod def _get_image_type(image): if imghdr.what(image) == 'jpeg': return 'image/jpeg' elif imghdr.what(image) == 'gif': return 'image/gif' elif imghdr.what(image) == 'png': return 'image/png' def get_favorites(self, before=None, after=None): """ Get a list of the authenticated user's 10 most recent favorites (likes). Args: before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the authenticated user's set of Likes. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the authenticated user's set of Likes. Returns: List of SharedFile objects. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/favorites' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['favorites']] def get_user(self, user_id=None, user_name=None): """ Get a user object from the API. If no ``user_id`` or ``user_name`` is specified, it will return the User object for the currently authenticated user. Args: user_id (int): User ID of the user for whom you want to get information. [Optional] user_name(str): Username for the user for whom you want to get information. [Optional] Returns: A User object. """ if user_id: endpoint = '/api/user_id/{0}'.format(user_id) elif user_name: endpoint = '/api/user_name/{0}'.format(user_name) else: # Return currently authorized user endpoint = '/api/user' data = self._make_request(verb="GET", endpoint=endpoint) try: return User.NewFromJSON(data) except: return data def get_user_shakes(self): """ Get a list of Shake objects for the currently authenticated user. Returns: A list of Shake objects. """ endpoint = '/api/shakes' data = self._make_request(verb="GET", endpoint=endpoint) shakes = [Shake.NewFromJSON(shk) for shk in data['shakes']] return shakes def get_shared_file(self, sharekey=None): """ Returns a SharedFile object given by the sharekey. Args: sharekey (str): Sharekey of the SharedFile you want to retrieve. Returns: SharedFile """ if not sharekey: raise Exception("You must specify a sharekey.") endpoint = '/api/sharedfile/{0}'.format(sharekey) data = self._make_request('GET', endpoint) return SharedFile.NewFromJSON(data) def like_shared_file(self, sharekey=None): """ 'Like' a SharedFile. mlkshk doesn't allow you to unlike a sharedfile, so this is ~~permanent~~. Args: sharekey (str): Sharekey for the file you want to 'like'. Returns: Either a SharedFile on success, or an exception on error. """ if not sharekey: raise Exception( "You must specify a sharekey of the file you" "want to 'like'.") endpoint = '/api/sharedfile/{sharekey}/like'.format(sharekey=sharekey) data = self._make_request("POST", endpoint=endpoint, data=None) try: sf = SharedFile.NewFromJSON(data) sf.liked = True return sf except: raise Exception("{0}".format(data['error'])) def save_shared_file(self, sharekey=None): """ Save a SharedFile to your Shake. Args: sharekey (str): Sharekey for the file to save. Returns: SharedFile saved to your shake. """ endpoint = '/api/sharedfile/{sharekey}/save'.format(sharekey=sharekey) data = self._make_request("POST", endpoint=endpoint, data=None) try: sf = SharedFile.NewFromJSON(data) sf.saved = True return sf except: raise Exception("{0}".format(data['error'])) def get_friends_shake(self, before=None, after=None): """ Contrary to the endpoint naming, this resource is for a list of SharedFiles from your friends on mlkshk. Returns: List of SharedFiles. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/friends' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['friend_shake']] def get_incoming_shake(self, before=None, after=None): """ Returns a list of the most recent SharedFiles on mlkshk.com Args: before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the Incoming Shake. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the Incoming Shake. Returns: List of SharedFile objects. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/incoming' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['incoming']] def get_magic_shake(self, before=None, after=None): """ From the API: Returns the 10 most recent files accepted by the 'magic' file selection algorithm. Currently any files with 10 or more likes are magic. Returns: List of SharedFile objects """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/magicfiles' if before: endpoint += '/before/{key}'.format(key=before) elif after: endpoint += '/after/{key}'.format(key=after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['magicfiles']] def get_comments(self, sharekey=None): """ Retrieve comments on a SharedFile Args: sharekey (str): Sharekey for the file from which you want to return the set of comments. Returns: List of Comment objects. """ if not sharekey: raise Exception( "You must specify a sharekey of the file you" "want to 'like'.") endpoint = '/api/sharedfile/{0}/comments'.format(sharekey) data = self._make_request("GET", endpoint=endpoint) return [Comment.NewFromJSON(c) for c in data['comments']] def post_comment(self, sharekey=None, comment=None): """ Post a comment on behalf of the current user to the SharedFile with the given sharekey. Args: sharekey (str): Sharekey of the SharedFile to which you'd like to post a comment. comment (str): Text of the comment to post. Returns: Comment object. """ endpoint = '/api/sharedfile/{0}/comments'.format(sharekey) post_data = {'body': comment} data = self._make_request("POST", endpoint=endpoint, data=post_data) return Comment.NewFromJSON(data) def post_shared_file(self, image_file=None, source_link=None, shake_id=None, title=None, description=None): """ Upload an image. TODO: Don't have a pro account to test (or even write) code to upload a shared filed to a particular shake. Args: image_file (str): path to an image (jpg/gif) on your computer. source_link (str): URL of a source (youtube/vine/etc.) shake_id (int): shake to which to upload the file or source_link [optional] title (str): title of the SharedFile [optional] description (str): description of the SharedFile Returns: SharedFile key. """ if image_file and source_link: raise Exception('You can only specify an image file or ' 'a source link, not both.') if not image_file and not source_link: raise Exception('You must specify an image file or a source link') content_type = self._get_image_type(image_file) if not title: title = os.path.basename(image_file) f = open(image_file, 'rb') endpoint = '/api/upload' files = {'file': (title, f, content_type)} data = self._make_request('POST', endpoint=endpoint, files=files) f.close() return data def update_shared_file(self, sharekey=None, title=None, description=None): """ Update the editable details (just the title and description) of a SharedFile. Args: sharekey (str): Sharekey of the SharedFile to update. title (Optional[str]): Title of the SharedFile. description (Optional[str]): Description of the SharedFile Returns: SharedFile on success, 404 on Sharekey not found, 403 on unauthorized. """ if not sharekey: raise Exception( "You must specify a sharekey for the sharedfile" "you wish to update.") if not (title or description): raise Exception("You must specify a title or description.") post_data = {} if title: post_data['title'] = title if description: post_data['description'] = description endpoint = '/api/sharedfile/{0}'.format(sharekey) data = self._make_request('POST', endpoint=endpoint, data=post_data) return SharedFile.NewFromJSON(data)
jeremylow/pyshk
pyshk/api.py
Api.get_shared_file
python
def get_shared_file(self, sharekey=None): if not sharekey: raise Exception("You must specify a sharekey.") endpoint = '/api/sharedfile/{0}'.format(sharekey) data = self._make_request('GET', endpoint) return SharedFile.NewFromJSON(data)
Returns a SharedFile object given by the sharekey. Args: sharekey (str): Sharekey of the SharedFile you want to retrieve. Returns: SharedFile
train
https://github.com/jeremylow/pyshk/blob/3ab92f6706397cde7a18367266eba9e0f1ada868/pyshk/api.py#L342-L356
[ "def _make_request(self, verb, endpoint=None, data=None, files=None):\n if not self.authenticated:\n raise ApiInstanceUnauthorized\n\n resource_url = self._get_url_endpoint(endpoint)\n\n timestamp = int(time.mktime(datetime.datetime.utcnow().timetuple()))\n nonce = self.get_nonce()\n\n authorization_header = self._make_headers(\n verb=verb,\n endpoint=endpoint,\n nonce=nonce,\n timestamp=timestamp)\n\n if verb == \"GET\":\n req = requests.get(\n resource_url,\n headers={'Authorization': authorization_header},\n verify=False)\n elif verb == \"POST\":\n if data:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n data=data)\n elif files:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n files=files)\n elif data and files:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n files=files,\n data=data)\n else:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header})\n\n if req.status_code == 401:\n raise ApiResponseUnauthorized(req)\n elif req.status_code == 404:\n raise NotFound404(req)\n elif req.status_code == 500:\n raise Exception(req)\n\n if self.testing:\n return req\n\n try:\n return req.json()\n except:\n print('returning req', req._content)\n return req\n", "def NewFromJSON(data):\n \"\"\"\n Create a new SharedFile instance from a JSON dict.\n\n Args:\n data (dict): JSON dictionary representing a SharedFile.\n\n Returns:\n A SharedFile instance.\n \"\"\"\n return SharedFile(\n sharekey=data.get('sharekey', None),\n name=data.get('name', None),\n user=User.NewFromJSON(data.get('user', None)),\n title=data.get('title', None),\n description=data.get('description', None),\n posted_at=data.get('posted_at', None),\n permalink=data.get('permalink', None),\n width=data.get('width', None),\n height=data.get('height', None),\n views=data.get('views', 0),\n likes=data.get('likes', 0),\n saves=data.get('saves', 0),\n comments=data.get('comments', None),\n nsfw=data.get('nsfw', False),\n image_url=data.get('image_url', None),\n source_url=data.get('source_url', None),\n saved=data.get('saved', False),\n liked=data.get('liked', False),\n )\n" ]
class Api(object): def __init__(self, consumer_key=None, consumer_secret=None, access_token_key=None, access_token_secret=None, base_url=None, testing=False): if base_url is None: self.base_url = 'http://mlkshk.com' else: self.base_url = base_url self.port = 80 self.authenticated = False self.testing = False if testing: self.testing = True self.consumer_key = consumer_key self.consumer_secret = consumer_secret self.access_token_key = access_token_key self.access_token_secret = access_token_secret auth_list = [consumer_key, consumer_secret, access_token_key, access_token_secret] if all(auth_list): self.authenticated = True # Set headers, client info for requests. # default_headers = {'User-Agent': 'PyShk v0.0.1'} # self.client_args = {} # self.client_args['headers'] = default_headers # Set up auth - TODO: # self.auth = None # if self.access_token_key: # token = { # 'token_type': 'mac', # 'hash_algorithm': 'hmac-sha-1', # 'access_token': self.access_token_key # } # self.auth = OAuth2(self.consumer_key, token=token) # self.client = requests.Session() # self.client.auth = self.auth def get_auth(self, redirect_uri=None): if not redirect_uri: redirect_uri = "http://localhost:8000" authentication_url = ( "https://mlkshk.com/api/authorize" "?response_type=code&client_id={key}&redirect_uri={uri}").format( key=self.consumer_key, uri=redirect_uri) access_token_url = 'https://mlkshk.com/api/token' if not self.testing: webbrowser.open(authentication_url, new=1) authorization_code = input("Enter the code from the redirected URL: ") else: authorization_code = 123456 message = { 'grant_type': "authorization_code", 'code': authorization_code, 'redirect_uri': redirect_uri, 'client_id': self.consumer_key, 'client_secret': self.consumer_secret} data = urlencode(message) req = requests.post(access_token_url, params=data, verify=False) json_resp = req.json() print(""" {full_token} >>> Your access token is: {token} >>> Your access secret is: {secret} """.format(full_token=json_resp, token=json_resp['access_token'], secret=json_resp['secret'])) self.access_token_key = json_resp['access_token'] self.access_token_secret = json_resp['secret'] def _get_url_endpoint(self, endpoint): return self.base_url + endpoint def _make_headers(self, verb=None, endpoint=None, nonce=None, timestamp=None): normalized_string = "{0}\n".format(self.access_token_key) normalized_string += "{0}\n".format(timestamp) normalized_string += "{0}\n".format(nonce) normalized_string += "{0}\n".format(verb) normalized_string += "mlkshk.com\n" normalized_string += "80\n" normalized_string += "{0}\n".format(endpoint) digest = hmac.new( self.access_token_secret.encode('ascii'), normalized_string.encode('ascii'), sha1).digest() if six.PY2: signature = base64.encodestring(digest).strip().decode('utf8') else: signature = base64.encodebytes(digest).strip().decode('utf8') auth_str = ( 'MAC token="{0}", ' 'timestamp="{1}", ' 'nonce="{2}", ' 'signature="{3}"').format( self.access_token_key, str(timestamp), nonce, signature) return auth_str def _make_request(self, verb, endpoint=None, data=None, files=None): if not self.authenticated: raise ApiInstanceUnauthorized resource_url = self._get_url_endpoint(endpoint) timestamp = int(time.mktime(datetime.datetime.utcnow().timetuple())) nonce = self.get_nonce() authorization_header = self._make_headers( verb=verb, endpoint=endpoint, nonce=nonce, timestamp=timestamp) if verb == "GET": req = requests.get( resource_url, headers={'Authorization': authorization_header}, verify=False) elif verb == "POST": if data: req = requests.post( resource_url, headers={'Authorization': authorization_header}, data=data) elif files: req = requests.post( resource_url, headers={'Authorization': authorization_header}, files=files) elif data and files: req = requests.post( resource_url, headers={'Authorization': authorization_header}, files=files, data=data) else: req = requests.post( resource_url, headers={'Authorization': authorization_header}) if req.status_code == 401: raise ApiResponseUnauthorized(req) elif req.status_code == 404: raise NotFound404(req) elif req.status_code == 500: raise Exception(req) if self.testing: return req try: return req.json() except: print('returning req', req._content) return req @staticmethod def get_nonce(): nonce = md5( str(random.SystemRandom().randint(0, 100000000)).encode('utf8') ).hexdigest() return nonce @staticmethod def _get_image_type(image): if imghdr.what(image) == 'jpeg': return 'image/jpeg' elif imghdr.what(image) == 'gif': return 'image/gif' elif imghdr.what(image) == 'png': return 'image/png' def get_favorites(self, before=None, after=None): """ Get a list of the authenticated user's 10 most recent favorites (likes). Args: before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the authenticated user's set of Likes. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the authenticated user's set of Likes. Returns: List of SharedFile objects. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/favorites' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['favorites']] def get_user(self, user_id=None, user_name=None): """ Get a user object from the API. If no ``user_id`` or ``user_name`` is specified, it will return the User object for the currently authenticated user. Args: user_id (int): User ID of the user for whom you want to get information. [Optional] user_name(str): Username for the user for whom you want to get information. [Optional] Returns: A User object. """ if user_id: endpoint = '/api/user_id/{0}'.format(user_id) elif user_name: endpoint = '/api/user_name/{0}'.format(user_name) else: # Return currently authorized user endpoint = '/api/user' data = self._make_request(verb="GET", endpoint=endpoint) try: return User.NewFromJSON(data) except: return data def get_user_shakes(self): """ Get a list of Shake objects for the currently authenticated user. Returns: A list of Shake objects. """ endpoint = '/api/shakes' data = self._make_request(verb="GET", endpoint=endpoint) shakes = [Shake.NewFromJSON(shk) for shk in data['shakes']] return shakes def get_shared_files_from_shake(self, shake_id=None, before=None, after=None): """ Returns a list of SharedFile objects from a particular shake. Args: shake_id (int): Shake from which to get a list of SharedFiles before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the given Shake. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the given Shake. Returns: List (list) of SharedFiles. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/shakes' if shake_id: endpoint += '/{0}'.format(shake_id) if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request(verb="GET", endpoint=endpoint) return [SharedFile.NewFromJSON(f) for f in data['sharedfiles']] def like_shared_file(self, sharekey=None): """ 'Like' a SharedFile. mlkshk doesn't allow you to unlike a sharedfile, so this is ~~permanent~~. Args: sharekey (str): Sharekey for the file you want to 'like'. Returns: Either a SharedFile on success, or an exception on error. """ if not sharekey: raise Exception( "You must specify a sharekey of the file you" "want to 'like'.") endpoint = '/api/sharedfile/{sharekey}/like'.format(sharekey=sharekey) data = self._make_request("POST", endpoint=endpoint, data=None) try: sf = SharedFile.NewFromJSON(data) sf.liked = True return sf except: raise Exception("{0}".format(data['error'])) def save_shared_file(self, sharekey=None): """ Save a SharedFile to your Shake. Args: sharekey (str): Sharekey for the file to save. Returns: SharedFile saved to your shake. """ endpoint = '/api/sharedfile/{sharekey}/save'.format(sharekey=sharekey) data = self._make_request("POST", endpoint=endpoint, data=None) try: sf = SharedFile.NewFromJSON(data) sf.saved = True return sf except: raise Exception("{0}".format(data['error'])) def get_friends_shake(self, before=None, after=None): """ Contrary to the endpoint naming, this resource is for a list of SharedFiles from your friends on mlkshk. Returns: List of SharedFiles. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/friends' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['friend_shake']] def get_incoming_shake(self, before=None, after=None): """ Returns a list of the most recent SharedFiles on mlkshk.com Args: before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the Incoming Shake. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the Incoming Shake. Returns: List of SharedFile objects. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/incoming' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['incoming']] def get_magic_shake(self, before=None, after=None): """ From the API: Returns the 10 most recent files accepted by the 'magic' file selection algorithm. Currently any files with 10 or more likes are magic. Returns: List of SharedFile objects """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/magicfiles' if before: endpoint += '/before/{key}'.format(key=before) elif after: endpoint += '/after/{key}'.format(key=after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['magicfiles']] def get_comments(self, sharekey=None): """ Retrieve comments on a SharedFile Args: sharekey (str): Sharekey for the file from which you want to return the set of comments. Returns: List of Comment objects. """ if not sharekey: raise Exception( "You must specify a sharekey of the file you" "want to 'like'.") endpoint = '/api/sharedfile/{0}/comments'.format(sharekey) data = self._make_request("GET", endpoint=endpoint) return [Comment.NewFromJSON(c) for c in data['comments']] def post_comment(self, sharekey=None, comment=None): """ Post a comment on behalf of the current user to the SharedFile with the given sharekey. Args: sharekey (str): Sharekey of the SharedFile to which you'd like to post a comment. comment (str): Text of the comment to post. Returns: Comment object. """ endpoint = '/api/sharedfile/{0}/comments'.format(sharekey) post_data = {'body': comment} data = self._make_request("POST", endpoint=endpoint, data=post_data) return Comment.NewFromJSON(data) def post_shared_file(self, image_file=None, source_link=None, shake_id=None, title=None, description=None): """ Upload an image. TODO: Don't have a pro account to test (or even write) code to upload a shared filed to a particular shake. Args: image_file (str): path to an image (jpg/gif) on your computer. source_link (str): URL of a source (youtube/vine/etc.) shake_id (int): shake to which to upload the file or source_link [optional] title (str): title of the SharedFile [optional] description (str): description of the SharedFile Returns: SharedFile key. """ if image_file and source_link: raise Exception('You can only specify an image file or ' 'a source link, not both.') if not image_file and not source_link: raise Exception('You must specify an image file or a source link') content_type = self._get_image_type(image_file) if not title: title = os.path.basename(image_file) f = open(image_file, 'rb') endpoint = '/api/upload' files = {'file': (title, f, content_type)} data = self._make_request('POST', endpoint=endpoint, files=files) f.close() return data def update_shared_file(self, sharekey=None, title=None, description=None): """ Update the editable details (just the title and description) of a SharedFile. Args: sharekey (str): Sharekey of the SharedFile to update. title (Optional[str]): Title of the SharedFile. description (Optional[str]): Description of the SharedFile Returns: SharedFile on success, 404 on Sharekey not found, 403 on unauthorized. """ if not sharekey: raise Exception( "You must specify a sharekey for the sharedfile" "you wish to update.") if not (title or description): raise Exception("You must specify a title or description.") post_data = {} if title: post_data['title'] = title if description: post_data['description'] = description endpoint = '/api/sharedfile/{0}'.format(sharekey) data = self._make_request('POST', endpoint=endpoint, data=post_data) return SharedFile.NewFromJSON(data)
jeremylow/pyshk
pyshk/api.py
Api.like_shared_file
python
def like_shared_file(self, sharekey=None): if not sharekey: raise Exception( "You must specify a sharekey of the file you" "want to 'like'.") endpoint = '/api/sharedfile/{sharekey}/like'.format(sharekey=sharekey) data = self._make_request("POST", endpoint=endpoint, data=None) try: sf = SharedFile.NewFromJSON(data) sf.liked = True return sf except: raise Exception("{0}".format(data['error']))
'Like' a SharedFile. mlkshk doesn't allow you to unlike a sharedfile, so this is ~~permanent~~. Args: sharekey (str): Sharekey for the file you want to 'like'. Returns: Either a SharedFile on success, or an exception on error.
train
https://github.com/jeremylow/pyshk/blob/3ab92f6706397cde7a18367266eba9e0f1ada868/pyshk/api.py#L358-L382
[ "def _make_request(self, verb, endpoint=None, data=None, files=None):\n if not self.authenticated:\n raise ApiInstanceUnauthorized\n\n resource_url = self._get_url_endpoint(endpoint)\n\n timestamp = int(time.mktime(datetime.datetime.utcnow().timetuple()))\n nonce = self.get_nonce()\n\n authorization_header = self._make_headers(\n verb=verb,\n endpoint=endpoint,\n nonce=nonce,\n timestamp=timestamp)\n\n if verb == \"GET\":\n req = requests.get(\n resource_url,\n headers={'Authorization': authorization_header},\n verify=False)\n elif verb == \"POST\":\n if data:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n data=data)\n elif files:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n files=files)\n elif data and files:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n files=files,\n data=data)\n else:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header})\n\n if req.status_code == 401:\n raise ApiResponseUnauthorized(req)\n elif req.status_code == 404:\n raise NotFound404(req)\n elif req.status_code == 500:\n raise Exception(req)\n\n if self.testing:\n return req\n\n try:\n return req.json()\n except:\n print('returning req', req._content)\n return req\n", "def NewFromJSON(data):\n \"\"\"\n Create a new SharedFile instance from a JSON dict.\n\n Args:\n data (dict): JSON dictionary representing a SharedFile.\n\n Returns:\n A SharedFile instance.\n \"\"\"\n return SharedFile(\n sharekey=data.get('sharekey', None),\n name=data.get('name', None),\n user=User.NewFromJSON(data.get('user', None)),\n title=data.get('title', None),\n description=data.get('description', None),\n posted_at=data.get('posted_at', None),\n permalink=data.get('permalink', None),\n width=data.get('width', None),\n height=data.get('height', None),\n views=data.get('views', 0),\n likes=data.get('likes', 0),\n saves=data.get('saves', 0),\n comments=data.get('comments', None),\n nsfw=data.get('nsfw', False),\n image_url=data.get('image_url', None),\n source_url=data.get('source_url', None),\n saved=data.get('saved', False),\n liked=data.get('liked', False),\n )\n" ]
class Api(object): def __init__(self, consumer_key=None, consumer_secret=None, access_token_key=None, access_token_secret=None, base_url=None, testing=False): if base_url is None: self.base_url = 'http://mlkshk.com' else: self.base_url = base_url self.port = 80 self.authenticated = False self.testing = False if testing: self.testing = True self.consumer_key = consumer_key self.consumer_secret = consumer_secret self.access_token_key = access_token_key self.access_token_secret = access_token_secret auth_list = [consumer_key, consumer_secret, access_token_key, access_token_secret] if all(auth_list): self.authenticated = True # Set headers, client info for requests. # default_headers = {'User-Agent': 'PyShk v0.0.1'} # self.client_args = {} # self.client_args['headers'] = default_headers # Set up auth - TODO: # self.auth = None # if self.access_token_key: # token = { # 'token_type': 'mac', # 'hash_algorithm': 'hmac-sha-1', # 'access_token': self.access_token_key # } # self.auth = OAuth2(self.consumer_key, token=token) # self.client = requests.Session() # self.client.auth = self.auth def get_auth(self, redirect_uri=None): if not redirect_uri: redirect_uri = "http://localhost:8000" authentication_url = ( "https://mlkshk.com/api/authorize" "?response_type=code&client_id={key}&redirect_uri={uri}").format( key=self.consumer_key, uri=redirect_uri) access_token_url = 'https://mlkshk.com/api/token' if not self.testing: webbrowser.open(authentication_url, new=1) authorization_code = input("Enter the code from the redirected URL: ") else: authorization_code = 123456 message = { 'grant_type': "authorization_code", 'code': authorization_code, 'redirect_uri': redirect_uri, 'client_id': self.consumer_key, 'client_secret': self.consumer_secret} data = urlencode(message) req = requests.post(access_token_url, params=data, verify=False) json_resp = req.json() print(""" {full_token} >>> Your access token is: {token} >>> Your access secret is: {secret} """.format(full_token=json_resp, token=json_resp['access_token'], secret=json_resp['secret'])) self.access_token_key = json_resp['access_token'] self.access_token_secret = json_resp['secret'] def _get_url_endpoint(self, endpoint): return self.base_url + endpoint def _make_headers(self, verb=None, endpoint=None, nonce=None, timestamp=None): normalized_string = "{0}\n".format(self.access_token_key) normalized_string += "{0}\n".format(timestamp) normalized_string += "{0}\n".format(nonce) normalized_string += "{0}\n".format(verb) normalized_string += "mlkshk.com\n" normalized_string += "80\n" normalized_string += "{0}\n".format(endpoint) digest = hmac.new( self.access_token_secret.encode('ascii'), normalized_string.encode('ascii'), sha1).digest() if six.PY2: signature = base64.encodestring(digest).strip().decode('utf8') else: signature = base64.encodebytes(digest).strip().decode('utf8') auth_str = ( 'MAC token="{0}", ' 'timestamp="{1}", ' 'nonce="{2}", ' 'signature="{3}"').format( self.access_token_key, str(timestamp), nonce, signature) return auth_str def _make_request(self, verb, endpoint=None, data=None, files=None): if not self.authenticated: raise ApiInstanceUnauthorized resource_url = self._get_url_endpoint(endpoint) timestamp = int(time.mktime(datetime.datetime.utcnow().timetuple())) nonce = self.get_nonce() authorization_header = self._make_headers( verb=verb, endpoint=endpoint, nonce=nonce, timestamp=timestamp) if verb == "GET": req = requests.get( resource_url, headers={'Authorization': authorization_header}, verify=False) elif verb == "POST": if data: req = requests.post( resource_url, headers={'Authorization': authorization_header}, data=data) elif files: req = requests.post( resource_url, headers={'Authorization': authorization_header}, files=files) elif data and files: req = requests.post( resource_url, headers={'Authorization': authorization_header}, files=files, data=data) else: req = requests.post( resource_url, headers={'Authorization': authorization_header}) if req.status_code == 401: raise ApiResponseUnauthorized(req) elif req.status_code == 404: raise NotFound404(req) elif req.status_code == 500: raise Exception(req) if self.testing: return req try: return req.json() except: print('returning req', req._content) return req @staticmethod def get_nonce(): nonce = md5( str(random.SystemRandom().randint(0, 100000000)).encode('utf8') ).hexdigest() return nonce @staticmethod def _get_image_type(image): if imghdr.what(image) == 'jpeg': return 'image/jpeg' elif imghdr.what(image) == 'gif': return 'image/gif' elif imghdr.what(image) == 'png': return 'image/png' def get_favorites(self, before=None, after=None): """ Get a list of the authenticated user's 10 most recent favorites (likes). Args: before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the authenticated user's set of Likes. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the authenticated user's set of Likes. Returns: List of SharedFile objects. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/favorites' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['favorites']] def get_user(self, user_id=None, user_name=None): """ Get a user object from the API. If no ``user_id`` or ``user_name`` is specified, it will return the User object for the currently authenticated user. Args: user_id (int): User ID of the user for whom you want to get information. [Optional] user_name(str): Username for the user for whom you want to get information. [Optional] Returns: A User object. """ if user_id: endpoint = '/api/user_id/{0}'.format(user_id) elif user_name: endpoint = '/api/user_name/{0}'.format(user_name) else: # Return currently authorized user endpoint = '/api/user' data = self._make_request(verb="GET", endpoint=endpoint) try: return User.NewFromJSON(data) except: return data def get_user_shakes(self): """ Get a list of Shake objects for the currently authenticated user. Returns: A list of Shake objects. """ endpoint = '/api/shakes' data = self._make_request(verb="GET", endpoint=endpoint) shakes = [Shake.NewFromJSON(shk) for shk in data['shakes']] return shakes def get_shared_files_from_shake(self, shake_id=None, before=None, after=None): """ Returns a list of SharedFile objects from a particular shake. Args: shake_id (int): Shake from which to get a list of SharedFiles before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the given Shake. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the given Shake. Returns: List (list) of SharedFiles. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/shakes' if shake_id: endpoint += '/{0}'.format(shake_id) if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request(verb="GET", endpoint=endpoint) return [SharedFile.NewFromJSON(f) for f in data['sharedfiles']] def get_shared_file(self, sharekey=None): """ Returns a SharedFile object given by the sharekey. Args: sharekey (str): Sharekey of the SharedFile you want to retrieve. Returns: SharedFile """ if not sharekey: raise Exception("You must specify a sharekey.") endpoint = '/api/sharedfile/{0}'.format(sharekey) data = self._make_request('GET', endpoint) return SharedFile.NewFromJSON(data) def save_shared_file(self, sharekey=None): """ Save a SharedFile to your Shake. Args: sharekey (str): Sharekey for the file to save. Returns: SharedFile saved to your shake. """ endpoint = '/api/sharedfile/{sharekey}/save'.format(sharekey=sharekey) data = self._make_request("POST", endpoint=endpoint, data=None) try: sf = SharedFile.NewFromJSON(data) sf.saved = True return sf except: raise Exception("{0}".format(data['error'])) def get_friends_shake(self, before=None, after=None): """ Contrary to the endpoint naming, this resource is for a list of SharedFiles from your friends on mlkshk. Returns: List of SharedFiles. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/friends' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['friend_shake']] def get_incoming_shake(self, before=None, after=None): """ Returns a list of the most recent SharedFiles on mlkshk.com Args: before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the Incoming Shake. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the Incoming Shake. Returns: List of SharedFile objects. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/incoming' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['incoming']] def get_magic_shake(self, before=None, after=None): """ From the API: Returns the 10 most recent files accepted by the 'magic' file selection algorithm. Currently any files with 10 or more likes are magic. Returns: List of SharedFile objects """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/magicfiles' if before: endpoint += '/before/{key}'.format(key=before) elif after: endpoint += '/after/{key}'.format(key=after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['magicfiles']] def get_comments(self, sharekey=None): """ Retrieve comments on a SharedFile Args: sharekey (str): Sharekey for the file from which you want to return the set of comments. Returns: List of Comment objects. """ if not sharekey: raise Exception( "You must specify a sharekey of the file you" "want to 'like'.") endpoint = '/api/sharedfile/{0}/comments'.format(sharekey) data = self._make_request("GET", endpoint=endpoint) return [Comment.NewFromJSON(c) for c in data['comments']] def post_comment(self, sharekey=None, comment=None): """ Post a comment on behalf of the current user to the SharedFile with the given sharekey. Args: sharekey (str): Sharekey of the SharedFile to which you'd like to post a comment. comment (str): Text of the comment to post. Returns: Comment object. """ endpoint = '/api/sharedfile/{0}/comments'.format(sharekey) post_data = {'body': comment} data = self._make_request("POST", endpoint=endpoint, data=post_data) return Comment.NewFromJSON(data) def post_shared_file(self, image_file=None, source_link=None, shake_id=None, title=None, description=None): """ Upload an image. TODO: Don't have a pro account to test (or even write) code to upload a shared filed to a particular shake. Args: image_file (str): path to an image (jpg/gif) on your computer. source_link (str): URL of a source (youtube/vine/etc.) shake_id (int): shake to which to upload the file or source_link [optional] title (str): title of the SharedFile [optional] description (str): description of the SharedFile Returns: SharedFile key. """ if image_file and source_link: raise Exception('You can only specify an image file or ' 'a source link, not both.') if not image_file and not source_link: raise Exception('You must specify an image file or a source link') content_type = self._get_image_type(image_file) if not title: title = os.path.basename(image_file) f = open(image_file, 'rb') endpoint = '/api/upload' files = {'file': (title, f, content_type)} data = self._make_request('POST', endpoint=endpoint, files=files) f.close() return data def update_shared_file(self, sharekey=None, title=None, description=None): """ Update the editable details (just the title and description) of a SharedFile. Args: sharekey (str): Sharekey of the SharedFile to update. title (Optional[str]): Title of the SharedFile. description (Optional[str]): Description of the SharedFile Returns: SharedFile on success, 404 on Sharekey not found, 403 on unauthorized. """ if not sharekey: raise Exception( "You must specify a sharekey for the sharedfile" "you wish to update.") if not (title or description): raise Exception("You must specify a title or description.") post_data = {} if title: post_data['title'] = title if description: post_data['description'] = description endpoint = '/api/sharedfile/{0}'.format(sharekey) data = self._make_request('POST', endpoint=endpoint, data=post_data) return SharedFile.NewFromJSON(data)
jeremylow/pyshk
pyshk/api.py
Api.save_shared_file
python
def save_shared_file(self, sharekey=None): endpoint = '/api/sharedfile/{sharekey}/save'.format(sharekey=sharekey) data = self._make_request("POST", endpoint=endpoint, data=None) try: sf = SharedFile.NewFromJSON(data) sf.saved = True return sf except: raise Exception("{0}".format(data['error']))
Save a SharedFile to your Shake. Args: sharekey (str): Sharekey for the file to save. Returns: SharedFile saved to your shake.
train
https://github.com/jeremylow/pyshk/blob/3ab92f6706397cde7a18367266eba9e0f1ada868/pyshk/api.py#L384-L402
[ "def _make_request(self, verb, endpoint=None, data=None, files=None):\n if not self.authenticated:\n raise ApiInstanceUnauthorized\n\n resource_url = self._get_url_endpoint(endpoint)\n\n timestamp = int(time.mktime(datetime.datetime.utcnow().timetuple()))\n nonce = self.get_nonce()\n\n authorization_header = self._make_headers(\n verb=verb,\n endpoint=endpoint,\n nonce=nonce,\n timestamp=timestamp)\n\n if verb == \"GET\":\n req = requests.get(\n resource_url,\n headers={'Authorization': authorization_header},\n verify=False)\n elif verb == \"POST\":\n if data:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n data=data)\n elif files:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n files=files)\n elif data and files:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n files=files,\n data=data)\n else:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header})\n\n if req.status_code == 401:\n raise ApiResponseUnauthorized(req)\n elif req.status_code == 404:\n raise NotFound404(req)\n elif req.status_code == 500:\n raise Exception(req)\n\n if self.testing:\n return req\n\n try:\n return req.json()\n except:\n print('returning req', req._content)\n return req\n", "def NewFromJSON(data):\n \"\"\"\n Create a new SharedFile instance from a JSON dict.\n\n Args:\n data (dict): JSON dictionary representing a SharedFile.\n\n Returns:\n A SharedFile instance.\n \"\"\"\n return SharedFile(\n sharekey=data.get('sharekey', None),\n name=data.get('name', None),\n user=User.NewFromJSON(data.get('user', None)),\n title=data.get('title', None),\n description=data.get('description', None),\n posted_at=data.get('posted_at', None),\n permalink=data.get('permalink', None),\n width=data.get('width', None),\n height=data.get('height', None),\n views=data.get('views', 0),\n likes=data.get('likes', 0),\n saves=data.get('saves', 0),\n comments=data.get('comments', None),\n nsfw=data.get('nsfw', False),\n image_url=data.get('image_url', None),\n source_url=data.get('source_url', None),\n saved=data.get('saved', False),\n liked=data.get('liked', False),\n )\n" ]
class Api(object): def __init__(self, consumer_key=None, consumer_secret=None, access_token_key=None, access_token_secret=None, base_url=None, testing=False): if base_url is None: self.base_url = 'http://mlkshk.com' else: self.base_url = base_url self.port = 80 self.authenticated = False self.testing = False if testing: self.testing = True self.consumer_key = consumer_key self.consumer_secret = consumer_secret self.access_token_key = access_token_key self.access_token_secret = access_token_secret auth_list = [consumer_key, consumer_secret, access_token_key, access_token_secret] if all(auth_list): self.authenticated = True # Set headers, client info for requests. # default_headers = {'User-Agent': 'PyShk v0.0.1'} # self.client_args = {} # self.client_args['headers'] = default_headers # Set up auth - TODO: # self.auth = None # if self.access_token_key: # token = { # 'token_type': 'mac', # 'hash_algorithm': 'hmac-sha-1', # 'access_token': self.access_token_key # } # self.auth = OAuth2(self.consumer_key, token=token) # self.client = requests.Session() # self.client.auth = self.auth def get_auth(self, redirect_uri=None): if not redirect_uri: redirect_uri = "http://localhost:8000" authentication_url = ( "https://mlkshk.com/api/authorize" "?response_type=code&client_id={key}&redirect_uri={uri}").format( key=self.consumer_key, uri=redirect_uri) access_token_url = 'https://mlkshk.com/api/token' if not self.testing: webbrowser.open(authentication_url, new=1) authorization_code = input("Enter the code from the redirected URL: ") else: authorization_code = 123456 message = { 'grant_type': "authorization_code", 'code': authorization_code, 'redirect_uri': redirect_uri, 'client_id': self.consumer_key, 'client_secret': self.consumer_secret} data = urlencode(message) req = requests.post(access_token_url, params=data, verify=False) json_resp = req.json() print(""" {full_token} >>> Your access token is: {token} >>> Your access secret is: {secret} """.format(full_token=json_resp, token=json_resp['access_token'], secret=json_resp['secret'])) self.access_token_key = json_resp['access_token'] self.access_token_secret = json_resp['secret'] def _get_url_endpoint(self, endpoint): return self.base_url + endpoint def _make_headers(self, verb=None, endpoint=None, nonce=None, timestamp=None): normalized_string = "{0}\n".format(self.access_token_key) normalized_string += "{0}\n".format(timestamp) normalized_string += "{0}\n".format(nonce) normalized_string += "{0}\n".format(verb) normalized_string += "mlkshk.com\n" normalized_string += "80\n" normalized_string += "{0}\n".format(endpoint) digest = hmac.new( self.access_token_secret.encode('ascii'), normalized_string.encode('ascii'), sha1).digest() if six.PY2: signature = base64.encodestring(digest).strip().decode('utf8') else: signature = base64.encodebytes(digest).strip().decode('utf8') auth_str = ( 'MAC token="{0}", ' 'timestamp="{1}", ' 'nonce="{2}", ' 'signature="{3}"').format( self.access_token_key, str(timestamp), nonce, signature) return auth_str def _make_request(self, verb, endpoint=None, data=None, files=None): if not self.authenticated: raise ApiInstanceUnauthorized resource_url = self._get_url_endpoint(endpoint) timestamp = int(time.mktime(datetime.datetime.utcnow().timetuple())) nonce = self.get_nonce() authorization_header = self._make_headers( verb=verb, endpoint=endpoint, nonce=nonce, timestamp=timestamp) if verb == "GET": req = requests.get( resource_url, headers={'Authorization': authorization_header}, verify=False) elif verb == "POST": if data: req = requests.post( resource_url, headers={'Authorization': authorization_header}, data=data) elif files: req = requests.post( resource_url, headers={'Authorization': authorization_header}, files=files) elif data and files: req = requests.post( resource_url, headers={'Authorization': authorization_header}, files=files, data=data) else: req = requests.post( resource_url, headers={'Authorization': authorization_header}) if req.status_code == 401: raise ApiResponseUnauthorized(req) elif req.status_code == 404: raise NotFound404(req) elif req.status_code == 500: raise Exception(req) if self.testing: return req try: return req.json() except: print('returning req', req._content) return req @staticmethod def get_nonce(): nonce = md5( str(random.SystemRandom().randint(0, 100000000)).encode('utf8') ).hexdigest() return nonce @staticmethod def _get_image_type(image): if imghdr.what(image) == 'jpeg': return 'image/jpeg' elif imghdr.what(image) == 'gif': return 'image/gif' elif imghdr.what(image) == 'png': return 'image/png' def get_favorites(self, before=None, after=None): """ Get a list of the authenticated user's 10 most recent favorites (likes). Args: before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the authenticated user's set of Likes. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the authenticated user's set of Likes. Returns: List of SharedFile objects. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/favorites' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['favorites']] def get_user(self, user_id=None, user_name=None): """ Get a user object from the API. If no ``user_id`` or ``user_name`` is specified, it will return the User object for the currently authenticated user. Args: user_id (int): User ID of the user for whom you want to get information. [Optional] user_name(str): Username for the user for whom you want to get information. [Optional] Returns: A User object. """ if user_id: endpoint = '/api/user_id/{0}'.format(user_id) elif user_name: endpoint = '/api/user_name/{0}'.format(user_name) else: # Return currently authorized user endpoint = '/api/user' data = self._make_request(verb="GET", endpoint=endpoint) try: return User.NewFromJSON(data) except: return data def get_user_shakes(self): """ Get a list of Shake objects for the currently authenticated user. Returns: A list of Shake objects. """ endpoint = '/api/shakes' data = self._make_request(verb="GET", endpoint=endpoint) shakes = [Shake.NewFromJSON(shk) for shk in data['shakes']] return shakes def get_shared_files_from_shake(self, shake_id=None, before=None, after=None): """ Returns a list of SharedFile objects from a particular shake. Args: shake_id (int): Shake from which to get a list of SharedFiles before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the given Shake. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the given Shake. Returns: List (list) of SharedFiles. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/shakes' if shake_id: endpoint += '/{0}'.format(shake_id) if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request(verb="GET", endpoint=endpoint) return [SharedFile.NewFromJSON(f) for f in data['sharedfiles']] def get_shared_file(self, sharekey=None): """ Returns a SharedFile object given by the sharekey. Args: sharekey (str): Sharekey of the SharedFile you want to retrieve. Returns: SharedFile """ if not sharekey: raise Exception("You must specify a sharekey.") endpoint = '/api/sharedfile/{0}'.format(sharekey) data = self._make_request('GET', endpoint) return SharedFile.NewFromJSON(data) def like_shared_file(self, sharekey=None): """ 'Like' a SharedFile. mlkshk doesn't allow you to unlike a sharedfile, so this is ~~permanent~~. Args: sharekey (str): Sharekey for the file you want to 'like'. Returns: Either a SharedFile on success, or an exception on error. """ if not sharekey: raise Exception( "You must specify a sharekey of the file you" "want to 'like'.") endpoint = '/api/sharedfile/{sharekey}/like'.format(sharekey=sharekey) data = self._make_request("POST", endpoint=endpoint, data=None) try: sf = SharedFile.NewFromJSON(data) sf.liked = True return sf except: raise Exception("{0}".format(data['error'])) def get_friends_shake(self, before=None, after=None): """ Contrary to the endpoint naming, this resource is for a list of SharedFiles from your friends on mlkshk. Returns: List of SharedFiles. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/friends' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['friend_shake']] def get_incoming_shake(self, before=None, after=None): """ Returns a list of the most recent SharedFiles on mlkshk.com Args: before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the Incoming Shake. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the Incoming Shake. Returns: List of SharedFile objects. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/incoming' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['incoming']] def get_magic_shake(self, before=None, after=None): """ From the API: Returns the 10 most recent files accepted by the 'magic' file selection algorithm. Currently any files with 10 or more likes are magic. Returns: List of SharedFile objects """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/magicfiles' if before: endpoint += '/before/{key}'.format(key=before) elif after: endpoint += '/after/{key}'.format(key=after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['magicfiles']] def get_comments(self, sharekey=None): """ Retrieve comments on a SharedFile Args: sharekey (str): Sharekey for the file from which you want to return the set of comments. Returns: List of Comment objects. """ if not sharekey: raise Exception( "You must specify a sharekey of the file you" "want to 'like'.") endpoint = '/api/sharedfile/{0}/comments'.format(sharekey) data = self._make_request("GET", endpoint=endpoint) return [Comment.NewFromJSON(c) for c in data['comments']] def post_comment(self, sharekey=None, comment=None): """ Post a comment on behalf of the current user to the SharedFile with the given sharekey. Args: sharekey (str): Sharekey of the SharedFile to which you'd like to post a comment. comment (str): Text of the comment to post. Returns: Comment object. """ endpoint = '/api/sharedfile/{0}/comments'.format(sharekey) post_data = {'body': comment} data = self._make_request("POST", endpoint=endpoint, data=post_data) return Comment.NewFromJSON(data) def post_shared_file(self, image_file=None, source_link=None, shake_id=None, title=None, description=None): """ Upload an image. TODO: Don't have a pro account to test (or even write) code to upload a shared filed to a particular shake. Args: image_file (str): path to an image (jpg/gif) on your computer. source_link (str): URL of a source (youtube/vine/etc.) shake_id (int): shake to which to upload the file or source_link [optional] title (str): title of the SharedFile [optional] description (str): description of the SharedFile Returns: SharedFile key. """ if image_file and source_link: raise Exception('You can only specify an image file or ' 'a source link, not both.') if not image_file and not source_link: raise Exception('You must specify an image file or a source link') content_type = self._get_image_type(image_file) if not title: title = os.path.basename(image_file) f = open(image_file, 'rb') endpoint = '/api/upload' files = {'file': (title, f, content_type)} data = self._make_request('POST', endpoint=endpoint, files=files) f.close() return data def update_shared_file(self, sharekey=None, title=None, description=None): """ Update the editable details (just the title and description) of a SharedFile. Args: sharekey (str): Sharekey of the SharedFile to update. title (Optional[str]): Title of the SharedFile. description (Optional[str]): Description of the SharedFile Returns: SharedFile on success, 404 on Sharekey not found, 403 on unauthorized. """ if not sharekey: raise Exception( "You must specify a sharekey for the sharedfile" "you wish to update.") if not (title or description): raise Exception("You must specify a title or description.") post_data = {} if title: post_data['title'] = title if description: post_data['description'] = description endpoint = '/api/sharedfile/{0}'.format(sharekey) data = self._make_request('POST', endpoint=endpoint, data=post_data) return SharedFile.NewFromJSON(data)
jeremylow/pyshk
pyshk/api.py
Api.get_friends_shake
python
def get_friends_shake(self, before=None, after=None): if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/friends' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['friend_shake']]
Contrary to the endpoint naming, this resource is for a list of SharedFiles from your friends on mlkshk. Returns: List of SharedFiles.
train
https://github.com/jeremylow/pyshk/blob/3ab92f6706397cde7a18367266eba9e0f1ada868/pyshk/api.py#L404-L423
[ "def _make_request(self, verb, endpoint=None, data=None, files=None):\n if not self.authenticated:\n raise ApiInstanceUnauthorized\n\n resource_url = self._get_url_endpoint(endpoint)\n\n timestamp = int(time.mktime(datetime.datetime.utcnow().timetuple()))\n nonce = self.get_nonce()\n\n authorization_header = self._make_headers(\n verb=verb,\n endpoint=endpoint,\n nonce=nonce,\n timestamp=timestamp)\n\n if verb == \"GET\":\n req = requests.get(\n resource_url,\n headers={'Authorization': authorization_header},\n verify=False)\n elif verb == \"POST\":\n if data:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n data=data)\n elif files:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n files=files)\n elif data and files:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n files=files,\n data=data)\n else:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header})\n\n if req.status_code == 401:\n raise ApiResponseUnauthorized(req)\n elif req.status_code == 404:\n raise NotFound404(req)\n elif req.status_code == 500:\n raise Exception(req)\n\n if self.testing:\n return req\n\n try:\n return req.json()\n except:\n print('returning req', req._content)\n return req\n" ]
class Api(object): def __init__(self, consumer_key=None, consumer_secret=None, access_token_key=None, access_token_secret=None, base_url=None, testing=False): if base_url is None: self.base_url = 'http://mlkshk.com' else: self.base_url = base_url self.port = 80 self.authenticated = False self.testing = False if testing: self.testing = True self.consumer_key = consumer_key self.consumer_secret = consumer_secret self.access_token_key = access_token_key self.access_token_secret = access_token_secret auth_list = [consumer_key, consumer_secret, access_token_key, access_token_secret] if all(auth_list): self.authenticated = True # Set headers, client info for requests. # default_headers = {'User-Agent': 'PyShk v0.0.1'} # self.client_args = {} # self.client_args['headers'] = default_headers # Set up auth - TODO: # self.auth = None # if self.access_token_key: # token = { # 'token_type': 'mac', # 'hash_algorithm': 'hmac-sha-1', # 'access_token': self.access_token_key # } # self.auth = OAuth2(self.consumer_key, token=token) # self.client = requests.Session() # self.client.auth = self.auth def get_auth(self, redirect_uri=None): if not redirect_uri: redirect_uri = "http://localhost:8000" authentication_url = ( "https://mlkshk.com/api/authorize" "?response_type=code&client_id={key}&redirect_uri={uri}").format( key=self.consumer_key, uri=redirect_uri) access_token_url = 'https://mlkshk.com/api/token' if not self.testing: webbrowser.open(authentication_url, new=1) authorization_code = input("Enter the code from the redirected URL: ") else: authorization_code = 123456 message = { 'grant_type': "authorization_code", 'code': authorization_code, 'redirect_uri': redirect_uri, 'client_id': self.consumer_key, 'client_secret': self.consumer_secret} data = urlencode(message) req = requests.post(access_token_url, params=data, verify=False) json_resp = req.json() print(""" {full_token} >>> Your access token is: {token} >>> Your access secret is: {secret} """.format(full_token=json_resp, token=json_resp['access_token'], secret=json_resp['secret'])) self.access_token_key = json_resp['access_token'] self.access_token_secret = json_resp['secret'] def _get_url_endpoint(self, endpoint): return self.base_url + endpoint def _make_headers(self, verb=None, endpoint=None, nonce=None, timestamp=None): normalized_string = "{0}\n".format(self.access_token_key) normalized_string += "{0}\n".format(timestamp) normalized_string += "{0}\n".format(nonce) normalized_string += "{0}\n".format(verb) normalized_string += "mlkshk.com\n" normalized_string += "80\n" normalized_string += "{0}\n".format(endpoint) digest = hmac.new( self.access_token_secret.encode('ascii'), normalized_string.encode('ascii'), sha1).digest() if six.PY2: signature = base64.encodestring(digest).strip().decode('utf8') else: signature = base64.encodebytes(digest).strip().decode('utf8') auth_str = ( 'MAC token="{0}", ' 'timestamp="{1}", ' 'nonce="{2}", ' 'signature="{3}"').format( self.access_token_key, str(timestamp), nonce, signature) return auth_str def _make_request(self, verb, endpoint=None, data=None, files=None): if not self.authenticated: raise ApiInstanceUnauthorized resource_url = self._get_url_endpoint(endpoint) timestamp = int(time.mktime(datetime.datetime.utcnow().timetuple())) nonce = self.get_nonce() authorization_header = self._make_headers( verb=verb, endpoint=endpoint, nonce=nonce, timestamp=timestamp) if verb == "GET": req = requests.get( resource_url, headers={'Authorization': authorization_header}, verify=False) elif verb == "POST": if data: req = requests.post( resource_url, headers={'Authorization': authorization_header}, data=data) elif files: req = requests.post( resource_url, headers={'Authorization': authorization_header}, files=files) elif data and files: req = requests.post( resource_url, headers={'Authorization': authorization_header}, files=files, data=data) else: req = requests.post( resource_url, headers={'Authorization': authorization_header}) if req.status_code == 401: raise ApiResponseUnauthorized(req) elif req.status_code == 404: raise NotFound404(req) elif req.status_code == 500: raise Exception(req) if self.testing: return req try: return req.json() except: print('returning req', req._content) return req @staticmethod def get_nonce(): nonce = md5( str(random.SystemRandom().randint(0, 100000000)).encode('utf8') ).hexdigest() return nonce @staticmethod def _get_image_type(image): if imghdr.what(image) == 'jpeg': return 'image/jpeg' elif imghdr.what(image) == 'gif': return 'image/gif' elif imghdr.what(image) == 'png': return 'image/png' def get_favorites(self, before=None, after=None): """ Get a list of the authenticated user's 10 most recent favorites (likes). Args: before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the authenticated user's set of Likes. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the authenticated user's set of Likes. Returns: List of SharedFile objects. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/favorites' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['favorites']] def get_user(self, user_id=None, user_name=None): """ Get a user object from the API. If no ``user_id`` or ``user_name`` is specified, it will return the User object for the currently authenticated user. Args: user_id (int): User ID of the user for whom you want to get information. [Optional] user_name(str): Username for the user for whom you want to get information. [Optional] Returns: A User object. """ if user_id: endpoint = '/api/user_id/{0}'.format(user_id) elif user_name: endpoint = '/api/user_name/{0}'.format(user_name) else: # Return currently authorized user endpoint = '/api/user' data = self._make_request(verb="GET", endpoint=endpoint) try: return User.NewFromJSON(data) except: return data def get_user_shakes(self): """ Get a list of Shake objects for the currently authenticated user. Returns: A list of Shake objects. """ endpoint = '/api/shakes' data = self._make_request(verb="GET", endpoint=endpoint) shakes = [Shake.NewFromJSON(shk) for shk in data['shakes']] return shakes def get_shared_files_from_shake(self, shake_id=None, before=None, after=None): """ Returns a list of SharedFile objects from a particular shake. Args: shake_id (int): Shake from which to get a list of SharedFiles before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the given Shake. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the given Shake. Returns: List (list) of SharedFiles. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/shakes' if shake_id: endpoint += '/{0}'.format(shake_id) if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request(verb="GET", endpoint=endpoint) return [SharedFile.NewFromJSON(f) for f in data['sharedfiles']] def get_shared_file(self, sharekey=None): """ Returns a SharedFile object given by the sharekey. Args: sharekey (str): Sharekey of the SharedFile you want to retrieve. Returns: SharedFile """ if not sharekey: raise Exception("You must specify a sharekey.") endpoint = '/api/sharedfile/{0}'.format(sharekey) data = self._make_request('GET', endpoint) return SharedFile.NewFromJSON(data) def like_shared_file(self, sharekey=None): """ 'Like' a SharedFile. mlkshk doesn't allow you to unlike a sharedfile, so this is ~~permanent~~. Args: sharekey (str): Sharekey for the file you want to 'like'. Returns: Either a SharedFile on success, or an exception on error. """ if not sharekey: raise Exception( "You must specify a sharekey of the file you" "want to 'like'.") endpoint = '/api/sharedfile/{sharekey}/like'.format(sharekey=sharekey) data = self._make_request("POST", endpoint=endpoint, data=None) try: sf = SharedFile.NewFromJSON(data) sf.liked = True return sf except: raise Exception("{0}".format(data['error'])) def save_shared_file(self, sharekey=None): """ Save a SharedFile to your Shake. Args: sharekey (str): Sharekey for the file to save. Returns: SharedFile saved to your shake. """ endpoint = '/api/sharedfile/{sharekey}/save'.format(sharekey=sharekey) data = self._make_request("POST", endpoint=endpoint, data=None) try: sf = SharedFile.NewFromJSON(data) sf.saved = True return sf except: raise Exception("{0}".format(data['error'])) def get_incoming_shake(self, before=None, after=None): """ Returns a list of the most recent SharedFiles on mlkshk.com Args: before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the Incoming Shake. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the Incoming Shake. Returns: List of SharedFile objects. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/incoming' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['incoming']] def get_magic_shake(self, before=None, after=None): """ From the API: Returns the 10 most recent files accepted by the 'magic' file selection algorithm. Currently any files with 10 or more likes are magic. Returns: List of SharedFile objects """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/magicfiles' if before: endpoint += '/before/{key}'.format(key=before) elif after: endpoint += '/after/{key}'.format(key=after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['magicfiles']] def get_comments(self, sharekey=None): """ Retrieve comments on a SharedFile Args: sharekey (str): Sharekey for the file from which you want to return the set of comments. Returns: List of Comment objects. """ if not sharekey: raise Exception( "You must specify a sharekey of the file you" "want to 'like'.") endpoint = '/api/sharedfile/{0}/comments'.format(sharekey) data = self._make_request("GET", endpoint=endpoint) return [Comment.NewFromJSON(c) for c in data['comments']] def post_comment(self, sharekey=None, comment=None): """ Post a comment on behalf of the current user to the SharedFile with the given sharekey. Args: sharekey (str): Sharekey of the SharedFile to which you'd like to post a comment. comment (str): Text of the comment to post. Returns: Comment object. """ endpoint = '/api/sharedfile/{0}/comments'.format(sharekey) post_data = {'body': comment} data = self._make_request("POST", endpoint=endpoint, data=post_data) return Comment.NewFromJSON(data) def post_shared_file(self, image_file=None, source_link=None, shake_id=None, title=None, description=None): """ Upload an image. TODO: Don't have a pro account to test (or even write) code to upload a shared filed to a particular shake. Args: image_file (str): path to an image (jpg/gif) on your computer. source_link (str): URL of a source (youtube/vine/etc.) shake_id (int): shake to which to upload the file or source_link [optional] title (str): title of the SharedFile [optional] description (str): description of the SharedFile Returns: SharedFile key. """ if image_file and source_link: raise Exception('You can only specify an image file or ' 'a source link, not both.') if not image_file and not source_link: raise Exception('You must specify an image file or a source link') content_type = self._get_image_type(image_file) if not title: title = os.path.basename(image_file) f = open(image_file, 'rb') endpoint = '/api/upload' files = {'file': (title, f, content_type)} data = self._make_request('POST', endpoint=endpoint, files=files) f.close() return data def update_shared_file(self, sharekey=None, title=None, description=None): """ Update the editable details (just the title and description) of a SharedFile. Args: sharekey (str): Sharekey of the SharedFile to update. title (Optional[str]): Title of the SharedFile. description (Optional[str]): Description of the SharedFile Returns: SharedFile on success, 404 on Sharekey not found, 403 on unauthorized. """ if not sharekey: raise Exception( "You must specify a sharekey for the sharedfile" "you wish to update.") if not (title or description): raise Exception("You must specify a title or description.") post_data = {} if title: post_data['title'] = title if description: post_data['description'] = description endpoint = '/api/sharedfile/{0}'.format(sharekey) data = self._make_request('POST', endpoint=endpoint, data=post_data) return SharedFile.NewFromJSON(data)
jeremylow/pyshk
pyshk/api.py
Api.get_magic_shake
python
def get_magic_shake(self, before=None, after=None): if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/magicfiles' if before: endpoint += '/before/{key}'.format(key=before) elif after: endpoint += '/after/{key}'.format(key=after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['magicfiles']]
From the API: Returns the 10 most recent files accepted by the 'magic' file selection algorithm. Currently any files with 10 or more likes are magic. Returns: List of SharedFile objects
train
https://github.com/jeremylow/pyshk/blob/3ab92f6706397cde7a18367266eba9e0f1ada868/pyshk/api.py#L451-L472
[ "def _make_request(self, verb, endpoint=None, data=None, files=None):\n if not self.authenticated:\n raise ApiInstanceUnauthorized\n\n resource_url = self._get_url_endpoint(endpoint)\n\n timestamp = int(time.mktime(datetime.datetime.utcnow().timetuple()))\n nonce = self.get_nonce()\n\n authorization_header = self._make_headers(\n verb=verb,\n endpoint=endpoint,\n nonce=nonce,\n timestamp=timestamp)\n\n if verb == \"GET\":\n req = requests.get(\n resource_url,\n headers={'Authorization': authorization_header},\n verify=False)\n elif verb == \"POST\":\n if data:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n data=data)\n elif files:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n files=files)\n elif data and files:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n files=files,\n data=data)\n else:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header})\n\n if req.status_code == 401:\n raise ApiResponseUnauthorized(req)\n elif req.status_code == 404:\n raise NotFound404(req)\n elif req.status_code == 500:\n raise Exception(req)\n\n if self.testing:\n return req\n\n try:\n return req.json()\n except:\n print('returning req', req._content)\n return req\n" ]
class Api(object): def __init__(self, consumer_key=None, consumer_secret=None, access_token_key=None, access_token_secret=None, base_url=None, testing=False): if base_url is None: self.base_url = 'http://mlkshk.com' else: self.base_url = base_url self.port = 80 self.authenticated = False self.testing = False if testing: self.testing = True self.consumer_key = consumer_key self.consumer_secret = consumer_secret self.access_token_key = access_token_key self.access_token_secret = access_token_secret auth_list = [consumer_key, consumer_secret, access_token_key, access_token_secret] if all(auth_list): self.authenticated = True # Set headers, client info for requests. # default_headers = {'User-Agent': 'PyShk v0.0.1'} # self.client_args = {} # self.client_args['headers'] = default_headers # Set up auth - TODO: # self.auth = None # if self.access_token_key: # token = { # 'token_type': 'mac', # 'hash_algorithm': 'hmac-sha-1', # 'access_token': self.access_token_key # } # self.auth = OAuth2(self.consumer_key, token=token) # self.client = requests.Session() # self.client.auth = self.auth def get_auth(self, redirect_uri=None): if not redirect_uri: redirect_uri = "http://localhost:8000" authentication_url = ( "https://mlkshk.com/api/authorize" "?response_type=code&client_id={key}&redirect_uri={uri}").format( key=self.consumer_key, uri=redirect_uri) access_token_url = 'https://mlkshk.com/api/token' if not self.testing: webbrowser.open(authentication_url, new=1) authorization_code = input("Enter the code from the redirected URL: ") else: authorization_code = 123456 message = { 'grant_type': "authorization_code", 'code': authorization_code, 'redirect_uri': redirect_uri, 'client_id': self.consumer_key, 'client_secret': self.consumer_secret} data = urlencode(message) req = requests.post(access_token_url, params=data, verify=False) json_resp = req.json() print(""" {full_token} >>> Your access token is: {token} >>> Your access secret is: {secret} """.format(full_token=json_resp, token=json_resp['access_token'], secret=json_resp['secret'])) self.access_token_key = json_resp['access_token'] self.access_token_secret = json_resp['secret'] def _get_url_endpoint(self, endpoint): return self.base_url + endpoint def _make_headers(self, verb=None, endpoint=None, nonce=None, timestamp=None): normalized_string = "{0}\n".format(self.access_token_key) normalized_string += "{0}\n".format(timestamp) normalized_string += "{0}\n".format(nonce) normalized_string += "{0}\n".format(verb) normalized_string += "mlkshk.com\n" normalized_string += "80\n" normalized_string += "{0}\n".format(endpoint) digest = hmac.new( self.access_token_secret.encode('ascii'), normalized_string.encode('ascii'), sha1).digest() if six.PY2: signature = base64.encodestring(digest).strip().decode('utf8') else: signature = base64.encodebytes(digest).strip().decode('utf8') auth_str = ( 'MAC token="{0}", ' 'timestamp="{1}", ' 'nonce="{2}", ' 'signature="{3}"').format( self.access_token_key, str(timestamp), nonce, signature) return auth_str def _make_request(self, verb, endpoint=None, data=None, files=None): if not self.authenticated: raise ApiInstanceUnauthorized resource_url = self._get_url_endpoint(endpoint) timestamp = int(time.mktime(datetime.datetime.utcnow().timetuple())) nonce = self.get_nonce() authorization_header = self._make_headers( verb=verb, endpoint=endpoint, nonce=nonce, timestamp=timestamp) if verb == "GET": req = requests.get( resource_url, headers={'Authorization': authorization_header}, verify=False) elif verb == "POST": if data: req = requests.post( resource_url, headers={'Authorization': authorization_header}, data=data) elif files: req = requests.post( resource_url, headers={'Authorization': authorization_header}, files=files) elif data and files: req = requests.post( resource_url, headers={'Authorization': authorization_header}, files=files, data=data) else: req = requests.post( resource_url, headers={'Authorization': authorization_header}) if req.status_code == 401: raise ApiResponseUnauthorized(req) elif req.status_code == 404: raise NotFound404(req) elif req.status_code == 500: raise Exception(req) if self.testing: return req try: return req.json() except: print('returning req', req._content) return req @staticmethod def get_nonce(): nonce = md5( str(random.SystemRandom().randint(0, 100000000)).encode('utf8') ).hexdigest() return nonce @staticmethod def _get_image_type(image): if imghdr.what(image) == 'jpeg': return 'image/jpeg' elif imghdr.what(image) == 'gif': return 'image/gif' elif imghdr.what(image) == 'png': return 'image/png' def get_favorites(self, before=None, after=None): """ Get a list of the authenticated user's 10 most recent favorites (likes). Args: before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the authenticated user's set of Likes. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the authenticated user's set of Likes. Returns: List of SharedFile objects. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/favorites' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['favorites']] def get_user(self, user_id=None, user_name=None): """ Get a user object from the API. If no ``user_id`` or ``user_name`` is specified, it will return the User object for the currently authenticated user. Args: user_id (int): User ID of the user for whom you want to get information. [Optional] user_name(str): Username for the user for whom you want to get information. [Optional] Returns: A User object. """ if user_id: endpoint = '/api/user_id/{0}'.format(user_id) elif user_name: endpoint = '/api/user_name/{0}'.format(user_name) else: # Return currently authorized user endpoint = '/api/user' data = self._make_request(verb="GET", endpoint=endpoint) try: return User.NewFromJSON(data) except: return data def get_user_shakes(self): """ Get a list of Shake objects for the currently authenticated user. Returns: A list of Shake objects. """ endpoint = '/api/shakes' data = self._make_request(verb="GET", endpoint=endpoint) shakes = [Shake.NewFromJSON(shk) for shk in data['shakes']] return shakes def get_shared_files_from_shake(self, shake_id=None, before=None, after=None): """ Returns a list of SharedFile objects from a particular shake. Args: shake_id (int): Shake from which to get a list of SharedFiles before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the given Shake. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the given Shake. Returns: List (list) of SharedFiles. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/shakes' if shake_id: endpoint += '/{0}'.format(shake_id) if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request(verb="GET", endpoint=endpoint) return [SharedFile.NewFromJSON(f) for f in data['sharedfiles']] def get_shared_file(self, sharekey=None): """ Returns a SharedFile object given by the sharekey. Args: sharekey (str): Sharekey of the SharedFile you want to retrieve. Returns: SharedFile """ if not sharekey: raise Exception("You must specify a sharekey.") endpoint = '/api/sharedfile/{0}'.format(sharekey) data = self._make_request('GET', endpoint) return SharedFile.NewFromJSON(data) def like_shared_file(self, sharekey=None): """ 'Like' a SharedFile. mlkshk doesn't allow you to unlike a sharedfile, so this is ~~permanent~~. Args: sharekey (str): Sharekey for the file you want to 'like'. Returns: Either a SharedFile on success, or an exception on error. """ if not sharekey: raise Exception( "You must specify a sharekey of the file you" "want to 'like'.") endpoint = '/api/sharedfile/{sharekey}/like'.format(sharekey=sharekey) data = self._make_request("POST", endpoint=endpoint, data=None) try: sf = SharedFile.NewFromJSON(data) sf.liked = True return sf except: raise Exception("{0}".format(data['error'])) def save_shared_file(self, sharekey=None): """ Save a SharedFile to your Shake. Args: sharekey (str): Sharekey for the file to save. Returns: SharedFile saved to your shake. """ endpoint = '/api/sharedfile/{sharekey}/save'.format(sharekey=sharekey) data = self._make_request("POST", endpoint=endpoint, data=None) try: sf = SharedFile.NewFromJSON(data) sf.saved = True return sf except: raise Exception("{0}".format(data['error'])) def get_friends_shake(self, before=None, after=None): """ Contrary to the endpoint naming, this resource is for a list of SharedFiles from your friends on mlkshk. Returns: List of SharedFiles. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/friends' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['friend_shake']] def get_incoming_shake(self, before=None, after=None): """ Returns a list of the most recent SharedFiles on mlkshk.com Args: before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the Incoming Shake. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the Incoming Shake. Returns: List of SharedFile objects. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/incoming' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['incoming']] def get_comments(self, sharekey=None): """ Retrieve comments on a SharedFile Args: sharekey (str): Sharekey for the file from which you want to return the set of comments. Returns: List of Comment objects. """ if not sharekey: raise Exception( "You must specify a sharekey of the file you" "want to 'like'.") endpoint = '/api/sharedfile/{0}/comments'.format(sharekey) data = self._make_request("GET", endpoint=endpoint) return [Comment.NewFromJSON(c) for c in data['comments']] def post_comment(self, sharekey=None, comment=None): """ Post a comment on behalf of the current user to the SharedFile with the given sharekey. Args: sharekey (str): Sharekey of the SharedFile to which you'd like to post a comment. comment (str): Text of the comment to post. Returns: Comment object. """ endpoint = '/api/sharedfile/{0}/comments'.format(sharekey) post_data = {'body': comment} data = self._make_request("POST", endpoint=endpoint, data=post_data) return Comment.NewFromJSON(data) def post_shared_file(self, image_file=None, source_link=None, shake_id=None, title=None, description=None): """ Upload an image. TODO: Don't have a pro account to test (or even write) code to upload a shared filed to a particular shake. Args: image_file (str): path to an image (jpg/gif) on your computer. source_link (str): URL of a source (youtube/vine/etc.) shake_id (int): shake to which to upload the file or source_link [optional] title (str): title of the SharedFile [optional] description (str): description of the SharedFile Returns: SharedFile key. """ if image_file and source_link: raise Exception('You can only specify an image file or ' 'a source link, not both.') if not image_file and not source_link: raise Exception('You must specify an image file or a source link') content_type = self._get_image_type(image_file) if not title: title = os.path.basename(image_file) f = open(image_file, 'rb') endpoint = '/api/upload' files = {'file': (title, f, content_type)} data = self._make_request('POST', endpoint=endpoint, files=files) f.close() return data def update_shared_file(self, sharekey=None, title=None, description=None): """ Update the editable details (just the title and description) of a SharedFile. Args: sharekey (str): Sharekey of the SharedFile to update. title (Optional[str]): Title of the SharedFile. description (Optional[str]): Description of the SharedFile Returns: SharedFile on success, 404 on Sharekey not found, 403 on unauthorized. """ if not sharekey: raise Exception( "You must specify a sharekey for the sharedfile" "you wish to update.") if not (title or description): raise Exception("You must specify a title or description.") post_data = {} if title: post_data['title'] = title if description: post_data['description'] = description endpoint = '/api/sharedfile/{0}'.format(sharekey) data = self._make_request('POST', endpoint=endpoint, data=post_data) return SharedFile.NewFromJSON(data)
jeremylow/pyshk
pyshk/api.py
Api.get_comments
python
def get_comments(self, sharekey=None): if not sharekey: raise Exception( "You must specify a sharekey of the file you" "want to 'like'.") endpoint = '/api/sharedfile/{0}/comments'.format(sharekey) data = self._make_request("GET", endpoint=endpoint) return [Comment.NewFromJSON(c) for c in data['comments']]
Retrieve comments on a SharedFile Args: sharekey (str): Sharekey for the file from which you want to return the set of comments. Returns: List of Comment objects.
train
https://github.com/jeremylow/pyshk/blob/3ab92f6706397cde7a18367266eba9e0f1ada868/pyshk/api.py#L474-L494
[ "def _make_request(self, verb, endpoint=None, data=None, files=None):\n if not self.authenticated:\n raise ApiInstanceUnauthorized\n\n resource_url = self._get_url_endpoint(endpoint)\n\n timestamp = int(time.mktime(datetime.datetime.utcnow().timetuple()))\n nonce = self.get_nonce()\n\n authorization_header = self._make_headers(\n verb=verb,\n endpoint=endpoint,\n nonce=nonce,\n timestamp=timestamp)\n\n if verb == \"GET\":\n req = requests.get(\n resource_url,\n headers={'Authorization': authorization_header},\n verify=False)\n elif verb == \"POST\":\n if data:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n data=data)\n elif files:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n files=files)\n elif data and files:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n files=files,\n data=data)\n else:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header})\n\n if req.status_code == 401:\n raise ApiResponseUnauthorized(req)\n elif req.status_code == 404:\n raise NotFound404(req)\n elif req.status_code == 500:\n raise Exception(req)\n\n if self.testing:\n return req\n\n try:\n return req.json()\n except:\n print('returning req', req._content)\n return req\n" ]
class Api(object): def __init__(self, consumer_key=None, consumer_secret=None, access_token_key=None, access_token_secret=None, base_url=None, testing=False): if base_url is None: self.base_url = 'http://mlkshk.com' else: self.base_url = base_url self.port = 80 self.authenticated = False self.testing = False if testing: self.testing = True self.consumer_key = consumer_key self.consumer_secret = consumer_secret self.access_token_key = access_token_key self.access_token_secret = access_token_secret auth_list = [consumer_key, consumer_secret, access_token_key, access_token_secret] if all(auth_list): self.authenticated = True # Set headers, client info for requests. # default_headers = {'User-Agent': 'PyShk v0.0.1'} # self.client_args = {} # self.client_args['headers'] = default_headers # Set up auth - TODO: # self.auth = None # if self.access_token_key: # token = { # 'token_type': 'mac', # 'hash_algorithm': 'hmac-sha-1', # 'access_token': self.access_token_key # } # self.auth = OAuth2(self.consumer_key, token=token) # self.client = requests.Session() # self.client.auth = self.auth def get_auth(self, redirect_uri=None): if not redirect_uri: redirect_uri = "http://localhost:8000" authentication_url = ( "https://mlkshk.com/api/authorize" "?response_type=code&client_id={key}&redirect_uri={uri}").format( key=self.consumer_key, uri=redirect_uri) access_token_url = 'https://mlkshk.com/api/token' if not self.testing: webbrowser.open(authentication_url, new=1) authorization_code = input("Enter the code from the redirected URL: ") else: authorization_code = 123456 message = { 'grant_type': "authorization_code", 'code': authorization_code, 'redirect_uri': redirect_uri, 'client_id': self.consumer_key, 'client_secret': self.consumer_secret} data = urlencode(message) req = requests.post(access_token_url, params=data, verify=False) json_resp = req.json() print(""" {full_token} >>> Your access token is: {token} >>> Your access secret is: {secret} """.format(full_token=json_resp, token=json_resp['access_token'], secret=json_resp['secret'])) self.access_token_key = json_resp['access_token'] self.access_token_secret = json_resp['secret'] def _get_url_endpoint(self, endpoint): return self.base_url + endpoint def _make_headers(self, verb=None, endpoint=None, nonce=None, timestamp=None): normalized_string = "{0}\n".format(self.access_token_key) normalized_string += "{0}\n".format(timestamp) normalized_string += "{0}\n".format(nonce) normalized_string += "{0}\n".format(verb) normalized_string += "mlkshk.com\n" normalized_string += "80\n" normalized_string += "{0}\n".format(endpoint) digest = hmac.new( self.access_token_secret.encode('ascii'), normalized_string.encode('ascii'), sha1).digest() if six.PY2: signature = base64.encodestring(digest).strip().decode('utf8') else: signature = base64.encodebytes(digest).strip().decode('utf8') auth_str = ( 'MAC token="{0}", ' 'timestamp="{1}", ' 'nonce="{2}", ' 'signature="{3}"').format( self.access_token_key, str(timestamp), nonce, signature) return auth_str def _make_request(self, verb, endpoint=None, data=None, files=None): if not self.authenticated: raise ApiInstanceUnauthorized resource_url = self._get_url_endpoint(endpoint) timestamp = int(time.mktime(datetime.datetime.utcnow().timetuple())) nonce = self.get_nonce() authorization_header = self._make_headers( verb=verb, endpoint=endpoint, nonce=nonce, timestamp=timestamp) if verb == "GET": req = requests.get( resource_url, headers={'Authorization': authorization_header}, verify=False) elif verb == "POST": if data: req = requests.post( resource_url, headers={'Authorization': authorization_header}, data=data) elif files: req = requests.post( resource_url, headers={'Authorization': authorization_header}, files=files) elif data and files: req = requests.post( resource_url, headers={'Authorization': authorization_header}, files=files, data=data) else: req = requests.post( resource_url, headers={'Authorization': authorization_header}) if req.status_code == 401: raise ApiResponseUnauthorized(req) elif req.status_code == 404: raise NotFound404(req) elif req.status_code == 500: raise Exception(req) if self.testing: return req try: return req.json() except: print('returning req', req._content) return req @staticmethod def get_nonce(): nonce = md5( str(random.SystemRandom().randint(0, 100000000)).encode('utf8') ).hexdigest() return nonce @staticmethod def _get_image_type(image): if imghdr.what(image) == 'jpeg': return 'image/jpeg' elif imghdr.what(image) == 'gif': return 'image/gif' elif imghdr.what(image) == 'png': return 'image/png' def get_favorites(self, before=None, after=None): """ Get a list of the authenticated user's 10 most recent favorites (likes). Args: before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the authenticated user's set of Likes. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the authenticated user's set of Likes. Returns: List of SharedFile objects. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/favorites' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['favorites']] def get_user(self, user_id=None, user_name=None): """ Get a user object from the API. If no ``user_id`` or ``user_name`` is specified, it will return the User object for the currently authenticated user. Args: user_id (int): User ID of the user for whom you want to get information. [Optional] user_name(str): Username for the user for whom you want to get information. [Optional] Returns: A User object. """ if user_id: endpoint = '/api/user_id/{0}'.format(user_id) elif user_name: endpoint = '/api/user_name/{0}'.format(user_name) else: # Return currently authorized user endpoint = '/api/user' data = self._make_request(verb="GET", endpoint=endpoint) try: return User.NewFromJSON(data) except: return data def get_user_shakes(self): """ Get a list of Shake objects for the currently authenticated user. Returns: A list of Shake objects. """ endpoint = '/api/shakes' data = self._make_request(verb="GET", endpoint=endpoint) shakes = [Shake.NewFromJSON(shk) for shk in data['shakes']] return shakes def get_shared_files_from_shake(self, shake_id=None, before=None, after=None): """ Returns a list of SharedFile objects from a particular shake. Args: shake_id (int): Shake from which to get a list of SharedFiles before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the given Shake. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the given Shake. Returns: List (list) of SharedFiles. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/shakes' if shake_id: endpoint += '/{0}'.format(shake_id) if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request(verb="GET", endpoint=endpoint) return [SharedFile.NewFromJSON(f) for f in data['sharedfiles']] def get_shared_file(self, sharekey=None): """ Returns a SharedFile object given by the sharekey. Args: sharekey (str): Sharekey of the SharedFile you want to retrieve. Returns: SharedFile """ if not sharekey: raise Exception("You must specify a sharekey.") endpoint = '/api/sharedfile/{0}'.format(sharekey) data = self._make_request('GET', endpoint) return SharedFile.NewFromJSON(data) def like_shared_file(self, sharekey=None): """ 'Like' a SharedFile. mlkshk doesn't allow you to unlike a sharedfile, so this is ~~permanent~~. Args: sharekey (str): Sharekey for the file you want to 'like'. Returns: Either a SharedFile on success, or an exception on error. """ if not sharekey: raise Exception( "You must specify a sharekey of the file you" "want to 'like'.") endpoint = '/api/sharedfile/{sharekey}/like'.format(sharekey=sharekey) data = self._make_request("POST", endpoint=endpoint, data=None) try: sf = SharedFile.NewFromJSON(data) sf.liked = True return sf except: raise Exception("{0}".format(data['error'])) def save_shared_file(self, sharekey=None): """ Save a SharedFile to your Shake. Args: sharekey (str): Sharekey for the file to save. Returns: SharedFile saved to your shake. """ endpoint = '/api/sharedfile/{sharekey}/save'.format(sharekey=sharekey) data = self._make_request("POST", endpoint=endpoint, data=None) try: sf = SharedFile.NewFromJSON(data) sf.saved = True return sf except: raise Exception("{0}".format(data['error'])) def get_friends_shake(self, before=None, after=None): """ Contrary to the endpoint naming, this resource is for a list of SharedFiles from your friends on mlkshk. Returns: List of SharedFiles. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/friends' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['friend_shake']] def get_incoming_shake(self, before=None, after=None): """ Returns a list of the most recent SharedFiles on mlkshk.com Args: before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the Incoming Shake. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the Incoming Shake. Returns: List of SharedFile objects. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/incoming' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['incoming']] def get_magic_shake(self, before=None, after=None): """ From the API: Returns the 10 most recent files accepted by the 'magic' file selection algorithm. Currently any files with 10 or more likes are magic. Returns: List of SharedFile objects """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/magicfiles' if before: endpoint += '/before/{key}'.format(key=before) elif after: endpoint += '/after/{key}'.format(key=after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['magicfiles']] def post_comment(self, sharekey=None, comment=None): """ Post a comment on behalf of the current user to the SharedFile with the given sharekey. Args: sharekey (str): Sharekey of the SharedFile to which you'd like to post a comment. comment (str): Text of the comment to post. Returns: Comment object. """ endpoint = '/api/sharedfile/{0}/comments'.format(sharekey) post_data = {'body': comment} data = self._make_request("POST", endpoint=endpoint, data=post_data) return Comment.NewFromJSON(data) def post_shared_file(self, image_file=None, source_link=None, shake_id=None, title=None, description=None): """ Upload an image. TODO: Don't have a pro account to test (or even write) code to upload a shared filed to a particular shake. Args: image_file (str): path to an image (jpg/gif) on your computer. source_link (str): URL of a source (youtube/vine/etc.) shake_id (int): shake to which to upload the file or source_link [optional] title (str): title of the SharedFile [optional] description (str): description of the SharedFile Returns: SharedFile key. """ if image_file and source_link: raise Exception('You can only specify an image file or ' 'a source link, not both.') if not image_file and not source_link: raise Exception('You must specify an image file or a source link') content_type = self._get_image_type(image_file) if not title: title = os.path.basename(image_file) f = open(image_file, 'rb') endpoint = '/api/upload' files = {'file': (title, f, content_type)} data = self._make_request('POST', endpoint=endpoint, files=files) f.close() return data def update_shared_file(self, sharekey=None, title=None, description=None): """ Update the editable details (just the title and description) of a SharedFile. Args: sharekey (str): Sharekey of the SharedFile to update. title (Optional[str]): Title of the SharedFile. description (Optional[str]): Description of the SharedFile Returns: SharedFile on success, 404 on Sharekey not found, 403 on unauthorized. """ if not sharekey: raise Exception( "You must specify a sharekey for the sharedfile" "you wish to update.") if not (title or description): raise Exception("You must specify a title or description.") post_data = {} if title: post_data['title'] = title if description: post_data['description'] = description endpoint = '/api/sharedfile/{0}'.format(sharekey) data = self._make_request('POST', endpoint=endpoint, data=post_data) return SharedFile.NewFromJSON(data)
jeremylow/pyshk
pyshk/api.py
Api.post_comment
python
def post_comment(self, sharekey=None, comment=None): endpoint = '/api/sharedfile/{0}/comments'.format(sharekey) post_data = {'body': comment} data = self._make_request("POST", endpoint=endpoint, data=post_data) return Comment.NewFromJSON(data)
Post a comment on behalf of the current user to the SharedFile with the given sharekey. Args: sharekey (str): Sharekey of the SharedFile to which you'd like to post a comment. comment (str): Text of the comment to post. Returns: Comment object.
train
https://github.com/jeremylow/pyshk/blob/3ab92f6706397cde7a18367266eba9e0f1ada868/pyshk/api.py#L496-L514
[ "def _make_request(self, verb, endpoint=None, data=None, files=None):\n if not self.authenticated:\n raise ApiInstanceUnauthorized\n\n resource_url = self._get_url_endpoint(endpoint)\n\n timestamp = int(time.mktime(datetime.datetime.utcnow().timetuple()))\n nonce = self.get_nonce()\n\n authorization_header = self._make_headers(\n verb=verb,\n endpoint=endpoint,\n nonce=nonce,\n timestamp=timestamp)\n\n if verb == \"GET\":\n req = requests.get(\n resource_url,\n headers={'Authorization': authorization_header},\n verify=False)\n elif verb == \"POST\":\n if data:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n data=data)\n elif files:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n files=files)\n elif data and files:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n files=files,\n data=data)\n else:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header})\n\n if req.status_code == 401:\n raise ApiResponseUnauthorized(req)\n elif req.status_code == 404:\n raise NotFound404(req)\n elif req.status_code == 500:\n raise Exception(req)\n\n if self.testing:\n return req\n\n try:\n return req.json()\n except:\n print('returning req', req._content)\n return req\n", "def NewFromJSON(data):\n \"\"\"\n Create a new Comment instance from a JSON dict.\n\n Args:\n data (dict): JSON dictionary representing a Comment.\n\n Returns:\n A Comment instance.\n \"\"\"\n return Comment(\n body=data.get('body', None),\n posted_at=data.get('posted_at', None),\n user=User.NewFromJSON(data.get('user', None))\n )\n" ]
class Api(object): def __init__(self, consumer_key=None, consumer_secret=None, access_token_key=None, access_token_secret=None, base_url=None, testing=False): if base_url is None: self.base_url = 'http://mlkshk.com' else: self.base_url = base_url self.port = 80 self.authenticated = False self.testing = False if testing: self.testing = True self.consumer_key = consumer_key self.consumer_secret = consumer_secret self.access_token_key = access_token_key self.access_token_secret = access_token_secret auth_list = [consumer_key, consumer_secret, access_token_key, access_token_secret] if all(auth_list): self.authenticated = True # Set headers, client info for requests. # default_headers = {'User-Agent': 'PyShk v0.0.1'} # self.client_args = {} # self.client_args['headers'] = default_headers # Set up auth - TODO: # self.auth = None # if self.access_token_key: # token = { # 'token_type': 'mac', # 'hash_algorithm': 'hmac-sha-1', # 'access_token': self.access_token_key # } # self.auth = OAuth2(self.consumer_key, token=token) # self.client = requests.Session() # self.client.auth = self.auth def get_auth(self, redirect_uri=None): if not redirect_uri: redirect_uri = "http://localhost:8000" authentication_url = ( "https://mlkshk.com/api/authorize" "?response_type=code&client_id={key}&redirect_uri={uri}").format( key=self.consumer_key, uri=redirect_uri) access_token_url = 'https://mlkshk.com/api/token' if not self.testing: webbrowser.open(authentication_url, new=1) authorization_code = input("Enter the code from the redirected URL: ") else: authorization_code = 123456 message = { 'grant_type': "authorization_code", 'code': authorization_code, 'redirect_uri': redirect_uri, 'client_id': self.consumer_key, 'client_secret': self.consumer_secret} data = urlencode(message) req = requests.post(access_token_url, params=data, verify=False) json_resp = req.json() print(""" {full_token} >>> Your access token is: {token} >>> Your access secret is: {secret} """.format(full_token=json_resp, token=json_resp['access_token'], secret=json_resp['secret'])) self.access_token_key = json_resp['access_token'] self.access_token_secret = json_resp['secret'] def _get_url_endpoint(self, endpoint): return self.base_url + endpoint def _make_headers(self, verb=None, endpoint=None, nonce=None, timestamp=None): normalized_string = "{0}\n".format(self.access_token_key) normalized_string += "{0}\n".format(timestamp) normalized_string += "{0}\n".format(nonce) normalized_string += "{0}\n".format(verb) normalized_string += "mlkshk.com\n" normalized_string += "80\n" normalized_string += "{0}\n".format(endpoint) digest = hmac.new( self.access_token_secret.encode('ascii'), normalized_string.encode('ascii'), sha1).digest() if six.PY2: signature = base64.encodestring(digest).strip().decode('utf8') else: signature = base64.encodebytes(digest).strip().decode('utf8') auth_str = ( 'MAC token="{0}", ' 'timestamp="{1}", ' 'nonce="{2}", ' 'signature="{3}"').format( self.access_token_key, str(timestamp), nonce, signature) return auth_str def _make_request(self, verb, endpoint=None, data=None, files=None): if not self.authenticated: raise ApiInstanceUnauthorized resource_url = self._get_url_endpoint(endpoint) timestamp = int(time.mktime(datetime.datetime.utcnow().timetuple())) nonce = self.get_nonce() authorization_header = self._make_headers( verb=verb, endpoint=endpoint, nonce=nonce, timestamp=timestamp) if verb == "GET": req = requests.get( resource_url, headers={'Authorization': authorization_header}, verify=False) elif verb == "POST": if data: req = requests.post( resource_url, headers={'Authorization': authorization_header}, data=data) elif files: req = requests.post( resource_url, headers={'Authorization': authorization_header}, files=files) elif data and files: req = requests.post( resource_url, headers={'Authorization': authorization_header}, files=files, data=data) else: req = requests.post( resource_url, headers={'Authorization': authorization_header}) if req.status_code == 401: raise ApiResponseUnauthorized(req) elif req.status_code == 404: raise NotFound404(req) elif req.status_code == 500: raise Exception(req) if self.testing: return req try: return req.json() except: print('returning req', req._content) return req @staticmethod def get_nonce(): nonce = md5( str(random.SystemRandom().randint(0, 100000000)).encode('utf8') ).hexdigest() return nonce @staticmethod def _get_image_type(image): if imghdr.what(image) == 'jpeg': return 'image/jpeg' elif imghdr.what(image) == 'gif': return 'image/gif' elif imghdr.what(image) == 'png': return 'image/png' def get_favorites(self, before=None, after=None): """ Get a list of the authenticated user's 10 most recent favorites (likes). Args: before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the authenticated user's set of Likes. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the authenticated user's set of Likes. Returns: List of SharedFile objects. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/favorites' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['favorites']] def get_user(self, user_id=None, user_name=None): """ Get a user object from the API. If no ``user_id`` or ``user_name`` is specified, it will return the User object for the currently authenticated user. Args: user_id (int): User ID of the user for whom you want to get information. [Optional] user_name(str): Username for the user for whom you want to get information. [Optional] Returns: A User object. """ if user_id: endpoint = '/api/user_id/{0}'.format(user_id) elif user_name: endpoint = '/api/user_name/{0}'.format(user_name) else: # Return currently authorized user endpoint = '/api/user' data = self._make_request(verb="GET", endpoint=endpoint) try: return User.NewFromJSON(data) except: return data def get_user_shakes(self): """ Get a list of Shake objects for the currently authenticated user. Returns: A list of Shake objects. """ endpoint = '/api/shakes' data = self._make_request(verb="GET", endpoint=endpoint) shakes = [Shake.NewFromJSON(shk) for shk in data['shakes']] return shakes def get_shared_files_from_shake(self, shake_id=None, before=None, after=None): """ Returns a list of SharedFile objects from a particular shake. Args: shake_id (int): Shake from which to get a list of SharedFiles before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the given Shake. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the given Shake. Returns: List (list) of SharedFiles. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/shakes' if shake_id: endpoint += '/{0}'.format(shake_id) if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request(verb="GET", endpoint=endpoint) return [SharedFile.NewFromJSON(f) for f in data['sharedfiles']] def get_shared_file(self, sharekey=None): """ Returns a SharedFile object given by the sharekey. Args: sharekey (str): Sharekey of the SharedFile you want to retrieve. Returns: SharedFile """ if not sharekey: raise Exception("You must specify a sharekey.") endpoint = '/api/sharedfile/{0}'.format(sharekey) data = self._make_request('GET', endpoint) return SharedFile.NewFromJSON(data) def like_shared_file(self, sharekey=None): """ 'Like' a SharedFile. mlkshk doesn't allow you to unlike a sharedfile, so this is ~~permanent~~. Args: sharekey (str): Sharekey for the file you want to 'like'. Returns: Either a SharedFile on success, or an exception on error. """ if not sharekey: raise Exception( "You must specify a sharekey of the file you" "want to 'like'.") endpoint = '/api/sharedfile/{sharekey}/like'.format(sharekey=sharekey) data = self._make_request("POST", endpoint=endpoint, data=None) try: sf = SharedFile.NewFromJSON(data) sf.liked = True return sf except: raise Exception("{0}".format(data['error'])) def save_shared_file(self, sharekey=None): """ Save a SharedFile to your Shake. Args: sharekey (str): Sharekey for the file to save. Returns: SharedFile saved to your shake. """ endpoint = '/api/sharedfile/{sharekey}/save'.format(sharekey=sharekey) data = self._make_request("POST", endpoint=endpoint, data=None) try: sf = SharedFile.NewFromJSON(data) sf.saved = True return sf except: raise Exception("{0}".format(data['error'])) def get_friends_shake(self, before=None, after=None): """ Contrary to the endpoint naming, this resource is for a list of SharedFiles from your friends on mlkshk. Returns: List of SharedFiles. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/friends' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['friend_shake']] def get_incoming_shake(self, before=None, after=None): """ Returns a list of the most recent SharedFiles on mlkshk.com Args: before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the Incoming Shake. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the Incoming Shake. Returns: List of SharedFile objects. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/incoming' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['incoming']] def get_magic_shake(self, before=None, after=None): """ From the API: Returns the 10 most recent files accepted by the 'magic' file selection algorithm. Currently any files with 10 or more likes are magic. Returns: List of SharedFile objects """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/magicfiles' if before: endpoint += '/before/{key}'.format(key=before) elif after: endpoint += '/after/{key}'.format(key=after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['magicfiles']] def get_comments(self, sharekey=None): """ Retrieve comments on a SharedFile Args: sharekey (str): Sharekey for the file from which you want to return the set of comments. Returns: List of Comment objects. """ if not sharekey: raise Exception( "You must specify a sharekey of the file you" "want to 'like'.") endpoint = '/api/sharedfile/{0}/comments'.format(sharekey) data = self._make_request("GET", endpoint=endpoint) return [Comment.NewFromJSON(c) for c in data['comments']] def post_shared_file(self, image_file=None, source_link=None, shake_id=None, title=None, description=None): """ Upload an image. TODO: Don't have a pro account to test (or even write) code to upload a shared filed to a particular shake. Args: image_file (str): path to an image (jpg/gif) on your computer. source_link (str): URL of a source (youtube/vine/etc.) shake_id (int): shake to which to upload the file or source_link [optional] title (str): title of the SharedFile [optional] description (str): description of the SharedFile Returns: SharedFile key. """ if image_file and source_link: raise Exception('You can only specify an image file or ' 'a source link, not both.') if not image_file and not source_link: raise Exception('You must specify an image file or a source link') content_type = self._get_image_type(image_file) if not title: title = os.path.basename(image_file) f = open(image_file, 'rb') endpoint = '/api/upload' files = {'file': (title, f, content_type)} data = self._make_request('POST', endpoint=endpoint, files=files) f.close() return data def update_shared_file(self, sharekey=None, title=None, description=None): """ Update the editable details (just the title and description) of a SharedFile. Args: sharekey (str): Sharekey of the SharedFile to update. title (Optional[str]): Title of the SharedFile. description (Optional[str]): Description of the SharedFile Returns: SharedFile on success, 404 on Sharekey not found, 403 on unauthorized. """ if not sharekey: raise Exception( "You must specify a sharekey for the sharedfile" "you wish to update.") if not (title or description): raise Exception("You must specify a title or description.") post_data = {} if title: post_data['title'] = title if description: post_data['description'] = description endpoint = '/api/sharedfile/{0}'.format(sharekey) data = self._make_request('POST', endpoint=endpoint, data=post_data) return SharedFile.NewFromJSON(data)
jeremylow/pyshk
pyshk/api.py
Api.post_shared_file
python
def post_shared_file(self, image_file=None, source_link=None, shake_id=None, title=None, description=None): if image_file and source_link: raise Exception('You can only specify an image file or ' 'a source link, not both.') if not image_file and not source_link: raise Exception('You must specify an image file or a source link') content_type = self._get_image_type(image_file) if not title: title = os.path.basename(image_file) f = open(image_file, 'rb') endpoint = '/api/upload' files = {'file': (title, f, content_type)} data = self._make_request('POST', endpoint=endpoint, files=files) f.close() return data
Upload an image. TODO: Don't have a pro account to test (or even write) code to upload a shared filed to a particular shake. Args: image_file (str): path to an image (jpg/gif) on your computer. source_link (str): URL of a source (youtube/vine/etc.) shake_id (int): shake to which to upload the file or source_link [optional] title (str): title of the SharedFile [optional] description (str): description of the SharedFile Returns: SharedFile key.
train
https://github.com/jeremylow/pyshk/blob/3ab92f6706397cde7a18367266eba9e0f1ada868/pyshk/api.py#L516-L557
[ "def _make_request(self, verb, endpoint=None, data=None, files=None):\n if not self.authenticated:\n raise ApiInstanceUnauthorized\n\n resource_url = self._get_url_endpoint(endpoint)\n\n timestamp = int(time.mktime(datetime.datetime.utcnow().timetuple()))\n nonce = self.get_nonce()\n\n authorization_header = self._make_headers(\n verb=verb,\n endpoint=endpoint,\n nonce=nonce,\n timestamp=timestamp)\n\n if verb == \"GET\":\n req = requests.get(\n resource_url,\n headers={'Authorization': authorization_header},\n verify=False)\n elif verb == \"POST\":\n if data:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n data=data)\n elif files:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n files=files)\n elif data and files:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n files=files,\n data=data)\n else:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header})\n\n if req.status_code == 401:\n raise ApiResponseUnauthorized(req)\n elif req.status_code == 404:\n raise NotFound404(req)\n elif req.status_code == 500:\n raise Exception(req)\n\n if self.testing:\n return req\n\n try:\n return req.json()\n except:\n print('returning req', req._content)\n return req\n", "def _get_image_type(image):\n if imghdr.what(image) == 'jpeg':\n return 'image/jpeg'\n elif imghdr.what(image) == 'gif':\n return 'image/gif'\n elif imghdr.what(image) == 'png':\n return 'image/png'\n" ]
class Api(object): def __init__(self, consumer_key=None, consumer_secret=None, access_token_key=None, access_token_secret=None, base_url=None, testing=False): if base_url is None: self.base_url = 'http://mlkshk.com' else: self.base_url = base_url self.port = 80 self.authenticated = False self.testing = False if testing: self.testing = True self.consumer_key = consumer_key self.consumer_secret = consumer_secret self.access_token_key = access_token_key self.access_token_secret = access_token_secret auth_list = [consumer_key, consumer_secret, access_token_key, access_token_secret] if all(auth_list): self.authenticated = True # Set headers, client info for requests. # default_headers = {'User-Agent': 'PyShk v0.0.1'} # self.client_args = {} # self.client_args['headers'] = default_headers # Set up auth - TODO: # self.auth = None # if self.access_token_key: # token = { # 'token_type': 'mac', # 'hash_algorithm': 'hmac-sha-1', # 'access_token': self.access_token_key # } # self.auth = OAuth2(self.consumer_key, token=token) # self.client = requests.Session() # self.client.auth = self.auth def get_auth(self, redirect_uri=None): if not redirect_uri: redirect_uri = "http://localhost:8000" authentication_url = ( "https://mlkshk.com/api/authorize" "?response_type=code&client_id={key}&redirect_uri={uri}").format( key=self.consumer_key, uri=redirect_uri) access_token_url = 'https://mlkshk.com/api/token' if not self.testing: webbrowser.open(authentication_url, new=1) authorization_code = input("Enter the code from the redirected URL: ") else: authorization_code = 123456 message = { 'grant_type': "authorization_code", 'code': authorization_code, 'redirect_uri': redirect_uri, 'client_id': self.consumer_key, 'client_secret': self.consumer_secret} data = urlencode(message) req = requests.post(access_token_url, params=data, verify=False) json_resp = req.json() print(""" {full_token} >>> Your access token is: {token} >>> Your access secret is: {secret} """.format(full_token=json_resp, token=json_resp['access_token'], secret=json_resp['secret'])) self.access_token_key = json_resp['access_token'] self.access_token_secret = json_resp['secret'] def _get_url_endpoint(self, endpoint): return self.base_url + endpoint def _make_headers(self, verb=None, endpoint=None, nonce=None, timestamp=None): normalized_string = "{0}\n".format(self.access_token_key) normalized_string += "{0}\n".format(timestamp) normalized_string += "{0}\n".format(nonce) normalized_string += "{0}\n".format(verb) normalized_string += "mlkshk.com\n" normalized_string += "80\n" normalized_string += "{0}\n".format(endpoint) digest = hmac.new( self.access_token_secret.encode('ascii'), normalized_string.encode('ascii'), sha1).digest() if six.PY2: signature = base64.encodestring(digest).strip().decode('utf8') else: signature = base64.encodebytes(digest).strip().decode('utf8') auth_str = ( 'MAC token="{0}", ' 'timestamp="{1}", ' 'nonce="{2}", ' 'signature="{3}"').format( self.access_token_key, str(timestamp), nonce, signature) return auth_str def _make_request(self, verb, endpoint=None, data=None, files=None): if not self.authenticated: raise ApiInstanceUnauthorized resource_url = self._get_url_endpoint(endpoint) timestamp = int(time.mktime(datetime.datetime.utcnow().timetuple())) nonce = self.get_nonce() authorization_header = self._make_headers( verb=verb, endpoint=endpoint, nonce=nonce, timestamp=timestamp) if verb == "GET": req = requests.get( resource_url, headers={'Authorization': authorization_header}, verify=False) elif verb == "POST": if data: req = requests.post( resource_url, headers={'Authorization': authorization_header}, data=data) elif files: req = requests.post( resource_url, headers={'Authorization': authorization_header}, files=files) elif data and files: req = requests.post( resource_url, headers={'Authorization': authorization_header}, files=files, data=data) else: req = requests.post( resource_url, headers={'Authorization': authorization_header}) if req.status_code == 401: raise ApiResponseUnauthorized(req) elif req.status_code == 404: raise NotFound404(req) elif req.status_code == 500: raise Exception(req) if self.testing: return req try: return req.json() except: print('returning req', req._content) return req @staticmethod def get_nonce(): nonce = md5( str(random.SystemRandom().randint(0, 100000000)).encode('utf8') ).hexdigest() return nonce @staticmethod def _get_image_type(image): if imghdr.what(image) == 'jpeg': return 'image/jpeg' elif imghdr.what(image) == 'gif': return 'image/gif' elif imghdr.what(image) == 'png': return 'image/png' def get_favorites(self, before=None, after=None): """ Get a list of the authenticated user's 10 most recent favorites (likes). Args: before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the authenticated user's set of Likes. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the authenticated user's set of Likes. Returns: List of SharedFile objects. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/favorites' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['favorites']] def get_user(self, user_id=None, user_name=None): """ Get a user object from the API. If no ``user_id`` or ``user_name`` is specified, it will return the User object for the currently authenticated user. Args: user_id (int): User ID of the user for whom you want to get information. [Optional] user_name(str): Username for the user for whom you want to get information. [Optional] Returns: A User object. """ if user_id: endpoint = '/api/user_id/{0}'.format(user_id) elif user_name: endpoint = '/api/user_name/{0}'.format(user_name) else: # Return currently authorized user endpoint = '/api/user' data = self._make_request(verb="GET", endpoint=endpoint) try: return User.NewFromJSON(data) except: return data def get_user_shakes(self): """ Get a list of Shake objects for the currently authenticated user. Returns: A list of Shake objects. """ endpoint = '/api/shakes' data = self._make_request(verb="GET", endpoint=endpoint) shakes = [Shake.NewFromJSON(shk) for shk in data['shakes']] return shakes def get_shared_files_from_shake(self, shake_id=None, before=None, after=None): """ Returns a list of SharedFile objects from a particular shake. Args: shake_id (int): Shake from which to get a list of SharedFiles before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the given Shake. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the given Shake. Returns: List (list) of SharedFiles. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/shakes' if shake_id: endpoint += '/{0}'.format(shake_id) if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request(verb="GET", endpoint=endpoint) return [SharedFile.NewFromJSON(f) for f in data['sharedfiles']] def get_shared_file(self, sharekey=None): """ Returns a SharedFile object given by the sharekey. Args: sharekey (str): Sharekey of the SharedFile you want to retrieve. Returns: SharedFile """ if not sharekey: raise Exception("You must specify a sharekey.") endpoint = '/api/sharedfile/{0}'.format(sharekey) data = self._make_request('GET', endpoint) return SharedFile.NewFromJSON(data) def like_shared_file(self, sharekey=None): """ 'Like' a SharedFile. mlkshk doesn't allow you to unlike a sharedfile, so this is ~~permanent~~. Args: sharekey (str): Sharekey for the file you want to 'like'. Returns: Either a SharedFile on success, or an exception on error. """ if not sharekey: raise Exception( "You must specify a sharekey of the file you" "want to 'like'.") endpoint = '/api/sharedfile/{sharekey}/like'.format(sharekey=sharekey) data = self._make_request("POST", endpoint=endpoint, data=None) try: sf = SharedFile.NewFromJSON(data) sf.liked = True return sf except: raise Exception("{0}".format(data['error'])) def save_shared_file(self, sharekey=None): """ Save a SharedFile to your Shake. Args: sharekey (str): Sharekey for the file to save. Returns: SharedFile saved to your shake. """ endpoint = '/api/sharedfile/{sharekey}/save'.format(sharekey=sharekey) data = self._make_request("POST", endpoint=endpoint, data=None) try: sf = SharedFile.NewFromJSON(data) sf.saved = True return sf except: raise Exception("{0}".format(data['error'])) def get_friends_shake(self, before=None, after=None): """ Contrary to the endpoint naming, this resource is for a list of SharedFiles from your friends on mlkshk. Returns: List of SharedFiles. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/friends' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['friend_shake']] def get_incoming_shake(self, before=None, after=None): """ Returns a list of the most recent SharedFiles on mlkshk.com Args: before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the Incoming Shake. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the Incoming Shake. Returns: List of SharedFile objects. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/incoming' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['incoming']] def get_magic_shake(self, before=None, after=None): """ From the API: Returns the 10 most recent files accepted by the 'magic' file selection algorithm. Currently any files with 10 or more likes are magic. Returns: List of SharedFile objects """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/magicfiles' if before: endpoint += '/before/{key}'.format(key=before) elif after: endpoint += '/after/{key}'.format(key=after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['magicfiles']] def get_comments(self, sharekey=None): """ Retrieve comments on a SharedFile Args: sharekey (str): Sharekey for the file from which you want to return the set of comments. Returns: List of Comment objects. """ if not sharekey: raise Exception( "You must specify a sharekey of the file you" "want to 'like'.") endpoint = '/api/sharedfile/{0}/comments'.format(sharekey) data = self._make_request("GET", endpoint=endpoint) return [Comment.NewFromJSON(c) for c in data['comments']] def post_comment(self, sharekey=None, comment=None): """ Post a comment on behalf of the current user to the SharedFile with the given sharekey. Args: sharekey (str): Sharekey of the SharedFile to which you'd like to post a comment. comment (str): Text of the comment to post. Returns: Comment object. """ endpoint = '/api/sharedfile/{0}/comments'.format(sharekey) post_data = {'body': comment} data = self._make_request("POST", endpoint=endpoint, data=post_data) return Comment.NewFromJSON(data) def update_shared_file(self, sharekey=None, title=None, description=None): """ Update the editable details (just the title and description) of a SharedFile. Args: sharekey (str): Sharekey of the SharedFile to update. title (Optional[str]): Title of the SharedFile. description (Optional[str]): Description of the SharedFile Returns: SharedFile on success, 404 on Sharekey not found, 403 on unauthorized. """ if not sharekey: raise Exception( "You must specify a sharekey for the sharedfile" "you wish to update.") if not (title or description): raise Exception("You must specify a title or description.") post_data = {} if title: post_data['title'] = title if description: post_data['description'] = description endpoint = '/api/sharedfile/{0}'.format(sharekey) data = self._make_request('POST', endpoint=endpoint, data=post_data) return SharedFile.NewFromJSON(data)
jeremylow/pyshk
pyshk/api.py
Api.update_shared_file
python
def update_shared_file(self, sharekey=None, title=None, description=None): if not sharekey: raise Exception( "You must specify a sharekey for the sharedfile" "you wish to update.") if not (title or description): raise Exception("You must specify a title or description.") post_data = {} if title: post_data['title'] = title if description: post_data['description'] = description endpoint = '/api/sharedfile/{0}'.format(sharekey) data = self._make_request('POST', endpoint=endpoint, data=post_data) return SharedFile.NewFromJSON(data)
Update the editable details (just the title and description) of a SharedFile. Args: sharekey (str): Sharekey of the SharedFile to update. title (Optional[str]): Title of the SharedFile. description (Optional[str]): Description of the SharedFile Returns: SharedFile on success, 404 on Sharekey not found, 403 on unauthorized.
train
https://github.com/jeremylow/pyshk/blob/3ab92f6706397cde7a18367266eba9e0f1ada868/pyshk/api.py#L559-L595
[ "def _make_request(self, verb, endpoint=None, data=None, files=None):\n if not self.authenticated:\n raise ApiInstanceUnauthorized\n\n resource_url = self._get_url_endpoint(endpoint)\n\n timestamp = int(time.mktime(datetime.datetime.utcnow().timetuple()))\n nonce = self.get_nonce()\n\n authorization_header = self._make_headers(\n verb=verb,\n endpoint=endpoint,\n nonce=nonce,\n timestamp=timestamp)\n\n if verb == \"GET\":\n req = requests.get(\n resource_url,\n headers={'Authorization': authorization_header},\n verify=False)\n elif verb == \"POST\":\n if data:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n data=data)\n elif files:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n files=files)\n elif data and files:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header},\n files=files,\n data=data)\n else:\n req = requests.post(\n resource_url,\n headers={'Authorization': authorization_header})\n\n if req.status_code == 401:\n raise ApiResponseUnauthorized(req)\n elif req.status_code == 404:\n raise NotFound404(req)\n elif req.status_code == 500:\n raise Exception(req)\n\n if self.testing:\n return req\n\n try:\n return req.json()\n except:\n print('returning req', req._content)\n return req\n", "def NewFromJSON(data):\n \"\"\"\n Create a new SharedFile instance from a JSON dict.\n\n Args:\n data (dict): JSON dictionary representing a SharedFile.\n\n Returns:\n A SharedFile instance.\n \"\"\"\n return SharedFile(\n sharekey=data.get('sharekey', None),\n name=data.get('name', None),\n user=User.NewFromJSON(data.get('user', None)),\n title=data.get('title', None),\n description=data.get('description', None),\n posted_at=data.get('posted_at', None),\n permalink=data.get('permalink', None),\n width=data.get('width', None),\n height=data.get('height', None),\n views=data.get('views', 0),\n likes=data.get('likes', 0),\n saves=data.get('saves', 0),\n comments=data.get('comments', None),\n nsfw=data.get('nsfw', False),\n image_url=data.get('image_url', None),\n source_url=data.get('source_url', None),\n saved=data.get('saved', False),\n liked=data.get('liked', False),\n )\n" ]
class Api(object): def __init__(self, consumer_key=None, consumer_secret=None, access_token_key=None, access_token_secret=None, base_url=None, testing=False): if base_url is None: self.base_url = 'http://mlkshk.com' else: self.base_url = base_url self.port = 80 self.authenticated = False self.testing = False if testing: self.testing = True self.consumer_key = consumer_key self.consumer_secret = consumer_secret self.access_token_key = access_token_key self.access_token_secret = access_token_secret auth_list = [consumer_key, consumer_secret, access_token_key, access_token_secret] if all(auth_list): self.authenticated = True # Set headers, client info for requests. # default_headers = {'User-Agent': 'PyShk v0.0.1'} # self.client_args = {} # self.client_args['headers'] = default_headers # Set up auth - TODO: # self.auth = None # if self.access_token_key: # token = { # 'token_type': 'mac', # 'hash_algorithm': 'hmac-sha-1', # 'access_token': self.access_token_key # } # self.auth = OAuth2(self.consumer_key, token=token) # self.client = requests.Session() # self.client.auth = self.auth def get_auth(self, redirect_uri=None): if not redirect_uri: redirect_uri = "http://localhost:8000" authentication_url = ( "https://mlkshk.com/api/authorize" "?response_type=code&client_id={key}&redirect_uri={uri}").format( key=self.consumer_key, uri=redirect_uri) access_token_url = 'https://mlkshk.com/api/token' if not self.testing: webbrowser.open(authentication_url, new=1) authorization_code = input("Enter the code from the redirected URL: ") else: authorization_code = 123456 message = { 'grant_type': "authorization_code", 'code': authorization_code, 'redirect_uri': redirect_uri, 'client_id': self.consumer_key, 'client_secret': self.consumer_secret} data = urlencode(message) req = requests.post(access_token_url, params=data, verify=False) json_resp = req.json() print(""" {full_token} >>> Your access token is: {token} >>> Your access secret is: {secret} """.format(full_token=json_resp, token=json_resp['access_token'], secret=json_resp['secret'])) self.access_token_key = json_resp['access_token'] self.access_token_secret = json_resp['secret'] def _get_url_endpoint(self, endpoint): return self.base_url + endpoint def _make_headers(self, verb=None, endpoint=None, nonce=None, timestamp=None): normalized_string = "{0}\n".format(self.access_token_key) normalized_string += "{0}\n".format(timestamp) normalized_string += "{0}\n".format(nonce) normalized_string += "{0}\n".format(verb) normalized_string += "mlkshk.com\n" normalized_string += "80\n" normalized_string += "{0}\n".format(endpoint) digest = hmac.new( self.access_token_secret.encode('ascii'), normalized_string.encode('ascii'), sha1).digest() if six.PY2: signature = base64.encodestring(digest).strip().decode('utf8') else: signature = base64.encodebytes(digest).strip().decode('utf8') auth_str = ( 'MAC token="{0}", ' 'timestamp="{1}", ' 'nonce="{2}", ' 'signature="{3}"').format( self.access_token_key, str(timestamp), nonce, signature) return auth_str def _make_request(self, verb, endpoint=None, data=None, files=None): if not self.authenticated: raise ApiInstanceUnauthorized resource_url = self._get_url_endpoint(endpoint) timestamp = int(time.mktime(datetime.datetime.utcnow().timetuple())) nonce = self.get_nonce() authorization_header = self._make_headers( verb=verb, endpoint=endpoint, nonce=nonce, timestamp=timestamp) if verb == "GET": req = requests.get( resource_url, headers={'Authorization': authorization_header}, verify=False) elif verb == "POST": if data: req = requests.post( resource_url, headers={'Authorization': authorization_header}, data=data) elif files: req = requests.post( resource_url, headers={'Authorization': authorization_header}, files=files) elif data and files: req = requests.post( resource_url, headers={'Authorization': authorization_header}, files=files, data=data) else: req = requests.post( resource_url, headers={'Authorization': authorization_header}) if req.status_code == 401: raise ApiResponseUnauthorized(req) elif req.status_code == 404: raise NotFound404(req) elif req.status_code == 500: raise Exception(req) if self.testing: return req try: return req.json() except: print('returning req', req._content) return req @staticmethod def get_nonce(): nonce = md5( str(random.SystemRandom().randint(0, 100000000)).encode('utf8') ).hexdigest() return nonce @staticmethod def _get_image_type(image): if imghdr.what(image) == 'jpeg': return 'image/jpeg' elif imghdr.what(image) == 'gif': return 'image/gif' elif imghdr.what(image) == 'png': return 'image/png' def get_favorites(self, before=None, after=None): """ Get a list of the authenticated user's 10 most recent favorites (likes). Args: before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the authenticated user's set of Likes. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the authenticated user's set of Likes. Returns: List of SharedFile objects. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/favorites' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['favorites']] def get_user(self, user_id=None, user_name=None): """ Get a user object from the API. If no ``user_id`` or ``user_name`` is specified, it will return the User object for the currently authenticated user. Args: user_id (int): User ID of the user for whom you want to get information. [Optional] user_name(str): Username for the user for whom you want to get information. [Optional] Returns: A User object. """ if user_id: endpoint = '/api/user_id/{0}'.format(user_id) elif user_name: endpoint = '/api/user_name/{0}'.format(user_name) else: # Return currently authorized user endpoint = '/api/user' data = self._make_request(verb="GET", endpoint=endpoint) try: return User.NewFromJSON(data) except: return data def get_user_shakes(self): """ Get a list of Shake objects for the currently authenticated user. Returns: A list of Shake objects. """ endpoint = '/api/shakes' data = self._make_request(verb="GET", endpoint=endpoint) shakes = [Shake.NewFromJSON(shk) for shk in data['shakes']] return shakes def get_shared_files_from_shake(self, shake_id=None, before=None, after=None): """ Returns a list of SharedFile objects from a particular shake. Args: shake_id (int): Shake from which to get a list of SharedFiles before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the given Shake. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the given Shake. Returns: List (list) of SharedFiles. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/shakes' if shake_id: endpoint += '/{0}'.format(shake_id) if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request(verb="GET", endpoint=endpoint) return [SharedFile.NewFromJSON(f) for f in data['sharedfiles']] def get_shared_file(self, sharekey=None): """ Returns a SharedFile object given by the sharekey. Args: sharekey (str): Sharekey of the SharedFile you want to retrieve. Returns: SharedFile """ if not sharekey: raise Exception("You must specify a sharekey.") endpoint = '/api/sharedfile/{0}'.format(sharekey) data = self._make_request('GET', endpoint) return SharedFile.NewFromJSON(data) def like_shared_file(self, sharekey=None): """ 'Like' a SharedFile. mlkshk doesn't allow you to unlike a sharedfile, so this is ~~permanent~~. Args: sharekey (str): Sharekey for the file you want to 'like'. Returns: Either a SharedFile on success, or an exception on error. """ if not sharekey: raise Exception( "You must specify a sharekey of the file you" "want to 'like'.") endpoint = '/api/sharedfile/{sharekey}/like'.format(sharekey=sharekey) data = self._make_request("POST", endpoint=endpoint, data=None) try: sf = SharedFile.NewFromJSON(data) sf.liked = True return sf except: raise Exception("{0}".format(data['error'])) def save_shared_file(self, sharekey=None): """ Save a SharedFile to your Shake. Args: sharekey (str): Sharekey for the file to save. Returns: SharedFile saved to your shake. """ endpoint = '/api/sharedfile/{sharekey}/save'.format(sharekey=sharekey) data = self._make_request("POST", endpoint=endpoint, data=None) try: sf = SharedFile.NewFromJSON(data) sf.saved = True return sf except: raise Exception("{0}".format(data['error'])) def get_friends_shake(self, before=None, after=None): """ Contrary to the endpoint naming, this resource is for a list of SharedFiles from your friends on mlkshk. Returns: List of SharedFiles. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/friends' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['friend_shake']] def get_incoming_shake(self, before=None, after=None): """ Returns a list of the most recent SharedFiles on mlkshk.com Args: before (str): get 10 SharedFile objects before (but not including) the SharedFile given by `before` for the Incoming Shake. after (str): get 10 SharedFile objects after (but not including) the SharedFile give by `after' for the Incoming Shake. Returns: List of SharedFile objects. """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/incoming' if before: endpoint += '/before/{0}'.format(before) elif after: endpoint += '/after/{0}'.format(after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['incoming']] def get_magic_shake(self, before=None, after=None): """ From the API: Returns the 10 most recent files accepted by the 'magic' file selection algorithm. Currently any files with 10 or more likes are magic. Returns: List of SharedFile objects """ if before and after: raise Exception("You cannot specify both before and after keys") endpoint = '/api/magicfiles' if before: endpoint += '/before/{key}'.format(key=before) elif after: endpoint += '/after/{key}'.format(key=after) data = self._make_request("GET", endpoint=endpoint) return [SharedFile.NewFromJSON(sf) for sf in data['magicfiles']] def get_comments(self, sharekey=None): """ Retrieve comments on a SharedFile Args: sharekey (str): Sharekey for the file from which you want to return the set of comments. Returns: List of Comment objects. """ if not sharekey: raise Exception( "You must specify a sharekey of the file you" "want to 'like'.") endpoint = '/api/sharedfile/{0}/comments'.format(sharekey) data = self._make_request("GET", endpoint=endpoint) return [Comment.NewFromJSON(c) for c in data['comments']] def post_comment(self, sharekey=None, comment=None): """ Post a comment on behalf of the current user to the SharedFile with the given sharekey. Args: sharekey (str): Sharekey of the SharedFile to which you'd like to post a comment. comment (str): Text of the comment to post. Returns: Comment object. """ endpoint = '/api/sharedfile/{0}/comments'.format(sharekey) post_data = {'body': comment} data = self._make_request("POST", endpoint=endpoint, data=post_data) return Comment.NewFromJSON(data) def post_shared_file(self, image_file=None, source_link=None, shake_id=None, title=None, description=None): """ Upload an image. TODO: Don't have a pro account to test (or even write) code to upload a shared filed to a particular shake. Args: image_file (str): path to an image (jpg/gif) on your computer. source_link (str): URL of a source (youtube/vine/etc.) shake_id (int): shake to which to upload the file or source_link [optional] title (str): title of the SharedFile [optional] description (str): description of the SharedFile Returns: SharedFile key. """ if image_file and source_link: raise Exception('You can only specify an image file or ' 'a source link, not both.') if not image_file and not source_link: raise Exception('You must specify an image file or a source link') content_type = self._get_image_type(image_file) if not title: title = os.path.basename(image_file) f = open(image_file, 'rb') endpoint = '/api/upload' files = {'file': (title, f, content_type)} data = self._make_request('POST', endpoint=endpoint, files=files) f.close() return data
jeremylow/pyshk
pyshk/models.py
User.AsDict
python
def AsDict(self, dt=True): data = {} if self.name: data['name'] = self.name data['mlkshk_url'] = self.mlkshk_url if self.profile_image_url: data['profile_image_url'] = self.profile_image_url if self.id: data['id'] = self.id if self.about: data['about'] = self.about if self.website: data['website'] = self.website if self.shakes: data['shakes'] = [shk.AsDict(dt=dt) for shk in self.shakes] data['shake_count'] = self.shake_count return data
A dict representation of this User instance. The return value uses the same key names as the JSON representation. Args: dt (bool): If True, return dates as python datetime objects. If False, return dates as ISO strings. Return: A dict representing this User instance
train
https://github.com/jeremylow/pyshk/blob/3ab92f6706397cde7a18367266eba9e0f1ada868/pyshk/models.py#L51-L79
null
class User(object): """ A class representing a MLKSHK user. Exposes the following properties of a user: user.id user.name user.profile_image_url user.about user.website user.shakes """ def __init__(self, **kwargs): param_defaults = { 'id': None, 'name': None, 'profile_image_url': None, 'about': None, 'website': None, 'shakes': None} for (param, default) in param_defaults.items(): setattr(self, param, kwargs.get(param, default)) @property def mlkshk_url(self): return "https://mlkshk.com/user/{0}".format(self.name) @property def shake_count(self): if self.shakes: return len(self.shakes) else: return 0 def AsJsonString(self): """A JSON string representation of this User instance. Returns: A JSON string representation of this User instance """ return json.dumps(self.AsDict(dt=False), sort_keys=True) @staticmethod def NewFromJSON(data): """ Create a new User instance from a JSON dict. Args: data (dict): JSON dictionary representing a user. Returns: A User instance. """ if data.get('shakes', None): shakes = [Shake.NewFromJSON(shk) for shk in data.get('shakes')] else: shakes = None return User( id=data.get('id', None), name=data.get('name', None), profile_image_url=data.get('profile_image_url', None), about=data.get('about', None), website=data.get('website', None), shakes=shakes) def __eq__(self, other): """ Compare two user objects against one another. Args: other (User): another User object against which to compare the current user. """ try: return other and \ self.id == other.id and \ self.name == other.name and \ self.profile_image_url == other.profile_image_url and \ self.about == other.about and \ self.website == other.website and \ self.shakes == other.shakes except AttributeError: return False def __ne__(self, other): return not self.__eq__(other) def __repr__(self): """ String representation of this User instance. """ return self.AsJsonString()
jeremylow/pyshk
pyshk/models.py
User.AsJsonString
python
def AsJsonString(self): return json.dumps(self.AsDict(dt=False), sort_keys=True)
A JSON string representation of this User instance. Returns: A JSON string representation of this User instance
train
https://github.com/jeremylow/pyshk/blob/3ab92f6706397cde7a18367266eba9e0f1ada868/pyshk/models.py#L81-L87
[ "def AsDict(self, dt=True):\n \"\"\"\n A dict representation of this User instance.\n\n The return value uses the same key names as the JSON representation.\n\n Args:\n dt (bool): If True, return dates as python datetime objects. If\n False, return dates as ISO strings.\n\n Return:\n A dict representing this User instance\n \"\"\"\n data = {}\n if self.name:\n data['name'] = self.name\n data['mlkshk_url'] = self.mlkshk_url\n if self.profile_image_url:\n data['profile_image_url'] = self.profile_image_url\n if self.id:\n data['id'] = self.id\n if self.about:\n data['about'] = self.about\n if self.website:\n data['website'] = self.website\n if self.shakes:\n data['shakes'] = [shk.AsDict(dt=dt) for shk in self.shakes]\n data['shake_count'] = self.shake_count\n return data\n" ]
class User(object): """ A class representing a MLKSHK user. Exposes the following properties of a user: user.id user.name user.profile_image_url user.about user.website user.shakes """ def __init__(self, **kwargs): param_defaults = { 'id': None, 'name': None, 'profile_image_url': None, 'about': None, 'website': None, 'shakes': None} for (param, default) in param_defaults.items(): setattr(self, param, kwargs.get(param, default)) @property def mlkshk_url(self): return "https://mlkshk.com/user/{0}".format(self.name) @property def shake_count(self): if self.shakes: return len(self.shakes) else: return 0 def AsDict(self, dt=True): """ A dict representation of this User instance. The return value uses the same key names as the JSON representation. Args: dt (bool): If True, return dates as python datetime objects. If False, return dates as ISO strings. Return: A dict representing this User instance """ data = {} if self.name: data['name'] = self.name data['mlkshk_url'] = self.mlkshk_url if self.profile_image_url: data['profile_image_url'] = self.profile_image_url if self.id: data['id'] = self.id if self.about: data['about'] = self.about if self.website: data['website'] = self.website if self.shakes: data['shakes'] = [shk.AsDict(dt=dt) for shk in self.shakes] data['shake_count'] = self.shake_count return data @staticmethod def NewFromJSON(data): """ Create a new User instance from a JSON dict. Args: data (dict): JSON dictionary representing a user. Returns: A User instance. """ if data.get('shakes', None): shakes = [Shake.NewFromJSON(shk) for shk in data.get('shakes')] else: shakes = None return User( id=data.get('id', None), name=data.get('name', None), profile_image_url=data.get('profile_image_url', None), about=data.get('about', None), website=data.get('website', None), shakes=shakes) def __eq__(self, other): """ Compare two user objects against one another. Args: other (User): another User object against which to compare the current user. """ try: return other and \ self.id == other.id and \ self.name == other.name and \ self.profile_image_url == other.profile_image_url and \ self.about == other.about and \ self.website == other.website and \ self.shakes == other.shakes except AttributeError: return False def __ne__(self, other): return not self.__eq__(other) def __repr__(self): """ String representation of this User instance. """ return self.AsJsonString()
jeremylow/pyshk
pyshk/models.py
User.NewFromJSON
python
def NewFromJSON(data): if data.get('shakes', None): shakes = [Shake.NewFromJSON(shk) for shk in data.get('shakes')] else: shakes = None return User( id=data.get('id', None), name=data.get('name', None), profile_image_url=data.get('profile_image_url', None), about=data.get('about', None), website=data.get('website', None), shakes=shakes)
Create a new User instance from a JSON dict. Args: data (dict): JSON dictionary representing a user. Returns: A User instance.
train
https://github.com/jeremylow/pyshk/blob/3ab92f6706397cde7a18367266eba9e0f1ada868/pyshk/models.py#L90-L111
null
class User(object): """ A class representing a MLKSHK user. Exposes the following properties of a user: user.id user.name user.profile_image_url user.about user.website user.shakes """ def __init__(self, **kwargs): param_defaults = { 'id': None, 'name': None, 'profile_image_url': None, 'about': None, 'website': None, 'shakes': None} for (param, default) in param_defaults.items(): setattr(self, param, kwargs.get(param, default)) @property def mlkshk_url(self): return "https://mlkshk.com/user/{0}".format(self.name) @property def shake_count(self): if self.shakes: return len(self.shakes) else: return 0 def AsDict(self, dt=True): """ A dict representation of this User instance. The return value uses the same key names as the JSON representation. Args: dt (bool): If True, return dates as python datetime objects. If False, return dates as ISO strings. Return: A dict representing this User instance """ data = {} if self.name: data['name'] = self.name data['mlkshk_url'] = self.mlkshk_url if self.profile_image_url: data['profile_image_url'] = self.profile_image_url if self.id: data['id'] = self.id if self.about: data['about'] = self.about if self.website: data['website'] = self.website if self.shakes: data['shakes'] = [shk.AsDict(dt=dt) for shk in self.shakes] data['shake_count'] = self.shake_count return data def AsJsonString(self): """A JSON string representation of this User instance. Returns: A JSON string representation of this User instance """ return json.dumps(self.AsDict(dt=False), sort_keys=True) @staticmethod def __eq__(self, other): """ Compare two user objects against one another. Args: other (User): another User object against which to compare the current user. """ try: return other and \ self.id == other.id and \ self.name == other.name and \ self.profile_image_url == other.profile_image_url and \ self.about == other.about and \ self.website == other.website and \ self.shakes == other.shakes except AttributeError: return False def __ne__(self, other): return not self.__eq__(other) def __repr__(self): """ String representation of this User instance. """ return self.AsJsonString()
jeremylow/pyshk
pyshk/models.py
Comment.AsDict
python
def AsDict(self, dt=True): data = {} if self.body: data['body'] = self.body if self.posted_at: data['posted_at'] = self.posted_at if self.user: data['user'] = self.user.AsDict() return data
A dict representation of this Comment instance. The return value uses the same key names as the JSON representation. Args: dt (bool): If True, return dates as python datetime objects. If False, return dates as ISO strings. Return: A dict representing this Comment instance
train
https://github.com/jeremylow/pyshk/blob/3ab92f6706397cde7a18367266eba9e0f1ada868/pyshk/models.py#L159-L181
null
class Comment(object): """ A class representing a Comment on mlkshk. Exposes the following properties of a Comment: comment.body comment.posted_at comment.user """ def __init__(self, **kwargs): param_defaults = { 'body': None, 'posted_at': None, 'user': None} for (param, default) in param_defaults.items(): setattr(self, param, kwargs.get(param, default)) def AsJsonString(self): """ A JSON string representation of this Comment instance. Returns: A JSON string representation of this Comment instance """ return json.dumps(self.AsDict(dt=False), sort_keys=True) @staticmethod def NewFromJSON(data): """ Create a new Comment instance from a JSON dict. Args: data (dict): JSON dictionary representing a Comment. Returns: A Comment instance. """ return Comment( body=data.get('body', None), posted_at=data.get('posted_at', None), user=User.NewFromJSON(data.get('user', None)) ) def __eq__(self, other): try: return other and \ self.body == other.body and \ self.posted_at == other.posted_at and \ self.user == other.user except AttributeError: return False def __ne__(self, other): return not self.__eq__(other) def __repr__(self): """ String representation of this Comment instance. """ return self.AsJsonString()
jeremylow/pyshk
pyshk/models.py
Comment.NewFromJSON
python
def NewFromJSON(data): return Comment( body=data.get('body', None), posted_at=data.get('posted_at', None), user=User.NewFromJSON(data.get('user', None)) )
Create a new Comment instance from a JSON dict. Args: data (dict): JSON dictionary representing a Comment. Returns: A Comment instance.
train
https://github.com/jeremylow/pyshk/blob/3ab92f6706397cde7a18367266eba9e0f1ada868/pyshk/models.py#L193-L207
[ "def NewFromJSON(data):\n \"\"\"\n Create a new User instance from a JSON dict.\n\n Args:\n data (dict): JSON dictionary representing a user.\n\n Returns:\n A User instance.\n \"\"\"\n if data.get('shakes', None):\n shakes = [Shake.NewFromJSON(shk) for shk in data.get('shakes')]\n else:\n shakes = None\n\n return User(\n id=data.get('id', None),\n name=data.get('name', None),\n profile_image_url=data.get('profile_image_url', None),\n about=data.get('about', None),\n website=data.get('website', None),\n shakes=shakes)\n" ]
class Comment(object): """ A class representing a Comment on mlkshk. Exposes the following properties of a Comment: comment.body comment.posted_at comment.user """ def __init__(self, **kwargs): param_defaults = { 'body': None, 'posted_at': None, 'user': None} for (param, default) in param_defaults.items(): setattr(self, param, kwargs.get(param, default)) def AsDict(self, dt=True): """ A dict representation of this Comment instance. The return value uses the same key names as the JSON representation. Args: dt (bool): If True, return dates as python datetime objects. If False, return dates as ISO strings. Return: A dict representing this Comment instance """ data = {} if self.body: data['body'] = self.body if self.posted_at: data['posted_at'] = self.posted_at if self.user: data['user'] = self.user.AsDict() return data def AsJsonString(self): """ A JSON string representation of this Comment instance. Returns: A JSON string representation of this Comment instance """ return json.dumps(self.AsDict(dt=False), sort_keys=True) @staticmethod def __eq__(self, other): try: return other and \ self.body == other.body and \ self.posted_at == other.posted_at and \ self.user == other.user except AttributeError: return False def __ne__(self, other): return not self.__eq__(other) def __repr__(self): """ String representation of this Comment instance. """ return self.AsJsonString()
jeremylow/pyshk
pyshk/models.py
Shake.AsDict
python
def AsDict(self, dt=True): data = {} if self.id: data['id'] = self.id if self.name: data['name'] = self.name if self.owner: data['owner'] = self.owner.AsDict() if self.url: data['url'] = self.url if self.thumbnail_url: data['thumbnail_url'] = self.thumbnail_url if self.description: data['description'] = self.description if self.type: data['type'] = self.type if dt: if self.created_at: data['created_at'] = self.created_at if self.updated_at: data['updated_at'] = self.updated_at else: if self.created_at: data['created_at'] = self.created_at_iso if self.updated_at: data['updated_at'] = self.updated_at_iso return data
A dict representation of this Shake instance. The return value uses the same key names as the JSON representation. Return: A dict representing this Shake instance
train
https://github.com/jeremylow/pyshk/blob/3ab92f6706397cde7a18367266eba9e0f1ada868/pyshk/models.py#L283-L320
null
class Shake(object): """ A class representing a Shake on mlkshk. Exposes the following properties of a Shake: shake.id shake.name shake.owner shake.url shake.thumbnail_url shake.description shake.type shake.created_at shake.updated_at """ def __init__(self, **kwargs): param_defaults = { 'id': None, 'name': None, 'owner': None, 'url': None, 'thumbnail_url': None, 'description': None, 'type': None, 'created_at': None, 'updated_at': None} for (param, default) in param_defaults.items(): setattr(self, param, kwargs.get(param, default)) self.created_at = kwargs.get('created_at', None) self.updated_at = kwargs.get('updated_at', None) @property def created_at(self): return self._created_at @created_at.setter def created_at(self, value): self._created_at = convert_time(value) @property def created_at_iso(self): return self._created_at.isoformat() @property def updated_at(self): return self._updated_at @updated_at.setter def updated_at(self, value): self._updated_at = convert_time(value) @property def updated_at_iso(self): return self._updated_at.isoformat() def AsJsonString(self): """ A JSON string representation of this Shake instance. Returns: A JSON string representation of this Shake instance """ return json.dumps(self.AsDict(dt=False), sort_keys=True) @staticmethod def NewFromJSON(data): """ Create a new Shake instance from a JSON dict. Args: data (dict): JSON dictionary representing a Shake. Returns: A Shake instance. """ s = Shake( id=data.get('id', None), name=data.get('name', None), url=data.get('url', None), thumbnail_url=data.get('thumbnail_url', None), description=data.get('description', None), type=data.get('type', None), created_at=data.get('created_at', None), updated_at=data.get('updated_at', None) ) if data.get('owner', None): s.owner = User.NewFromJSON(data.get('owner', None)) return s def __eq__(self, other): try: return other and \ self.id == other.id and \ self.name == other.name and \ self.owner == other.owner and \ self.url == other.url and \ self.thumbnail_url == other.thumbnail_url and \ self.description == other.description and \ self.type == other.type and \ self.created_at == other.created_at and \ self.updated_at == other.updated_at except AttributeError: return False def __ne__(self, other): return not self.__eq__(other) def __repr__(self): """ String representation of this Shake instance. """ return self.AsJsonString()
jeremylow/pyshk
pyshk/models.py
Shake.NewFromJSON
python
def NewFromJSON(data): s = Shake( id=data.get('id', None), name=data.get('name', None), url=data.get('url', None), thumbnail_url=data.get('thumbnail_url', None), description=data.get('description', None), type=data.get('type', None), created_at=data.get('created_at', None), updated_at=data.get('updated_at', None) ) if data.get('owner', None): s.owner = User.NewFromJSON(data.get('owner', None)) return s
Create a new Shake instance from a JSON dict. Args: data (dict): JSON dictionary representing a Shake. Returns: A Shake instance.
train
https://github.com/jeremylow/pyshk/blob/3ab92f6706397cde7a18367266eba9e0f1ada868/pyshk/models.py#L332-L354
[ "def NewFromJSON(data):\n \"\"\"\n Create a new User instance from a JSON dict.\n\n Args:\n data (dict): JSON dictionary representing a user.\n\n Returns:\n A User instance.\n \"\"\"\n if data.get('shakes', None):\n shakes = [Shake.NewFromJSON(shk) for shk in data.get('shakes')]\n else:\n shakes = None\n\n return User(\n id=data.get('id', None),\n name=data.get('name', None),\n profile_image_url=data.get('profile_image_url', None),\n about=data.get('about', None),\n website=data.get('website', None),\n shakes=shakes)\n" ]
class Shake(object): """ A class representing a Shake on mlkshk. Exposes the following properties of a Shake: shake.id shake.name shake.owner shake.url shake.thumbnail_url shake.description shake.type shake.created_at shake.updated_at """ def __init__(self, **kwargs): param_defaults = { 'id': None, 'name': None, 'owner': None, 'url': None, 'thumbnail_url': None, 'description': None, 'type': None, 'created_at': None, 'updated_at': None} for (param, default) in param_defaults.items(): setattr(self, param, kwargs.get(param, default)) self.created_at = kwargs.get('created_at', None) self.updated_at = kwargs.get('updated_at', None) @property def created_at(self): return self._created_at @created_at.setter def created_at(self, value): self._created_at = convert_time(value) @property def created_at_iso(self): return self._created_at.isoformat() @property def updated_at(self): return self._updated_at @updated_at.setter def updated_at(self, value): self._updated_at = convert_time(value) @property def updated_at_iso(self): return self._updated_at.isoformat() def AsDict(self, dt=True): """ A dict representation of this Shake instance. The return value uses the same key names as the JSON representation. Return: A dict representing this Shake instance """ data = {} if self.id: data['id'] = self.id if self.name: data['name'] = self.name if self.owner: data['owner'] = self.owner.AsDict() if self.url: data['url'] = self.url if self.thumbnail_url: data['thumbnail_url'] = self.thumbnail_url if self.description: data['description'] = self.description if self.type: data['type'] = self.type if dt: if self.created_at: data['created_at'] = self.created_at if self.updated_at: data['updated_at'] = self.updated_at else: if self.created_at: data['created_at'] = self.created_at_iso if self.updated_at: data['updated_at'] = self.updated_at_iso return data def AsJsonString(self): """ A JSON string representation of this Shake instance. Returns: A JSON string representation of this Shake instance """ return json.dumps(self.AsDict(dt=False), sort_keys=True) @staticmethod def __eq__(self, other): try: return other and \ self.id == other.id and \ self.name == other.name and \ self.owner == other.owner and \ self.url == other.url and \ self.thumbnail_url == other.thumbnail_url and \ self.description == other.description and \ self.type == other.type and \ self.created_at == other.created_at and \ self.updated_at == other.updated_at except AttributeError: return False def __ne__(self, other): return not self.__eq__(other) def __repr__(self): """ String representation of this Shake instance. """ return self.AsJsonString()
jeremylow/pyshk
pyshk/models.py
SharedFile.AsDict
python
def AsDict(self, dt=True): data = {} if self.sharekey: data['sharekey'] = self.sharekey if self.name: data['name'] = self.name if self.user: data['user'] = self.user.AsDict() if self.title: data['title'] = self.title if self.description: data['description'] = self.description if self.posted_at: if dt: data['posted_at'] = self.posted_at else: data['posted_at'] = self.posted_at_iso if self.permalink: data['permalink'] = self.permalink if self.width: data['width'] = self.width if self.height: data['height'] = self.height if self.image_url: data['image_url'] = self.image_url if self.source_url: data['source_url'] = self.source_url data['views'] = self.views data['likes'] = self.likes data['saves'] = self.saves data['comments'] = self.comments data['nsfw'] = self.nsfw data['saved'] = self.saved data['liked'] = self.liked return data
A dict representation of this Shake instance. The return value uses the same key names as the JSON representation. Args: dt (bool): If True, return dates as python datetime objects. If False, return dates as ISO strings. Return: A dict representing this Shake instance
train
https://github.com/jeremylow/pyshk/blob/3ab92f6706397cde7a18367266eba9e0f1ada868/pyshk/models.py#L462-L510
null
class SharedFile(object): """ A class representing a file shared on MLKSHK. Exposes the following properties of a sharedfile: sharedfile.sharekey sharedfile.name sharedfile.user sharedfile.title sharedfile.description sharedfile.posted_at sharedfile.permalink sharedfile.width sharedfile.height sharedfile.views sharedfile.likes sharedfile.saves sharedfile.comments sharedfile.nsfw sharedfile.image_url sharedfile.source_url sharedfile.saved sharedfile.liked Args: sharedfile.sharekey sharedfile.name sharedfile.user sharedfile.title sharedfile.description sharedfile.posted_at sharedfile.permalink sharedfile.width sharedfile.height sharedfile.views sharedfile.likes sharedfile.saves sharedfile.comments sharedfile.nsfw sharedfile.image_url sharedfile.source_url sharedfile.saved sharedfile.liked """ def __init__(self, *args, **kwargs): param_defaults = { 'sharekey': None, 'name': None, 'user': None, 'title': None, 'description': None, 'posted_at': None, 'permalink': None, 'width': None, 'height': None, 'views': None, 'likes': None, 'saves': None, 'comments': None, 'nsfw': None, 'image_url': None, 'source_url': None, 'saved': None, 'liked': None, } for (param, default) in param_defaults.items(): setattr(self, param, kwargs.get(param, default)) self.posted_at = kwargs.get('posted_at', None) @property def posted_at(self): return self._posted_at @posted_at.setter def posted_at(self, value): self._posted_at = convert_time(value) @property def posted_at_iso(self): return self._posted_at.isoformat() def AsJsonString(self): """ A JSON string representation of this SharedFile instance. Returns: A JSON string representation of this SharedFile instance """ return json.dumps(self.AsDict(dt=False), sort_keys=True) @staticmethod def NewFromJSON(data): """ Create a new SharedFile instance from a JSON dict. Args: data (dict): JSON dictionary representing a SharedFile. Returns: A SharedFile instance. """ return SharedFile( sharekey=data.get('sharekey', None), name=data.get('name', None), user=User.NewFromJSON(data.get('user', None)), title=data.get('title', None), description=data.get('description', None), posted_at=data.get('posted_at', None), permalink=data.get('permalink', None), width=data.get('width', None), height=data.get('height', None), views=data.get('views', 0), likes=data.get('likes', 0), saves=data.get('saves', 0), comments=data.get('comments', None), nsfw=data.get('nsfw', False), image_url=data.get('image_url', None), source_url=data.get('source_url', None), saved=data.get('saved', False), liked=data.get('liked', False), ) def __eq__(self, other): """ Compare two SharedFiles on all attributes **except** saved status and liked status. """ try: return other and \ self.sharekey == other.sharekey and \ self.name == other.name and \ self.user == other.user and \ self.title == other.title and \ self.description == other.description and \ self.posted_at == other.posted_at and \ self.permalink == other.permalink and \ self.width == other.width and \ self.height == other.height and \ self.views == other.views and \ self.likes == other.likes and \ self.saves == other.saves and \ self.comments == other.comments and \ self.nsfw == other.nsfw and \ self.image_url == other.image_url and \ self.source_url == other.source_url except AttributeError: return False def __ne__(self, other): return not self.__eq__(other) def __repr__(self): return self.AsJsonString()
jeremylow/pyshk
pyshk/models.py
SharedFile.NewFromJSON
python
def NewFromJSON(data): return SharedFile( sharekey=data.get('sharekey', None), name=data.get('name', None), user=User.NewFromJSON(data.get('user', None)), title=data.get('title', None), description=data.get('description', None), posted_at=data.get('posted_at', None), permalink=data.get('permalink', None), width=data.get('width', None), height=data.get('height', None), views=data.get('views', 0), likes=data.get('likes', 0), saves=data.get('saves', 0), comments=data.get('comments', None), nsfw=data.get('nsfw', False), image_url=data.get('image_url', None), source_url=data.get('source_url', None), saved=data.get('saved', False), liked=data.get('liked', False), )
Create a new SharedFile instance from a JSON dict. Args: data (dict): JSON dictionary representing a SharedFile. Returns: A SharedFile instance.
train
https://github.com/jeremylow/pyshk/blob/3ab92f6706397cde7a18367266eba9e0f1ada868/pyshk/models.py#L522-L551
[ "def NewFromJSON(data):\n \"\"\"\n Create a new User instance from a JSON dict.\n\n Args:\n data (dict): JSON dictionary representing a user.\n\n Returns:\n A User instance.\n \"\"\"\n if data.get('shakes', None):\n shakes = [Shake.NewFromJSON(shk) for shk in data.get('shakes')]\n else:\n shakes = None\n\n return User(\n id=data.get('id', None),\n name=data.get('name', None),\n profile_image_url=data.get('profile_image_url', None),\n about=data.get('about', None),\n website=data.get('website', None),\n shakes=shakes)\n" ]
class SharedFile(object): """ A class representing a file shared on MLKSHK. Exposes the following properties of a sharedfile: sharedfile.sharekey sharedfile.name sharedfile.user sharedfile.title sharedfile.description sharedfile.posted_at sharedfile.permalink sharedfile.width sharedfile.height sharedfile.views sharedfile.likes sharedfile.saves sharedfile.comments sharedfile.nsfw sharedfile.image_url sharedfile.source_url sharedfile.saved sharedfile.liked Args: sharedfile.sharekey sharedfile.name sharedfile.user sharedfile.title sharedfile.description sharedfile.posted_at sharedfile.permalink sharedfile.width sharedfile.height sharedfile.views sharedfile.likes sharedfile.saves sharedfile.comments sharedfile.nsfw sharedfile.image_url sharedfile.source_url sharedfile.saved sharedfile.liked """ def __init__(self, *args, **kwargs): param_defaults = { 'sharekey': None, 'name': None, 'user': None, 'title': None, 'description': None, 'posted_at': None, 'permalink': None, 'width': None, 'height': None, 'views': None, 'likes': None, 'saves': None, 'comments': None, 'nsfw': None, 'image_url': None, 'source_url': None, 'saved': None, 'liked': None, } for (param, default) in param_defaults.items(): setattr(self, param, kwargs.get(param, default)) self.posted_at = kwargs.get('posted_at', None) @property def posted_at(self): return self._posted_at @posted_at.setter def posted_at(self, value): self._posted_at = convert_time(value) @property def posted_at_iso(self): return self._posted_at.isoformat() def AsDict(self, dt=True): """ A dict representation of this Shake instance. The return value uses the same key names as the JSON representation. Args: dt (bool): If True, return dates as python datetime objects. If False, return dates as ISO strings. Return: A dict representing this Shake instance """ data = {} if self.sharekey: data['sharekey'] = self.sharekey if self.name: data['name'] = self.name if self.user: data['user'] = self.user.AsDict() if self.title: data['title'] = self.title if self.description: data['description'] = self.description if self.posted_at: if dt: data['posted_at'] = self.posted_at else: data['posted_at'] = self.posted_at_iso if self.permalink: data['permalink'] = self.permalink if self.width: data['width'] = self.width if self.height: data['height'] = self.height if self.image_url: data['image_url'] = self.image_url if self.source_url: data['source_url'] = self.source_url data['views'] = self.views data['likes'] = self.likes data['saves'] = self.saves data['comments'] = self.comments data['nsfw'] = self.nsfw data['saved'] = self.saved data['liked'] = self.liked return data def AsJsonString(self): """ A JSON string representation of this SharedFile instance. Returns: A JSON string representation of this SharedFile instance """ return json.dumps(self.AsDict(dt=False), sort_keys=True) @staticmethod def __eq__(self, other): """ Compare two SharedFiles on all attributes **except** saved status and liked status. """ try: return other and \ self.sharekey == other.sharekey and \ self.name == other.name and \ self.user == other.user and \ self.title == other.title and \ self.description == other.description and \ self.posted_at == other.posted_at and \ self.permalink == other.permalink and \ self.width == other.width and \ self.height == other.height and \ self.views == other.views and \ self.likes == other.likes and \ self.saves == other.saves and \ self.comments == other.comments and \ self.nsfw == other.nsfw and \ self.image_url == other.image_url and \ self.source_url == other.source_url except AttributeError: return False def __ne__(self, other): return not self.__eq__(other) def __repr__(self): return self.AsJsonString()
Dispersive-Hydrodynamics-Lab/PACE
PACE/PACE.py
manage_file_analysis
python
def manage_file_analysis(args: argparse.Namespace, filename: str, data: object) -> None: key = DataStore.hashfile(filename) print('Analyzing {} --> {}'.format(filename, key)) if data.check_key(key): # if exists in database, prepopulate fit = LineFit(filename, data=data.get_data(key)) else: fit = LineFit(filename) if args.time: noise, curvature, rnge, domn = fit.analyze(time=args.time) newrow = [args.time, noise, curvature, rnge, domn, fit.accepts[args.time]] data.update1(key, newrow, len(fit.noises)) else: fit.analyze_full() newrows = np.array([range(len(fit.noises)), fit.noises, fit.curves, fit.ranges, fit.domains, fit.accepts]) data.update(key, newrows) data.save()
Take care of the analysis of a datafile
train
https://github.com/Dispersive-Hydrodynamics-Lab/PACE/blob/4ce27d5fc9b02cc2ce55f6fea7fc8d6015317e1f/PACE/PACE.py#L79-L99
null
#!/usr/bin/env python3.5 """ PACE TODO: * model training/testing * more models (technically) * multithreading """ import sys import os import argparse import hashlib import typing from enforce import runtime_validation as types from tqdm import tqdm import numpy as np import numpy.linalg as linalg import matplotlib import matplotlib.pyplot as plt from matplotlib.colors import ListedColormap import scipy.integrate as si import scipy.io as sco import sklearn as sk from sklearn import svm from sklearn import preprocessing from sklearn import neighbors DATASTORE = 'linefitdata.mat' HEADER = (' ____ _ ____ _____\n' '| _ \ / \ / ___| ____|\n' '| |_) / _ \| | | _|\n' '| __/ ___ \ |___| |___\n' '|_| /_/ \_\____|_____|\n\n' 'PACE: Parameterization & Analysis of Conduit Edges\n' 'William Farmer - 2015\n') def main(): args = get_args() data = DataStore(DATASTORE) data.load() # Establish directory for img outputs if not os.path.exists('./img'): os.makedirs('./img') if args.plot: for filename in args.files: print('Plotting ' + filename) plot_name = './img/' + filename + '.general_fit.png' fit = LineFit(filename) fit.plot_file(name=plot_name, time=args.time) if args.analyze: for filename in args.files: manage_file_analysis(args, filename, data) if args.plotdata: data.plot_traindata() if args.machinetest: learner = ML(algo=args.model) if args.printdata: data.printdata() if args.printdatashort: data.printshort() @types class DataStore(object): def __init__(self, name: str): """ Uses a .mat as datastore for compatibility. Eventually may want to switch to SQLite, or some database? Not sure if ever needed. This class provides that extensible API structure however. Datafile has the following structure: learning_data = {filehash:[[trial_index, noise, curvature, range, domain, accept, viscosity] ,...],...} Conveniently, you can use the domain field as a check as to whether or not the row has been touched. If domain=0 (for that row) then that means that it hasn't been updated. :param: name of datastore """ self.name = name self.data = {} def load(self) -> None: """ Load datafile """ try: self.data = sco.loadmat(self.name) except FileNotFoundError: pass def save(self) -> None: """ Save datafile to disk """ sco.savemat(self.name, self.data) def get_data(self, key: str) -> np.ndarray: """ Returns the specified data. Warning, ZERO ERROR HANDLING :param key: name of file :return: 2d data array """ return self.data[key] @types def get_keys(self) -> typing.List[str]: """ Return list of SHA512 hash keys that exist in datafile :return: list of keys """ keys = [] for key in self.data.keys(): if key not in ['__header__', '__version__', '__globals__']: keys.append(key) return keys @types def check_key(self, key: str) -> bool: """ Checks if key exists in datastore. True if yes, False if no. :param: SHA512 hash key :return: whether or key not exists in datastore """ keys = self.get_keys() return key in keys def get_traindata(self) -> np.ndarray: """ Pulls all available data and concatenates for model training :return: 2d array of points """ traindata = None for key, value in self.data.items(): if key not in ['__header__', '__version__', '__globals__']: if traindata is None: traindata = value[np.where(value[:, 4] != 0)] else: traindata = np.concatenate((traindata, value[np.where(value[:, 4] != 0)])) return traindata @types def plot_traindata(self, name: str='dataplot') -> None: """ Plots traindata.... choo choo... """ traindata = self.get_traindata() plt.figure(figsize=(16, 16)) plt.scatter(traindata[:, 1], traindata[:, 2], c=traindata[:, 5], marker='o', label='Datastore Points') plt.xlabel(r'$\log_{10}$ Noise') plt.ylabel(r'$\log_{10}$ Curvature') plt.legend(loc=2, fontsize='xx-large') plt.savefig('./img/{}.png'.format(name)) def printdata(self) -> None: """ Prints data to stdout """ np.set_printoptions(threshold=np.nan) print(self.data) np.set_printoptions(threshold=1000) def printshort(self) -> None: """ Print shortened version of data to stdout""" print(self.data) @types def update(self, key: str, data: np.ndarray) -> None: """ Update entry in datastore """ self.data[key] = data def update1(self, key: str, data: np.ndarray, size: int) -> None: """ Update one entry in specific record in datastore """ print(data) if key in self.get_keys(): self.data[key][data[0]] = data else: newdata = np.zeros((size, 6)) newdata[data[0]] = data self.data[key] = newdata @staticmethod @types def hashfile(name: str) -> str: """ Gets a hash of a file using block parsing http://stackoverflow.com/questions/3431825/generating-a-md5-checksum-of-a-file Using SHA512 for long-term support (hehehehe) """ hasher = hashlib.sha512() with open(name, 'rb') as openfile: for chunk in iter(lambda: openfile.read(4096), b''): hasher.update(chunk) return hasher.hexdigest() class LineFit(object): def __init__(self, filename: str, data: np.ndarray=None, function_number: int=16, spread_number: int=22): """ Main class for line fitting and parameter determination :param: filename :param: data for fitting :param: number of functions :param: gaussian spread number """ self.filename = filename (self.averagedata, self.times, self.accepts, self.ratio, self.viscosity) = self._loadedges() self.domain = np.arange(len(self.averagedata[:, 0])) self.function_number = function_number self.spread_number = spread_number if data is None: self.noises = np.zeros(len(self.times)) self.curves = np.zeros(len(self.times)) self.ranges = np.zeros(len(self.times)) self.domains = np.zeros(len(self.times)) else: self.noises = data[:, 1] self.curves = data[:, 2] self.ranges = data[:, 3] self.domains = data[:, 4] @types def _loadedges(self) -> typing.Tuple[np.ndarray, np.ndarray, np.ndarray, float, np.ndarray]: """ Attempts to intelligently load the .mat file and take average of left and right edges :return: left and right averages :return: times for each column :return: accept/reject for each column :return: pixel-inch ratio """ data = sco.loadmat(self.filename) datakeys = [k for k in data.keys() if ('right' in k) or ('left' in k) or ('edge' in k)] averagedata = ((data[datakeys[0]] + data[datakeys[1]]) / 2) try: times = (data['times'] - data['times'].min())[0] except KeyError: times = np.arange(len(data[datakeys[0]][0])) try: accept = data['accept'] except KeyError: accept = np.zeros(len(times)) try: ratio = data['ratio'] except KeyError: ratio = 1 try: viscosity = data['viscosity'] except KeyError: viscosity = np.ones(len(times)) return averagedata, times, accept, ratio, viscosity def plot_file(self, name: str=None, time: int=None) -> None: """ Plot specific time for provided datafile. If no time provided, will plot middle. :param: savefile name :param: time/data column """ if not time: time = int(len(self.times) / 2) if not name: name = './img/' + self.filename + '.png' yhat, residuals, residual_mean, noise = self._get_fit(time) plt.figure() plt.scatter(self.domain, self.averagedata[:, time], alpha=0.2) plt.plot(yhat) plt.savefig(name) @staticmethod @types def ddiff(arr: np.ndarray) -> np.ndarray: """ Helper Function: Divided Differences input: array """ return arr[:-1] - arr[1:] @types def _gaussian_function(self, datalength: int, values: np.ndarray, height: int, index: int) -> np.ndarray: """ i'th Regression Model Gaussian :param: len(x) :param: x values :param: height of gaussian :param: position of gaussian :return: gaussian bumps over domain """ return height * np.exp(-(1 / (self.spread_number * datalength)) * (values - ((datalength / self.function_number) * index)) ** 2) @types def _get_fit(self, time: int) -> typing.Tuple[np.ndarray, np.ndarray, float, float]: """ Fit regression model to data :param: time (column of data) :return: predicted points :return: residuals :return: mean residual :return: error """ rawdata = self.averagedata[:, time] domain = np.arange(len(rawdata)) datalength = len(domain) coefficients = np.zeros((datalength, self.function_number + 2)) coefficients[:, 0] = 1 coefficients[:, 1] = domain for i in range(self.function_number): coefficients[:, 2 + i] = self._gaussian_function(datalength, domain, 1, i) betas = linalg.inv(coefficients.transpose().dot(coefficients)).dot(coefficients.transpose().dot(rawdata)) predicted_values = coefficients.dot(betas) residuals = rawdata - predicted_values error = np.sqrt(residuals.transpose().dot(residuals) / (datalength - (self.function_number + 2))) return predicted_values, residuals, residuals.mean(), error @types def _get_noise(self, residuals: np.ndarray) -> float: """ Determine Noise of Residuals. :param: residuals :return: noise """ return np.mean(np.abs(residuals)) @types def analyze(self, time: int=None) -> typing.Tuple[float, float, int, int]: """ Determine noise, curvature, range, and domain of specified array. :param: pixel to inch ratio :param: time (column) to use. :return: curvature :return: noise :return: range :return: domain """ if not time: time = int(len(self.times) / 2) if self.domains[time] == 0: yhat, residuals, mean_residual, error = self._get_fit(time) yhat_p = self.ddiff(yhat) yhat_pp = self.ddiff(yhat_p) noise = self._get_noise(residuals) curvature = (1 / self.ratio) * (1 / len(yhat_pp)) * np.sqrt(si.simps(yhat_pp ** 2)) rng = (self.ratio * (np.max(self.averagedata[:, time]) - np.min(self.averagedata[:, time]))) dmn = self.ratio * len(self.averagedata[:, time]) self.noises[time] = np.log10(noise) self.curves[time] = np.log10(curvature) self.ranges[time] = np.log10(rng) self.domains[time] = np.log10(dmn) return self.noises[time], self.curves[time], self.ranges[time], self.domains[time] @types def analyze_full(self) -> typing.Tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray]: """ Determine noise, curvature, range, and domain of specified data. Like analyze, except examines the entire file. :param: float->pixel to inch ratio :return: array->curvatures :return: array->noises :return: array->ranges :return: array->domains """ if self.noises[0] == 0: timelength = len(self.times) for i in tqdm(range(timelength)): self.analyze(time=i) return self.noises, self.curves, self.ranges, self.domains class ML(object): def __init__(self, args: argparse.Namespace, algo: str='nn'): """ Machine Learning to determine usability of data.... """ self.algo = self.get_algo(args, algo) def get_algo(self, args: argparse.Namespace, algo: str) -> object: """ Returns machine learning algorithm based on arguments """ if algo == 'nn': return NearestNeighbor(args.nnk) def train(self) -> None: """ Trains specified algorithm """ traindata = self.get_data() self.algo.train(traindata) def get_data(self) -> np.ndarray: """ Gets data for training We use the domain column to determine what fields have been filled out If the domain is zero (i.e. not in error) than we should probably ignore it anyway """ traindata = data.get_traindata() return traindata def plot_fitspace(self, name: str, X: np.ndarray, y: np.ndarray, clf: object) -> None: """ Plot 2dplane of fitspace """ cmap_light = ListedColormap(['#FFAAAA', '#AAFFAA', '#AAAAFF']) cmap_bold = ListedColormap(['#FF0000', '#00FF00', '#0000FF']) # Plot the decision boundary. For that, we will assign a color to each # point in the mesh [x_min, m_max]x[y_min, y_max]. h = 0.01 # Mesh step size x_min, x_max = X[:, 0].min() - 1, X[:, 0].max() + 1 y_min, y_max = X[:, 1].min() - 1, X[:, 1].max() + 1 xx, yy = np.meshgrid(np.arange(x_min, x_max, h), np.arange(y_min, y_max, h)) Z = clf.predict(np.c_[xx.ravel(), yy.ravel()]) # Put the result into a color plot Z = Z.reshape(xx.shape) plt.figure() plt.pcolormesh(xx, yy, Z, cmap=cmap_light) # Plot also the training points plt.scatter(X[:, 0], X[:, 1], c=y, cmap=cmap_bold) plt.xlim(xx.min(), xx.max()) plt.ylim(yy.min(), yy.max()) plt.xlabel(r'$\log_{10}$ Noise') plt.ylabel(r'$\log_{10}$ Curvature') plt.savefig(name) class NearestNeighbor(object): def __init__(self, k: int): """ An example machine learning model. EVERY MODEL NEEDS TO PROVIDE: 1. Train 2. Predict """ self.clf = neighbors.KNeighborsClassifier(k, weights='distance', p=2, algorithm='auto', n_jobs=8) def train(self, traindata: np.ndarray) -> None: """ Trains on dataset """ self.clf.fit(traindata[:, 1:5], traindata[:, 5]) def predict(self, predictdata: np.ndarray) -> np.ndarray: """ predict given points """ return self.clf.predict(predictdata) def get_args() -> argparse.Namespace: """ Get program arguments. Just use --help.... """ parser = argparse.ArgumentParser(prog='python3 linefit.py', description=('Parameterize and analyze ' 'usability of conduit edge data')) parser.add_argument('files', metavar='F', type=str, nargs='*', help=('File(s) for processing. ' 'Each file has a specific format: ' 'See README (or header) for specification.')) parser.add_argument('-p', '--plot', action='store_true', default=False, help=('Create Plot of file(s)? Note, unless --time flag used, ' 'will plot middle time.')) parser.add_argument('-pd', '--plotdata', action='store_true', default=False, help='Create plot of current datastore.') parser.add_argument('-a', '--analyze', action='store_true', default=False, help=('Analyze the file and determine Curvature/Noise parameters. ' 'If --time not specified, will examine entire file. ' 'This will add results to datastore with false flags ' 'in accept field if not provided.')) parser.add_argument('-mt', '--machinetest', action='store_true', default=False, help=('Determine if the times from the file are usable based on ' 'supervised learning model. If --time not specified, ' 'will examine entire file.')) parser.add_argument('-m', '--model', type=str, default='nn', help=('Learning Model to use. Options are ["nn", "svm", "forest", "sgd"]')) parser.add_argument('-nnk', '--nnk', type=int, default=10, help=('k-Parameter for k nearest neighbors. Google it.')) parser.add_argument('-t', '--time', type=int, default=None, help=('Time (column) of data to use for analysis OR plotting. ' 'Zero-Indexed')) parser.add_argument('-d', '--datastore', type=str, default=DATASTORE, help=("Datastore filename override. " "Don't do this unless you know what you're doing")) parser.add_argument('-pds', '--printdata', action='store_true', default=False, help=("Print data")) parser.add_argument('-pdss', '--printdatashort', action='store_true', default=False, help=("Print data short")) args = parser.parse_args() return args if __name__ == '__main__': sys.exit(main())
Dispersive-Hydrodynamics-Lab/PACE
PACE/PACE.py
get_args
python
def get_args() -> argparse.Namespace: parser = argparse.ArgumentParser(prog='python3 linefit.py', description=('Parameterize and analyze ' 'usability of conduit edge data')) parser.add_argument('files', metavar='F', type=str, nargs='*', help=('File(s) for processing. ' 'Each file has a specific format: ' 'See README (or header) for specification.')) parser.add_argument('-p', '--plot', action='store_true', default=False, help=('Create Plot of file(s)? Note, unless --time flag used, ' 'will plot middle time.')) parser.add_argument('-pd', '--plotdata', action='store_true', default=False, help='Create plot of current datastore.') parser.add_argument('-a', '--analyze', action='store_true', default=False, help=('Analyze the file and determine Curvature/Noise parameters. ' 'If --time not specified, will examine entire file. ' 'This will add results to datastore with false flags ' 'in accept field if not provided.')) parser.add_argument('-mt', '--machinetest', action='store_true', default=False, help=('Determine if the times from the file are usable based on ' 'supervised learning model. If --time not specified, ' 'will examine entire file.')) parser.add_argument('-m', '--model', type=str, default='nn', help=('Learning Model to use. Options are ["nn", "svm", "forest", "sgd"]')) parser.add_argument('-nnk', '--nnk', type=int, default=10, help=('k-Parameter for k nearest neighbors. Google it.')) parser.add_argument('-t', '--time', type=int, default=None, help=('Time (column) of data to use for analysis OR plotting. ' 'Zero-Indexed')) parser.add_argument('-d', '--datastore', type=str, default=DATASTORE, help=("Datastore filename override. " "Don't do this unless you know what you're doing")) parser.add_argument('-pds', '--printdata', action='store_true', default=False, help=("Print data")) parser.add_argument('-pdss', '--printdatashort', action='store_true', default=False, help=("Print data short")) args = parser.parse_args() return args
Get program arguments. Just use --help....
train
https://github.com/Dispersive-Hydrodynamics-Lab/PACE/blob/4ce27d5fc9b02cc2ce55f6fea7fc8d6015317e1f/PACE/PACE.py#L517-L559
null
#!/usr/bin/env python3.5 """ PACE TODO: * model training/testing * more models (technically) * multithreading """ import sys import os import argparse import hashlib import typing from enforce import runtime_validation as types from tqdm import tqdm import numpy as np import numpy.linalg as linalg import matplotlib import matplotlib.pyplot as plt from matplotlib.colors import ListedColormap import scipy.integrate as si import scipy.io as sco import sklearn as sk from sklearn import svm from sklearn import preprocessing from sklearn import neighbors DATASTORE = 'linefitdata.mat' HEADER = (' ____ _ ____ _____\n' '| _ \ / \ / ___| ____|\n' '| |_) / _ \| | | _|\n' '| __/ ___ \ |___| |___\n' '|_| /_/ \_\____|_____|\n\n' 'PACE: Parameterization & Analysis of Conduit Edges\n' 'William Farmer - 2015\n') def main(): args = get_args() data = DataStore(DATASTORE) data.load() # Establish directory for img outputs if not os.path.exists('./img'): os.makedirs('./img') if args.plot: for filename in args.files: print('Plotting ' + filename) plot_name = './img/' + filename + '.general_fit.png' fit = LineFit(filename) fit.plot_file(name=plot_name, time=args.time) if args.analyze: for filename in args.files: manage_file_analysis(args, filename, data) if args.plotdata: data.plot_traindata() if args.machinetest: learner = ML(algo=args.model) if args.printdata: data.printdata() if args.printdatashort: data.printshort() @types def manage_file_analysis(args: argparse.Namespace, filename: str, data: object) -> None: """ Take care of the analysis of a datafile """ key = DataStore.hashfile(filename) print('Analyzing {} --> {}'.format(filename, key)) if data.check_key(key): # if exists in database, prepopulate fit = LineFit(filename, data=data.get_data(key)) else: fit = LineFit(filename) if args.time: noise, curvature, rnge, domn = fit.analyze(time=args.time) newrow = [args.time, noise, curvature, rnge, domn, fit.accepts[args.time]] data.update1(key, newrow, len(fit.noises)) else: fit.analyze_full() newrows = np.array([range(len(fit.noises)), fit.noises, fit.curves, fit.ranges, fit.domains, fit.accepts]) data.update(key, newrows) data.save() class DataStore(object): def __init__(self, name: str): """ Uses a .mat as datastore for compatibility. Eventually may want to switch to SQLite, or some database? Not sure if ever needed. This class provides that extensible API structure however. Datafile has the following structure: learning_data = {filehash:[[trial_index, noise, curvature, range, domain, accept, viscosity] ,...],...} Conveniently, you can use the domain field as a check as to whether or not the row has been touched. If domain=0 (for that row) then that means that it hasn't been updated. :param: name of datastore """ self.name = name self.data = {} def load(self) -> None: """ Load datafile """ try: self.data = sco.loadmat(self.name) except FileNotFoundError: pass def save(self) -> None: """ Save datafile to disk """ sco.savemat(self.name, self.data) def get_data(self, key: str) -> np.ndarray: """ Returns the specified data. Warning, ZERO ERROR HANDLING :param key: name of file :return: 2d data array """ return self.data[key] @types def get_keys(self) -> typing.List[str]: """ Return list of SHA512 hash keys that exist in datafile :return: list of keys """ keys = [] for key in self.data.keys(): if key not in ['__header__', '__version__', '__globals__']: keys.append(key) return keys @types def check_key(self, key: str) -> bool: """ Checks if key exists in datastore. True if yes, False if no. :param: SHA512 hash key :return: whether or key not exists in datastore """ keys = self.get_keys() return key in keys def get_traindata(self) -> np.ndarray: """ Pulls all available data and concatenates for model training :return: 2d array of points """ traindata = None for key, value in self.data.items(): if key not in ['__header__', '__version__', '__globals__']: if traindata is None: traindata = value[np.where(value[:, 4] != 0)] else: traindata = np.concatenate((traindata, value[np.where(value[:, 4] != 0)])) return traindata @types def plot_traindata(self, name: str='dataplot') -> None: """ Plots traindata.... choo choo... """ traindata = self.get_traindata() plt.figure(figsize=(16, 16)) plt.scatter(traindata[:, 1], traindata[:, 2], c=traindata[:, 5], marker='o', label='Datastore Points') plt.xlabel(r'$\log_{10}$ Noise') plt.ylabel(r'$\log_{10}$ Curvature') plt.legend(loc=2, fontsize='xx-large') plt.savefig('./img/{}.png'.format(name)) def printdata(self) -> None: """ Prints data to stdout """ np.set_printoptions(threshold=np.nan) print(self.data) np.set_printoptions(threshold=1000) def printshort(self) -> None: """ Print shortened version of data to stdout""" print(self.data) @types def update(self, key: str, data: np.ndarray) -> None: """ Update entry in datastore """ self.data[key] = data def update1(self, key: str, data: np.ndarray, size: int) -> None: """ Update one entry in specific record in datastore """ print(data) if key in self.get_keys(): self.data[key][data[0]] = data else: newdata = np.zeros((size, 6)) newdata[data[0]] = data self.data[key] = newdata @staticmethod @types def hashfile(name: str) -> str: """ Gets a hash of a file using block parsing http://stackoverflow.com/questions/3431825/generating-a-md5-checksum-of-a-file Using SHA512 for long-term support (hehehehe) """ hasher = hashlib.sha512() with open(name, 'rb') as openfile: for chunk in iter(lambda: openfile.read(4096), b''): hasher.update(chunk) return hasher.hexdigest() class LineFit(object): def __init__(self, filename: str, data: np.ndarray=None, function_number: int=16, spread_number: int=22): """ Main class for line fitting and parameter determination :param: filename :param: data for fitting :param: number of functions :param: gaussian spread number """ self.filename = filename (self.averagedata, self.times, self.accepts, self.ratio, self.viscosity) = self._loadedges() self.domain = np.arange(len(self.averagedata[:, 0])) self.function_number = function_number self.spread_number = spread_number if data is None: self.noises = np.zeros(len(self.times)) self.curves = np.zeros(len(self.times)) self.ranges = np.zeros(len(self.times)) self.domains = np.zeros(len(self.times)) else: self.noises = data[:, 1] self.curves = data[:, 2] self.ranges = data[:, 3] self.domains = data[:, 4] @types def _loadedges(self) -> typing.Tuple[np.ndarray, np.ndarray, np.ndarray, float, np.ndarray]: """ Attempts to intelligently load the .mat file and take average of left and right edges :return: left and right averages :return: times for each column :return: accept/reject for each column :return: pixel-inch ratio """ data = sco.loadmat(self.filename) datakeys = [k for k in data.keys() if ('right' in k) or ('left' in k) or ('edge' in k)] averagedata = ((data[datakeys[0]] + data[datakeys[1]]) / 2) try: times = (data['times'] - data['times'].min())[0] except KeyError: times = np.arange(len(data[datakeys[0]][0])) try: accept = data['accept'] except KeyError: accept = np.zeros(len(times)) try: ratio = data['ratio'] except KeyError: ratio = 1 try: viscosity = data['viscosity'] except KeyError: viscosity = np.ones(len(times)) return averagedata, times, accept, ratio, viscosity def plot_file(self, name: str=None, time: int=None) -> None: """ Plot specific time for provided datafile. If no time provided, will plot middle. :param: savefile name :param: time/data column """ if not time: time = int(len(self.times) / 2) if not name: name = './img/' + self.filename + '.png' yhat, residuals, residual_mean, noise = self._get_fit(time) plt.figure() plt.scatter(self.domain, self.averagedata[:, time], alpha=0.2) plt.plot(yhat) plt.savefig(name) @staticmethod @types def ddiff(arr: np.ndarray) -> np.ndarray: """ Helper Function: Divided Differences input: array """ return arr[:-1] - arr[1:] @types def _gaussian_function(self, datalength: int, values: np.ndarray, height: int, index: int) -> np.ndarray: """ i'th Regression Model Gaussian :param: len(x) :param: x values :param: height of gaussian :param: position of gaussian :return: gaussian bumps over domain """ return height * np.exp(-(1 / (self.spread_number * datalength)) * (values - ((datalength / self.function_number) * index)) ** 2) @types def _get_fit(self, time: int) -> typing.Tuple[np.ndarray, np.ndarray, float, float]: """ Fit regression model to data :param: time (column of data) :return: predicted points :return: residuals :return: mean residual :return: error """ rawdata = self.averagedata[:, time] domain = np.arange(len(rawdata)) datalength = len(domain) coefficients = np.zeros((datalength, self.function_number + 2)) coefficients[:, 0] = 1 coefficients[:, 1] = domain for i in range(self.function_number): coefficients[:, 2 + i] = self._gaussian_function(datalength, domain, 1, i) betas = linalg.inv(coefficients.transpose().dot(coefficients)).dot(coefficients.transpose().dot(rawdata)) predicted_values = coefficients.dot(betas) residuals = rawdata - predicted_values error = np.sqrt(residuals.transpose().dot(residuals) / (datalength - (self.function_number + 2))) return predicted_values, residuals, residuals.mean(), error @types def _get_noise(self, residuals: np.ndarray) -> float: """ Determine Noise of Residuals. :param: residuals :return: noise """ return np.mean(np.abs(residuals)) @types def analyze(self, time: int=None) -> typing.Tuple[float, float, int, int]: """ Determine noise, curvature, range, and domain of specified array. :param: pixel to inch ratio :param: time (column) to use. :return: curvature :return: noise :return: range :return: domain """ if not time: time = int(len(self.times) / 2) if self.domains[time] == 0: yhat, residuals, mean_residual, error = self._get_fit(time) yhat_p = self.ddiff(yhat) yhat_pp = self.ddiff(yhat_p) noise = self._get_noise(residuals) curvature = (1 / self.ratio) * (1 / len(yhat_pp)) * np.sqrt(si.simps(yhat_pp ** 2)) rng = (self.ratio * (np.max(self.averagedata[:, time]) - np.min(self.averagedata[:, time]))) dmn = self.ratio * len(self.averagedata[:, time]) self.noises[time] = np.log10(noise) self.curves[time] = np.log10(curvature) self.ranges[time] = np.log10(rng) self.domains[time] = np.log10(dmn) return self.noises[time], self.curves[time], self.ranges[time], self.domains[time] @types def analyze_full(self) -> typing.Tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray]: """ Determine noise, curvature, range, and domain of specified data. Like analyze, except examines the entire file. :param: float->pixel to inch ratio :return: array->curvatures :return: array->noises :return: array->ranges :return: array->domains """ if self.noises[0] == 0: timelength = len(self.times) for i in tqdm(range(timelength)): self.analyze(time=i) return self.noises, self.curves, self.ranges, self.domains class ML(object): def __init__(self, args: argparse.Namespace, algo: str='nn'): """ Machine Learning to determine usability of data.... """ self.algo = self.get_algo(args, algo) def get_algo(self, args: argparse.Namespace, algo: str) -> object: """ Returns machine learning algorithm based on arguments """ if algo == 'nn': return NearestNeighbor(args.nnk) def train(self) -> None: """ Trains specified algorithm """ traindata = self.get_data() self.algo.train(traindata) def get_data(self) -> np.ndarray: """ Gets data for training We use the domain column to determine what fields have been filled out If the domain is zero (i.e. not in error) than we should probably ignore it anyway """ traindata = data.get_traindata() return traindata def plot_fitspace(self, name: str, X: np.ndarray, y: np.ndarray, clf: object) -> None: """ Plot 2dplane of fitspace """ cmap_light = ListedColormap(['#FFAAAA', '#AAFFAA', '#AAAAFF']) cmap_bold = ListedColormap(['#FF0000', '#00FF00', '#0000FF']) # Plot the decision boundary. For that, we will assign a color to each # point in the mesh [x_min, m_max]x[y_min, y_max]. h = 0.01 # Mesh step size x_min, x_max = X[:, 0].min() - 1, X[:, 0].max() + 1 y_min, y_max = X[:, 1].min() - 1, X[:, 1].max() + 1 xx, yy = np.meshgrid(np.arange(x_min, x_max, h), np.arange(y_min, y_max, h)) Z = clf.predict(np.c_[xx.ravel(), yy.ravel()]) # Put the result into a color plot Z = Z.reshape(xx.shape) plt.figure() plt.pcolormesh(xx, yy, Z, cmap=cmap_light) # Plot also the training points plt.scatter(X[:, 0], X[:, 1], c=y, cmap=cmap_bold) plt.xlim(xx.min(), xx.max()) plt.ylim(yy.min(), yy.max()) plt.xlabel(r'$\log_{10}$ Noise') plt.ylabel(r'$\log_{10}$ Curvature') plt.savefig(name) class NearestNeighbor(object): def __init__(self, k: int): """ An example machine learning model. EVERY MODEL NEEDS TO PROVIDE: 1. Train 2. Predict """ self.clf = neighbors.KNeighborsClassifier(k, weights='distance', p=2, algorithm='auto', n_jobs=8) def train(self, traindata: np.ndarray) -> None: """ Trains on dataset """ self.clf.fit(traindata[:, 1:5], traindata[:, 5]) def predict(self, predictdata: np.ndarray) -> np.ndarray: """ predict given points """ return self.clf.predict(predictdata) if __name__ == '__main__': sys.exit(main())
Dispersive-Hydrodynamics-Lab/PACE
PACE/PACE.py
DataStore.get_keys
python
def get_keys(self) -> typing.List[str]: keys = [] for key in self.data.keys(): if key not in ['__header__', '__version__', '__globals__']: keys.append(key) return keys
Return list of SHA512 hash keys that exist in datafile :return: list of keys
train
https://github.com/Dispersive-Hydrodynamics-Lab/PACE/blob/4ce27d5fc9b02cc2ce55f6fea7fc8d6015317e1f/PACE/PACE.py#L151-L161
null
class DataStore(object): def __init__(self, name: str): """ Uses a .mat as datastore for compatibility. Eventually may want to switch to SQLite, or some database? Not sure if ever needed. This class provides that extensible API structure however. Datafile has the following structure: learning_data = {filehash:[[trial_index, noise, curvature, range, domain, accept, viscosity] ,...],...} Conveniently, you can use the domain field as a check as to whether or not the row has been touched. If domain=0 (for that row) then that means that it hasn't been updated. :param: name of datastore """ self.name = name self.data = {} def load(self) -> None: """ Load datafile """ try: self.data = sco.loadmat(self.name) except FileNotFoundError: pass def save(self) -> None: """ Save datafile to disk """ sco.savemat(self.name, self.data) def get_data(self, key: str) -> np.ndarray: """ Returns the specified data. Warning, ZERO ERROR HANDLING :param key: name of file :return: 2d data array """ return self.data[key] @types @types def check_key(self, key: str) -> bool: """ Checks if key exists in datastore. True if yes, False if no. :param: SHA512 hash key :return: whether or key not exists in datastore """ keys = self.get_keys() return key in keys def get_traindata(self) -> np.ndarray: """ Pulls all available data and concatenates for model training :return: 2d array of points """ traindata = None for key, value in self.data.items(): if key not in ['__header__', '__version__', '__globals__']: if traindata is None: traindata = value[np.where(value[:, 4] != 0)] else: traindata = np.concatenate((traindata, value[np.where(value[:, 4] != 0)])) return traindata @types def plot_traindata(self, name: str='dataplot') -> None: """ Plots traindata.... choo choo... """ traindata = self.get_traindata() plt.figure(figsize=(16, 16)) plt.scatter(traindata[:, 1], traindata[:, 2], c=traindata[:, 5], marker='o', label='Datastore Points') plt.xlabel(r'$\log_{10}$ Noise') plt.ylabel(r'$\log_{10}$ Curvature') plt.legend(loc=2, fontsize='xx-large') plt.savefig('./img/{}.png'.format(name)) def printdata(self) -> None: """ Prints data to stdout """ np.set_printoptions(threshold=np.nan) print(self.data) np.set_printoptions(threshold=1000) def printshort(self) -> None: """ Print shortened version of data to stdout""" print(self.data) @types def update(self, key: str, data: np.ndarray) -> None: """ Update entry in datastore """ self.data[key] = data def update1(self, key: str, data: np.ndarray, size: int) -> None: """ Update one entry in specific record in datastore """ print(data) if key in self.get_keys(): self.data[key][data[0]] = data else: newdata = np.zeros((size, 6)) newdata[data[0]] = data self.data[key] = newdata @staticmethod @types def hashfile(name: str) -> str: """ Gets a hash of a file using block parsing http://stackoverflow.com/questions/3431825/generating-a-md5-checksum-of-a-file Using SHA512 for long-term support (hehehehe) """ hasher = hashlib.sha512() with open(name, 'rb') as openfile: for chunk in iter(lambda: openfile.read(4096), b''): hasher.update(chunk) return hasher.hexdigest()
Dispersive-Hydrodynamics-Lab/PACE
PACE/PACE.py
DataStore.check_key
python
def check_key(self, key: str) -> bool: keys = self.get_keys() return key in keys
Checks if key exists in datastore. True if yes, False if no. :param: SHA512 hash key :return: whether or key not exists in datastore
train
https://github.com/Dispersive-Hydrodynamics-Lab/PACE/blob/4ce27d5fc9b02cc2ce55f6fea7fc8d6015317e1f/PACE/PACE.py#L164-L173
null
class DataStore(object): def __init__(self, name: str): """ Uses a .mat as datastore for compatibility. Eventually may want to switch to SQLite, or some database? Not sure if ever needed. This class provides that extensible API structure however. Datafile has the following structure: learning_data = {filehash:[[trial_index, noise, curvature, range, domain, accept, viscosity] ,...],...} Conveniently, you can use the domain field as a check as to whether or not the row has been touched. If domain=0 (for that row) then that means that it hasn't been updated. :param: name of datastore """ self.name = name self.data = {} def load(self) -> None: """ Load datafile """ try: self.data = sco.loadmat(self.name) except FileNotFoundError: pass def save(self) -> None: """ Save datafile to disk """ sco.savemat(self.name, self.data) def get_data(self, key: str) -> np.ndarray: """ Returns the specified data. Warning, ZERO ERROR HANDLING :param key: name of file :return: 2d data array """ return self.data[key] @types def get_keys(self) -> typing.List[str]: """ Return list of SHA512 hash keys that exist in datafile :return: list of keys """ keys = [] for key in self.data.keys(): if key not in ['__header__', '__version__', '__globals__']: keys.append(key) return keys @types def get_traindata(self) -> np.ndarray: """ Pulls all available data and concatenates for model training :return: 2d array of points """ traindata = None for key, value in self.data.items(): if key not in ['__header__', '__version__', '__globals__']: if traindata is None: traindata = value[np.where(value[:, 4] != 0)] else: traindata = np.concatenate((traindata, value[np.where(value[:, 4] != 0)])) return traindata @types def plot_traindata(self, name: str='dataplot') -> None: """ Plots traindata.... choo choo... """ traindata = self.get_traindata() plt.figure(figsize=(16, 16)) plt.scatter(traindata[:, 1], traindata[:, 2], c=traindata[:, 5], marker='o', label='Datastore Points') plt.xlabel(r'$\log_{10}$ Noise') plt.ylabel(r'$\log_{10}$ Curvature') plt.legend(loc=2, fontsize='xx-large') plt.savefig('./img/{}.png'.format(name)) def printdata(self) -> None: """ Prints data to stdout """ np.set_printoptions(threshold=np.nan) print(self.data) np.set_printoptions(threshold=1000) def printshort(self) -> None: """ Print shortened version of data to stdout""" print(self.data) @types def update(self, key: str, data: np.ndarray) -> None: """ Update entry in datastore """ self.data[key] = data def update1(self, key: str, data: np.ndarray, size: int) -> None: """ Update one entry in specific record in datastore """ print(data) if key in self.get_keys(): self.data[key][data[0]] = data else: newdata = np.zeros((size, 6)) newdata[data[0]] = data self.data[key] = newdata @staticmethod @types def hashfile(name: str) -> str: """ Gets a hash of a file using block parsing http://stackoverflow.com/questions/3431825/generating-a-md5-checksum-of-a-file Using SHA512 for long-term support (hehehehe) """ hasher = hashlib.sha512() with open(name, 'rb') as openfile: for chunk in iter(lambda: openfile.read(4096), b''): hasher.update(chunk) return hasher.hexdigest()
Dispersive-Hydrodynamics-Lab/PACE
PACE/PACE.py
DataStore.get_traindata
python
def get_traindata(self) -> np.ndarray: traindata = None for key, value in self.data.items(): if key not in ['__header__', '__version__', '__globals__']: if traindata is None: traindata = value[np.where(value[:, 4] != 0)] else: traindata = np.concatenate((traindata, value[np.where(value[:, 4] != 0)])) return traindata
Pulls all available data and concatenates for model training :return: 2d array of points
train
https://github.com/Dispersive-Hydrodynamics-Lab/PACE/blob/4ce27d5fc9b02cc2ce55f6fea7fc8d6015317e1f/PACE/PACE.py#L175-L188
null
class DataStore(object): def __init__(self, name: str): """ Uses a .mat as datastore for compatibility. Eventually may want to switch to SQLite, or some database? Not sure if ever needed. This class provides that extensible API structure however. Datafile has the following structure: learning_data = {filehash:[[trial_index, noise, curvature, range, domain, accept, viscosity] ,...],...} Conveniently, you can use the domain field as a check as to whether or not the row has been touched. If domain=0 (for that row) then that means that it hasn't been updated. :param: name of datastore """ self.name = name self.data = {} def load(self) -> None: """ Load datafile """ try: self.data = sco.loadmat(self.name) except FileNotFoundError: pass def save(self) -> None: """ Save datafile to disk """ sco.savemat(self.name, self.data) def get_data(self, key: str) -> np.ndarray: """ Returns the specified data. Warning, ZERO ERROR HANDLING :param key: name of file :return: 2d data array """ return self.data[key] @types def get_keys(self) -> typing.List[str]: """ Return list of SHA512 hash keys that exist in datafile :return: list of keys """ keys = [] for key in self.data.keys(): if key not in ['__header__', '__version__', '__globals__']: keys.append(key) return keys @types def check_key(self, key: str) -> bool: """ Checks if key exists in datastore. True if yes, False if no. :param: SHA512 hash key :return: whether or key not exists in datastore """ keys = self.get_keys() return key in keys @types def plot_traindata(self, name: str='dataplot') -> None: """ Plots traindata.... choo choo... """ traindata = self.get_traindata() plt.figure(figsize=(16, 16)) plt.scatter(traindata[:, 1], traindata[:, 2], c=traindata[:, 5], marker='o', label='Datastore Points') plt.xlabel(r'$\log_{10}$ Noise') plt.ylabel(r'$\log_{10}$ Curvature') plt.legend(loc=2, fontsize='xx-large') plt.savefig('./img/{}.png'.format(name)) def printdata(self) -> None: """ Prints data to stdout """ np.set_printoptions(threshold=np.nan) print(self.data) np.set_printoptions(threshold=1000) def printshort(self) -> None: """ Print shortened version of data to stdout""" print(self.data) @types def update(self, key: str, data: np.ndarray) -> None: """ Update entry in datastore """ self.data[key] = data def update1(self, key: str, data: np.ndarray, size: int) -> None: """ Update one entry in specific record in datastore """ print(data) if key in self.get_keys(): self.data[key][data[0]] = data else: newdata = np.zeros((size, 6)) newdata[data[0]] = data self.data[key] = newdata @staticmethod @types def hashfile(name: str) -> str: """ Gets a hash of a file using block parsing http://stackoverflow.com/questions/3431825/generating-a-md5-checksum-of-a-file Using SHA512 for long-term support (hehehehe) """ hasher = hashlib.sha512() with open(name, 'rb') as openfile: for chunk in iter(lambda: openfile.read(4096), b''): hasher.update(chunk) return hasher.hexdigest()
Dispersive-Hydrodynamics-Lab/PACE
PACE/PACE.py
DataStore.plot_traindata
python
def plot_traindata(self, name: str='dataplot') -> None: traindata = self.get_traindata() plt.figure(figsize=(16, 16)) plt.scatter(traindata[:, 1], traindata[:, 2], c=traindata[:, 5], marker='o', label='Datastore Points') plt.xlabel(r'$\log_{10}$ Noise') plt.ylabel(r'$\log_{10}$ Curvature') plt.legend(loc=2, fontsize='xx-large') plt.savefig('./img/{}.png'.format(name))
Plots traindata.... choo choo...
train
https://github.com/Dispersive-Hydrodynamics-Lab/PACE/blob/4ce27d5fc9b02cc2ce55f6fea7fc8d6015317e1f/PACE/PACE.py#L191-L203
[ "def get_traindata(self) -> np.ndarray:\n \"\"\"\n Pulls all available data and concatenates for model training\n\n :return: 2d array of points\n \"\"\"\n traindata = None\n for key, value in self.data.items():\n if key not in ['__header__', '__version__', '__globals__']:\n if traindata is None:\n traindata = value[np.where(value[:, 4] != 0)]\n else:\n traindata = np.concatenate((traindata, value[np.where(value[:, 4] != 0)]))\n return traindata\n" ]
class DataStore(object): def __init__(self, name: str): """ Uses a .mat as datastore for compatibility. Eventually may want to switch to SQLite, or some database? Not sure if ever needed. This class provides that extensible API structure however. Datafile has the following structure: learning_data = {filehash:[[trial_index, noise, curvature, range, domain, accept, viscosity] ,...],...} Conveniently, you can use the domain field as a check as to whether or not the row has been touched. If domain=0 (for that row) then that means that it hasn't been updated. :param: name of datastore """ self.name = name self.data = {} def load(self) -> None: """ Load datafile """ try: self.data = sco.loadmat(self.name) except FileNotFoundError: pass def save(self) -> None: """ Save datafile to disk """ sco.savemat(self.name, self.data) def get_data(self, key: str) -> np.ndarray: """ Returns the specified data. Warning, ZERO ERROR HANDLING :param key: name of file :return: 2d data array """ return self.data[key] @types def get_keys(self) -> typing.List[str]: """ Return list of SHA512 hash keys that exist in datafile :return: list of keys """ keys = [] for key in self.data.keys(): if key not in ['__header__', '__version__', '__globals__']: keys.append(key) return keys @types def check_key(self, key: str) -> bool: """ Checks if key exists in datastore. True if yes, False if no. :param: SHA512 hash key :return: whether or key not exists in datastore """ keys = self.get_keys() return key in keys def get_traindata(self) -> np.ndarray: """ Pulls all available data and concatenates for model training :return: 2d array of points """ traindata = None for key, value in self.data.items(): if key not in ['__header__', '__version__', '__globals__']: if traindata is None: traindata = value[np.where(value[:, 4] != 0)] else: traindata = np.concatenate((traindata, value[np.where(value[:, 4] != 0)])) return traindata @types def printdata(self) -> None: """ Prints data to stdout """ np.set_printoptions(threshold=np.nan) print(self.data) np.set_printoptions(threshold=1000) def printshort(self) -> None: """ Print shortened version of data to stdout""" print(self.data) @types def update(self, key: str, data: np.ndarray) -> None: """ Update entry in datastore """ self.data[key] = data def update1(self, key: str, data: np.ndarray, size: int) -> None: """ Update one entry in specific record in datastore """ print(data) if key in self.get_keys(): self.data[key][data[0]] = data else: newdata = np.zeros((size, 6)) newdata[data[0]] = data self.data[key] = newdata @staticmethod @types def hashfile(name: str) -> str: """ Gets a hash of a file using block parsing http://stackoverflow.com/questions/3431825/generating-a-md5-checksum-of-a-file Using SHA512 for long-term support (hehehehe) """ hasher = hashlib.sha512() with open(name, 'rb') as openfile: for chunk in iter(lambda: openfile.read(4096), b''): hasher.update(chunk) return hasher.hexdigest()
Dispersive-Hydrodynamics-Lab/PACE
PACE/PACE.py
DataStore.printdata
python
def printdata(self) -> None: np.set_printoptions(threshold=np.nan) print(self.data) np.set_printoptions(threshold=1000)
Prints data to stdout
train
https://github.com/Dispersive-Hydrodynamics-Lab/PACE/blob/4ce27d5fc9b02cc2ce55f6fea7fc8d6015317e1f/PACE/PACE.py#L205-L209
null
class DataStore(object): def __init__(self, name: str): """ Uses a .mat as datastore for compatibility. Eventually may want to switch to SQLite, or some database? Not sure if ever needed. This class provides that extensible API structure however. Datafile has the following structure: learning_data = {filehash:[[trial_index, noise, curvature, range, domain, accept, viscosity] ,...],...} Conveniently, you can use the domain field as a check as to whether or not the row has been touched. If domain=0 (for that row) then that means that it hasn't been updated. :param: name of datastore """ self.name = name self.data = {} def load(self) -> None: """ Load datafile """ try: self.data = sco.loadmat(self.name) except FileNotFoundError: pass def save(self) -> None: """ Save datafile to disk """ sco.savemat(self.name, self.data) def get_data(self, key: str) -> np.ndarray: """ Returns the specified data. Warning, ZERO ERROR HANDLING :param key: name of file :return: 2d data array """ return self.data[key] @types def get_keys(self) -> typing.List[str]: """ Return list of SHA512 hash keys that exist in datafile :return: list of keys """ keys = [] for key in self.data.keys(): if key not in ['__header__', '__version__', '__globals__']: keys.append(key) return keys @types def check_key(self, key: str) -> bool: """ Checks if key exists in datastore. True if yes, False if no. :param: SHA512 hash key :return: whether or key not exists in datastore """ keys = self.get_keys() return key in keys def get_traindata(self) -> np.ndarray: """ Pulls all available data and concatenates for model training :return: 2d array of points """ traindata = None for key, value in self.data.items(): if key not in ['__header__', '__version__', '__globals__']: if traindata is None: traindata = value[np.where(value[:, 4] != 0)] else: traindata = np.concatenate((traindata, value[np.where(value[:, 4] != 0)])) return traindata @types def plot_traindata(self, name: str='dataplot') -> None: """ Plots traindata.... choo choo... """ traindata = self.get_traindata() plt.figure(figsize=(16, 16)) plt.scatter(traindata[:, 1], traindata[:, 2], c=traindata[:, 5], marker='o', label='Datastore Points') plt.xlabel(r'$\log_{10}$ Noise') plt.ylabel(r'$\log_{10}$ Curvature') plt.legend(loc=2, fontsize='xx-large') plt.savefig('./img/{}.png'.format(name)) def printshort(self) -> None: """ Print shortened version of data to stdout""" print(self.data) @types def update(self, key: str, data: np.ndarray) -> None: """ Update entry in datastore """ self.data[key] = data def update1(self, key: str, data: np.ndarray, size: int) -> None: """ Update one entry in specific record in datastore """ print(data) if key in self.get_keys(): self.data[key][data[0]] = data else: newdata = np.zeros((size, 6)) newdata[data[0]] = data self.data[key] = newdata @staticmethod @types def hashfile(name: str) -> str: """ Gets a hash of a file using block parsing http://stackoverflow.com/questions/3431825/generating-a-md5-checksum-of-a-file Using SHA512 for long-term support (hehehehe) """ hasher = hashlib.sha512() with open(name, 'rb') as openfile: for chunk in iter(lambda: openfile.read(4096), b''): hasher.update(chunk) return hasher.hexdigest()
Dispersive-Hydrodynamics-Lab/PACE
PACE/PACE.py
DataStore.update
python
def update(self, key: str, data: np.ndarray) -> None: self.data[key] = data
Update entry in datastore
train
https://github.com/Dispersive-Hydrodynamics-Lab/PACE/blob/4ce27d5fc9b02cc2ce55f6fea7fc8d6015317e1f/PACE/PACE.py#L216-L218
null
class DataStore(object): def __init__(self, name: str): """ Uses a .mat as datastore for compatibility. Eventually may want to switch to SQLite, or some database? Not sure if ever needed. This class provides that extensible API structure however. Datafile has the following structure: learning_data = {filehash:[[trial_index, noise, curvature, range, domain, accept, viscosity] ,...],...} Conveniently, you can use the domain field as a check as to whether or not the row has been touched. If domain=0 (for that row) then that means that it hasn't been updated. :param: name of datastore """ self.name = name self.data = {} def load(self) -> None: """ Load datafile """ try: self.data = sco.loadmat(self.name) except FileNotFoundError: pass def save(self) -> None: """ Save datafile to disk """ sco.savemat(self.name, self.data) def get_data(self, key: str) -> np.ndarray: """ Returns the specified data. Warning, ZERO ERROR HANDLING :param key: name of file :return: 2d data array """ return self.data[key] @types def get_keys(self) -> typing.List[str]: """ Return list of SHA512 hash keys that exist in datafile :return: list of keys """ keys = [] for key in self.data.keys(): if key not in ['__header__', '__version__', '__globals__']: keys.append(key) return keys @types def check_key(self, key: str) -> bool: """ Checks if key exists in datastore. True if yes, False if no. :param: SHA512 hash key :return: whether or key not exists in datastore """ keys = self.get_keys() return key in keys def get_traindata(self) -> np.ndarray: """ Pulls all available data and concatenates for model training :return: 2d array of points """ traindata = None for key, value in self.data.items(): if key not in ['__header__', '__version__', '__globals__']: if traindata is None: traindata = value[np.where(value[:, 4] != 0)] else: traindata = np.concatenate((traindata, value[np.where(value[:, 4] != 0)])) return traindata @types def plot_traindata(self, name: str='dataplot') -> None: """ Plots traindata.... choo choo... """ traindata = self.get_traindata() plt.figure(figsize=(16, 16)) plt.scatter(traindata[:, 1], traindata[:, 2], c=traindata[:, 5], marker='o', label='Datastore Points') plt.xlabel(r'$\log_{10}$ Noise') plt.ylabel(r'$\log_{10}$ Curvature') plt.legend(loc=2, fontsize='xx-large') plt.savefig('./img/{}.png'.format(name)) def printdata(self) -> None: """ Prints data to stdout """ np.set_printoptions(threshold=np.nan) print(self.data) np.set_printoptions(threshold=1000) def printshort(self) -> None: """ Print shortened version of data to stdout""" print(self.data) @types def update1(self, key: str, data: np.ndarray, size: int) -> None: """ Update one entry in specific record in datastore """ print(data) if key in self.get_keys(): self.data[key][data[0]] = data else: newdata = np.zeros((size, 6)) newdata[data[0]] = data self.data[key] = newdata @staticmethod @types def hashfile(name: str) -> str: """ Gets a hash of a file using block parsing http://stackoverflow.com/questions/3431825/generating-a-md5-checksum-of-a-file Using SHA512 for long-term support (hehehehe) """ hasher = hashlib.sha512() with open(name, 'rb') as openfile: for chunk in iter(lambda: openfile.read(4096), b''): hasher.update(chunk) return hasher.hexdigest()
Dispersive-Hydrodynamics-Lab/PACE
PACE/PACE.py
DataStore.update1
python
def update1(self, key: str, data: np.ndarray, size: int) -> None: print(data) if key in self.get_keys(): self.data[key][data[0]] = data else: newdata = np.zeros((size, 6)) newdata[data[0]] = data self.data[key] = newdata
Update one entry in specific record in datastore
train
https://github.com/Dispersive-Hydrodynamics-Lab/PACE/blob/4ce27d5fc9b02cc2ce55f6fea7fc8d6015317e1f/PACE/PACE.py#L220-L228
null
class DataStore(object): def __init__(self, name: str): """ Uses a .mat as datastore for compatibility. Eventually may want to switch to SQLite, or some database? Not sure if ever needed. This class provides that extensible API structure however. Datafile has the following structure: learning_data = {filehash:[[trial_index, noise, curvature, range, domain, accept, viscosity] ,...],...} Conveniently, you can use the domain field as a check as to whether or not the row has been touched. If domain=0 (for that row) then that means that it hasn't been updated. :param: name of datastore """ self.name = name self.data = {} def load(self) -> None: """ Load datafile """ try: self.data = sco.loadmat(self.name) except FileNotFoundError: pass def save(self) -> None: """ Save datafile to disk """ sco.savemat(self.name, self.data) def get_data(self, key: str) -> np.ndarray: """ Returns the specified data. Warning, ZERO ERROR HANDLING :param key: name of file :return: 2d data array """ return self.data[key] @types def get_keys(self) -> typing.List[str]: """ Return list of SHA512 hash keys that exist in datafile :return: list of keys """ keys = [] for key in self.data.keys(): if key not in ['__header__', '__version__', '__globals__']: keys.append(key) return keys @types def check_key(self, key: str) -> bool: """ Checks if key exists in datastore. True if yes, False if no. :param: SHA512 hash key :return: whether or key not exists in datastore """ keys = self.get_keys() return key in keys def get_traindata(self) -> np.ndarray: """ Pulls all available data and concatenates for model training :return: 2d array of points """ traindata = None for key, value in self.data.items(): if key not in ['__header__', '__version__', '__globals__']: if traindata is None: traindata = value[np.where(value[:, 4] != 0)] else: traindata = np.concatenate((traindata, value[np.where(value[:, 4] != 0)])) return traindata @types def plot_traindata(self, name: str='dataplot') -> None: """ Plots traindata.... choo choo... """ traindata = self.get_traindata() plt.figure(figsize=(16, 16)) plt.scatter(traindata[:, 1], traindata[:, 2], c=traindata[:, 5], marker='o', label='Datastore Points') plt.xlabel(r'$\log_{10}$ Noise') plt.ylabel(r'$\log_{10}$ Curvature') plt.legend(loc=2, fontsize='xx-large') plt.savefig('./img/{}.png'.format(name)) def printdata(self) -> None: """ Prints data to stdout """ np.set_printoptions(threshold=np.nan) print(self.data) np.set_printoptions(threshold=1000) def printshort(self) -> None: """ Print shortened version of data to stdout""" print(self.data) @types def update(self, key: str, data: np.ndarray) -> None: """ Update entry in datastore """ self.data[key] = data @staticmethod @types def hashfile(name: str) -> str: """ Gets a hash of a file using block parsing http://stackoverflow.com/questions/3431825/generating-a-md5-checksum-of-a-file Using SHA512 for long-term support (hehehehe) """ hasher = hashlib.sha512() with open(name, 'rb') as openfile: for chunk in iter(lambda: openfile.read(4096), b''): hasher.update(chunk) return hasher.hexdigest()
Dispersive-Hydrodynamics-Lab/PACE
PACE/PACE.py
DataStore.hashfile
python
def hashfile(name: str) -> str: hasher = hashlib.sha512() with open(name, 'rb') as openfile: for chunk in iter(lambda: openfile.read(4096), b''): hasher.update(chunk) return hasher.hexdigest()
Gets a hash of a file using block parsing http://stackoverflow.com/questions/3431825/generating-a-md5-checksum-of-a-file Using SHA512 for long-term support (hehehehe)
train
https://github.com/Dispersive-Hydrodynamics-Lab/PACE/blob/4ce27d5fc9b02cc2ce55f6fea7fc8d6015317e1f/PACE/PACE.py#L232-L243
null
class DataStore(object): def __init__(self, name: str): """ Uses a .mat as datastore for compatibility. Eventually may want to switch to SQLite, or some database? Not sure if ever needed. This class provides that extensible API structure however. Datafile has the following structure: learning_data = {filehash:[[trial_index, noise, curvature, range, domain, accept, viscosity] ,...],...} Conveniently, you can use the domain field as a check as to whether or not the row has been touched. If domain=0 (for that row) then that means that it hasn't been updated. :param: name of datastore """ self.name = name self.data = {} def load(self) -> None: """ Load datafile """ try: self.data = sco.loadmat(self.name) except FileNotFoundError: pass def save(self) -> None: """ Save datafile to disk """ sco.savemat(self.name, self.data) def get_data(self, key: str) -> np.ndarray: """ Returns the specified data. Warning, ZERO ERROR HANDLING :param key: name of file :return: 2d data array """ return self.data[key] @types def get_keys(self) -> typing.List[str]: """ Return list of SHA512 hash keys that exist in datafile :return: list of keys """ keys = [] for key in self.data.keys(): if key not in ['__header__', '__version__', '__globals__']: keys.append(key) return keys @types def check_key(self, key: str) -> bool: """ Checks if key exists in datastore. True if yes, False if no. :param: SHA512 hash key :return: whether or key not exists in datastore """ keys = self.get_keys() return key in keys def get_traindata(self) -> np.ndarray: """ Pulls all available data and concatenates for model training :return: 2d array of points """ traindata = None for key, value in self.data.items(): if key not in ['__header__', '__version__', '__globals__']: if traindata is None: traindata = value[np.where(value[:, 4] != 0)] else: traindata = np.concatenate((traindata, value[np.where(value[:, 4] != 0)])) return traindata @types def plot_traindata(self, name: str='dataplot') -> None: """ Plots traindata.... choo choo... """ traindata = self.get_traindata() plt.figure(figsize=(16, 16)) plt.scatter(traindata[:, 1], traindata[:, 2], c=traindata[:, 5], marker='o', label='Datastore Points') plt.xlabel(r'$\log_{10}$ Noise') plt.ylabel(r'$\log_{10}$ Curvature') plt.legend(loc=2, fontsize='xx-large') plt.savefig('./img/{}.png'.format(name)) def printdata(self) -> None: """ Prints data to stdout """ np.set_printoptions(threshold=np.nan) print(self.data) np.set_printoptions(threshold=1000) def printshort(self) -> None: """ Print shortened version of data to stdout""" print(self.data) @types def update(self, key: str, data: np.ndarray) -> None: """ Update entry in datastore """ self.data[key] = data def update1(self, key: str, data: np.ndarray, size: int) -> None: """ Update one entry in specific record in datastore """ print(data) if key in self.get_keys(): self.data[key][data[0]] = data else: newdata = np.zeros((size, 6)) newdata[data[0]] = data self.data[key] = newdata @staticmethod @types
Dispersive-Hydrodynamics-Lab/PACE
PACE/PACE.py
LineFit._loadedges
python
def _loadedges(self) -> typing.Tuple[np.ndarray, np.ndarray, np.ndarray, float, np.ndarray]: data = sco.loadmat(self.filename) datakeys = [k for k in data.keys() if ('right' in k) or ('left' in k) or ('edge' in k)] averagedata = ((data[datakeys[0]] + data[datakeys[1]]) / 2) try: times = (data['times'] - data['times'].min())[0] except KeyError: times = np.arange(len(data[datakeys[0]][0])) try: accept = data['accept'] except KeyError: accept = np.zeros(len(times)) try: ratio = data['ratio'] except KeyError: ratio = 1 try: viscosity = data['viscosity'] except KeyError: viscosity = np.ones(len(times)) return averagedata, times, accept, ratio, viscosity
Attempts to intelligently load the .mat file and take average of left and right edges :return: left and right averages :return: times for each column :return: accept/reject for each column :return: pixel-inch ratio
train
https://github.com/Dispersive-Hydrodynamics-Lab/PACE/blob/4ce27d5fc9b02cc2ce55f6fea7fc8d6015317e1f/PACE/PACE.py#L275-L308
null
class LineFit(object): def __init__(self, filename: str, data: np.ndarray=None, function_number: int=16, spread_number: int=22): """ Main class for line fitting and parameter determination :param: filename :param: data for fitting :param: number of functions :param: gaussian spread number """ self.filename = filename (self.averagedata, self.times, self.accepts, self.ratio, self.viscosity) = self._loadedges() self.domain = np.arange(len(self.averagedata[:, 0])) self.function_number = function_number self.spread_number = spread_number if data is None: self.noises = np.zeros(len(self.times)) self.curves = np.zeros(len(self.times)) self.ranges = np.zeros(len(self.times)) self.domains = np.zeros(len(self.times)) else: self.noises = data[:, 1] self.curves = data[:, 2] self.ranges = data[:, 3] self.domains = data[:, 4] @types def plot_file(self, name: str=None, time: int=None) -> None: """ Plot specific time for provided datafile. If no time provided, will plot middle. :param: savefile name :param: time/data column """ if not time: time = int(len(self.times) / 2) if not name: name = './img/' + self.filename + '.png' yhat, residuals, residual_mean, noise = self._get_fit(time) plt.figure() plt.scatter(self.domain, self.averagedata[:, time], alpha=0.2) plt.plot(yhat) plt.savefig(name) @staticmethod @types def ddiff(arr: np.ndarray) -> np.ndarray: """ Helper Function: Divided Differences input: array """ return arr[:-1] - arr[1:] @types def _gaussian_function(self, datalength: int, values: np.ndarray, height: int, index: int) -> np.ndarray: """ i'th Regression Model Gaussian :param: len(x) :param: x values :param: height of gaussian :param: position of gaussian :return: gaussian bumps over domain """ return height * np.exp(-(1 / (self.spread_number * datalength)) * (values - ((datalength / self.function_number) * index)) ** 2) @types def _get_fit(self, time: int) -> typing.Tuple[np.ndarray, np.ndarray, float, float]: """ Fit regression model to data :param: time (column of data) :return: predicted points :return: residuals :return: mean residual :return: error """ rawdata = self.averagedata[:, time] domain = np.arange(len(rawdata)) datalength = len(domain) coefficients = np.zeros((datalength, self.function_number + 2)) coefficients[:, 0] = 1 coefficients[:, 1] = domain for i in range(self.function_number): coefficients[:, 2 + i] = self._gaussian_function(datalength, domain, 1, i) betas = linalg.inv(coefficients.transpose().dot(coefficients)).dot(coefficients.transpose().dot(rawdata)) predicted_values = coefficients.dot(betas) residuals = rawdata - predicted_values error = np.sqrt(residuals.transpose().dot(residuals) / (datalength - (self.function_number + 2))) return predicted_values, residuals, residuals.mean(), error @types def _get_noise(self, residuals: np.ndarray) -> float: """ Determine Noise of Residuals. :param: residuals :return: noise """ return np.mean(np.abs(residuals)) @types def analyze(self, time: int=None) -> typing.Tuple[float, float, int, int]: """ Determine noise, curvature, range, and domain of specified array. :param: pixel to inch ratio :param: time (column) to use. :return: curvature :return: noise :return: range :return: domain """ if not time: time = int(len(self.times) / 2) if self.domains[time] == 0: yhat, residuals, mean_residual, error = self._get_fit(time) yhat_p = self.ddiff(yhat) yhat_pp = self.ddiff(yhat_p) noise = self._get_noise(residuals) curvature = (1 / self.ratio) * (1 / len(yhat_pp)) * np.sqrt(si.simps(yhat_pp ** 2)) rng = (self.ratio * (np.max(self.averagedata[:, time]) - np.min(self.averagedata[:, time]))) dmn = self.ratio * len(self.averagedata[:, time]) self.noises[time] = np.log10(noise) self.curves[time] = np.log10(curvature) self.ranges[time] = np.log10(rng) self.domains[time] = np.log10(dmn) return self.noises[time], self.curves[time], self.ranges[time], self.domains[time] @types def analyze_full(self) -> typing.Tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray]: """ Determine noise, curvature, range, and domain of specified data. Like analyze, except examines the entire file. :param: float->pixel to inch ratio :return: array->curvatures :return: array->noises :return: array->ranges :return: array->domains """ if self.noises[0] == 0: timelength = len(self.times) for i in tqdm(range(timelength)): self.analyze(time=i) return self.noises, self.curves, self.ranges, self.domains
Dispersive-Hydrodynamics-Lab/PACE
PACE/PACE.py
LineFit.plot_file
python
def plot_file(self, name: str=None, time: int=None) -> None: if not time: time = int(len(self.times) / 2) if not name: name = './img/' + self.filename + '.png' yhat, residuals, residual_mean, noise = self._get_fit(time) plt.figure() plt.scatter(self.domain, self.averagedata[:, time], alpha=0.2) plt.plot(yhat) plt.savefig(name)
Plot specific time for provided datafile. If no time provided, will plot middle. :param: savefile name :param: time/data column
train
https://github.com/Dispersive-Hydrodynamics-Lab/PACE/blob/4ce27d5fc9b02cc2ce55f6fea7fc8d6015317e1f/PACE/PACE.py#L310-L326
null
class LineFit(object): def __init__(self, filename: str, data: np.ndarray=None, function_number: int=16, spread_number: int=22): """ Main class for line fitting and parameter determination :param: filename :param: data for fitting :param: number of functions :param: gaussian spread number """ self.filename = filename (self.averagedata, self.times, self.accepts, self.ratio, self.viscosity) = self._loadedges() self.domain = np.arange(len(self.averagedata[:, 0])) self.function_number = function_number self.spread_number = spread_number if data is None: self.noises = np.zeros(len(self.times)) self.curves = np.zeros(len(self.times)) self.ranges = np.zeros(len(self.times)) self.domains = np.zeros(len(self.times)) else: self.noises = data[:, 1] self.curves = data[:, 2] self.ranges = data[:, 3] self.domains = data[:, 4] @types def _loadedges(self) -> typing.Tuple[np.ndarray, np.ndarray, np.ndarray, float, np.ndarray]: """ Attempts to intelligently load the .mat file and take average of left and right edges :return: left and right averages :return: times for each column :return: accept/reject for each column :return: pixel-inch ratio """ data = sco.loadmat(self.filename) datakeys = [k for k in data.keys() if ('right' in k) or ('left' in k) or ('edge' in k)] averagedata = ((data[datakeys[0]] + data[datakeys[1]]) / 2) try: times = (data['times'] - data['times'].min())[0] except KeyError: times = np.arange(len(data[datakeys[0]][0])) try: accept = data['accept'] except KeyError: accept = np.zeros(len(times)) try: ratio = data['ratio'] except KeyError: ratio = 1 try: viscosity = data['viscosity'] except KeyError: viscosity = np.ones(len(times)) return averagedata, times, accept, ratio, viscosity @staticmethod @types def ddiff(arr: np.ndarray) -> np.ndarray: """ Helper Function: Divided Differences input: array """ return arr[:-1] - arr[1:] @types def _gaussian_function(self, datalength: int, values: np.ndarray, height: int, index: int) -> np.ndarray: """ i'th Regression Model Gaussian :param: len(x) :param: x values :param: height of gaussian :param: position of gaussian :return: gaussian bumps over domain """ return height * np.exp(-(1 / (self.spread_number * datalength)) * (values - ((datalength / self.function_number) * index)) ** 2) @types def _get_fit(self, time: int) -> typing.Tuple[np.ndarray, np.ndarray, float, float]: """ Fit regression model to data :param: time (column of data) :return: predicted points :return: residuals :return: mean residual :return: error """ rawdata = self.averagedata[:, time] domain = np.arange(len(rawdata)) datalength = len(domain) coefficients = np.zeros((datalength, self.function_number + 2)) coefficients[:, 0] = 1 coefficients[:, 1] = domain for i in range(self.function_number): coefficients[:, 2 + i] = self._gaussian_function(datalength, domain, 1, i) betas = linalg.inv(coefficients.transpose().dot(coefficients)).dot(coefficients.transpose().dot(rawdata)) predicted_values = coefficients.dot(betas) residuals = rawdata - predicted_values error = np.sqrt(residuals.transpose().dot(residuals) / (datalength - (self.function_number + 2))) return predicted_values, residuals, residuals.mean(), error @types def _get_noise(self, residuals: np.ndarray) -> float: """ Determine Noise of Residuals. :param: residuals :return: noise """ return np.mean(np.abs(residuals)) @types def analyze(self, time: int=None) -> typing.Tuple[float, float, int, int]: """ Determine noise, curvature, range, and domain of specified array. :param: pixel to inch ratio :param: time (column) to use. :return: curvature :return: noise :return: range :return: domain """ if not time: time = int(len(self.times) / 2) if self.domains[time] == 0: yhat, residuals, mean_residual, error = self._get_fit(time) yhat_p = self.ddiff(yhat) yhat_pp = self.ddiff(yhat_p) noise = self._get_noise(residuals) curvature = (1 / self.ratio) * (1 / len(yhat_pp)) * np.sqrt(si.simps(yhat_pp ** 2)) rng = (self.ratio * (np.max(self.averagedata[:, time]) - np.min(self.averagedata[:, time]))) dmn = self.ratio * len(self.averagedata[:, time]) self.noises[time] = np.log10(noise) self.curves[time] = np.log10(curvature) self.ranges[time] = np.log10(rng) self.domains[time] = np.log10(dmn) return self.noises[time], self.curves[time], self.ranges[time], self.domains[time] @types def analyze_full(self) -> typing.Tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray]: """ Determine noise, curvature, range, and domain of specified data. Like analyze, except examines the entire file. :param: float->pixel to inch ratio :return: array->curvatures :return: array->noises :return: array->ranges :return: array->domains """ if self.noises[0] == 0: timelength = len(self.times) for i in tqdm(range(timelength)): self.analyze(time=i) return self.noises, self.curves, self.ranges, self.domains
Dispersive-Hydrodynamics-Lab/PACE
PACE/PACE.py
LineFit._gaussian_function
python
def _gaussian_function(self, datalength: int, values: np.ndarray, height: int, index: int) -> np.ndarray: return height * np.exp(-(1 / (self.spread_number * datalength)) * (values - ((datalength / self.function_number) * index)) ** 2)
i'th Regression Model Gaussian :param: len(x) :param: x values :param: height of gaussian :param: position of gaussian :return: gaussian bumps over domain
train
https://github.com/Dispersive-Hydrodynamics-Lab/PACE/blob/4ce27d5fc9b02cc2ce55f6fea7fc8d6015317e1f/PACE/PACE.py#L339-L352
null
class LineFit(object): def __init__(self, filename: str, data: np.ndarray=None, function_number: int=16, spread_number: int=22): """ Main class for line fitting and parameter determination :param: filename :param: data for fitting :param: number of functions :param: gaussian spread number """ self.filename = filename (self.averagedata, self.times, self.accepts, self.ratio, self.viscosity) = self._loadedges() self.domain = np.arange(len(self.averagedata[:, 0])) self.function_number = function_number self.spread_number = spread_number if data is None: self.noises = np.zeros(len(self.times)) self.curves = np.zeros(len(self.times)) self.ranges = np.zeros(len(self.times)) self.domains = np.zeros(len(self.times)) else: self.noises = data[:, 1] self.curves = data[:, 2] self.ranges = data[:, 3] self.domains = data[:, 4] @types def _loadedges(self) -> typing.Tuple[np.ndarray, np.ndarray, np.ndarray, float, np.ndarray]: """ Attempts to intelligently load the .mat file and take average of left and right edges :return: left and right averages :return: times for each column :return: accept/reject for each column :return: pixel-inch ratio """ data = sco.loadmat(self.filename) datakeys = [k for k in data.keys() if ('right' in k) or ('left' in k) or ('edge' in k)] averagedata = ((data[datakeys[0]] + data[datakeys[1]]) / 2) try: times = (data['times'] - data['times'].min())[0] except KeyError: times = np.arange(len(data[datakeys[0]][0])) try: accept = data['accept'] except KeyError: accept = np.zeros(len(times)) try: ratio = data['ratio'] except KeyError: ratio = 1 try: viscosity = data['viscosity'] except KeyError: viscosity = np.ones(len(times)) return averagedata, times, accept, ratio, viscosity def plot_file(self, name: str=None, time: int=None) -> None: """ Plot specific time for provided datafile. If no time provided, will plot middle. :param: savefile name :param: time/data column """ if not time: time = int(len(self.times) / 2) if not name: name = './img/' + self.filename + '.png' yhat, residuals, residual_mean, noise = self._get_fit(time) plt.figure() plt.scatter(self.domain, self.averagedata[:, time], alpha=0.2) plt.plot(yhat) plt.savefig(name) @staticmethod @types def ddiff(arr: np.ndarray) -> np.ndarray: """ Helper Function: Divided Differences input: array """ return arr[:-1] - arr[1:] @types @types def _get_fit(self, time: int) -> typing.Tuple[np.ndarray, np.ndarray, float, float]: """ Fit regression model to data :param: time (column of data) :return: predicted points :return: residuals :return: mean residual :return: error """ rawdata = self.averagedata[:, time] domain = np.arange(len(rawdata)) datalength = len(domain) coefficients = np.zeros((datalength, self.function_number + 2)) coefficients[:, 0] = 1 coefficients[:, 1] = domain for i in range(self.function_number): coefficients[:, 2 + i] = self._gaussian_function(datalength, domain, 1, i) betas = linalg.inv(coefficients.transpose().dot(coefficients)).dot(coefficients.transpose().dot(rawdata)) predicted_values = coefficients.dot(betas) residuals = rawdata - predicted_values error = np.sqrt(residuals.transpose().dot(residuals) / (datalength - (self.function_number + 2))) return predicted_values, residuals, residuals.mean(), error @types def _get_noise(self, residuals: np.ndarray) -> float: """ Determine Noise of Residuals. :param: residuals :return: noise """ return np.mean(np.abs(residuals)) @types def analyze(self, time: int=None) -> typing.Tuple[float, float, int, int]: """ Determine noise, curvature, range, and domain of specified array. :param: pixel to inch ratio :param: time (column) to use. :return: curvature :return: noise :return: range :return: domain """ if not time: time = int(len(self.times) / 2) if self.domains[time] == 0: yhat, residuals, mean_residual, error = self._get_fit(time) yhat_p = self.ddiff(yhat) yhat_pp = self.ddiff(yhat_p) noise = self._get_noise(residuals) curvature = (1 / self.ratio) * (1 / len(yhat_pp)) * np.sqrt(si.simps(yhat_pp ** 2)) rng = (self.ratio * (np.max(self.averagedata[:, time]) - np.min(self.averagedata[:, time]))) dmn = self.ratio * len(self.averagedata[:, time]) self.noises[time] = np.log10(noise) self.curves[time] = np.log10(curvature) self.ranges[time] = np.log10(rng) self.domains[time] = np.log10(dmn) return self.noises[time], self.curves[time], self.ranges[time], self.domains[time] @types def analyze_full(self) -> typing.Tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray]: """ Determine noise, curvature, range, and domain of specified data. Like analyze, except examines the entire file. :param: float->pixel to inch ratio :return: array->curvatures :return: array->noises :return: array->ranges :return: array->domains """ if self.noises[0] == 0: timelength = len(self.times) for i in tqdm(range(timelength)): self.analyze(time=i) return self.noises, self.curves, self.ranges, self.domains
Dispersive-Hydrodynamics-Lab/PACE
PACE/PACE.py
LineFit._get_fit
python
def _get_fit(self, time: int) -> typing.Tuple[np.ndarray, np.ndarray, float, float]: rawdata = self.averagedata[:, time] domain = np.arange(len(rawdata)) datalength = len(domain) coefficients = np.zeros((datalength, self.function_number + 2)) coefficients[:, 0] = 1 coefficients[:, 1] = domain for i in range(self.function_number): coefficients[:, 2 + i] = self._gaussian_function(datalength, domain, 1, i) betas = linalg.inv(coefficients.transpose().dot(coefficients)).dot(coefficients.transpose().dot(rawdata)) predicted_values = coefficients.dot(betas) residuals = rawdata - predicted_values error = np.sqrt(residuals.transpose().dot(residuals) / (datalength - (self.function_number + 2))) return predicted_values, residuals, residuals.mean(), error
Fit regression model to data :param: time (column of data) :return: predicted points :return: residuals :return: mean residual :return: error
train
https://github.com/Dispersive-Hydrodynamics-Lab/PACE/blob/4ce27d5fc9b02cc2ce55f6fea7fc8d6015317e1f/PACE/PACE.py#L355-L378
null
class LineFit(object): def __init__(self, filename: str, data: np.ndarray=None, function_number: int=16, spread_number: int=22): """ Main class for line fitting and parameter determination :param: filename :param: data for fitting :param: number of functions :param: gaussian spread number """ self.filename = filename (self.averagedata, self.times, self.accepts, self.ratio, self.viscosity) = self._loadedges() self.domain = np.arange(len(self.averagedata[:, 0])) self.function_number = function_number self.spread_number = spread_number if data is None: self.noises = np.zeros(len(self.times)) self.curves = np.zeros(len(self.times)) self.ranges = np.zeros(len(self.times)) self.domains = np.zeros(len(self.times)) else: self.noises = data[:, 1] self.curves = data[:, 2] self.ranges = data[:, 3] self.domains = data[:, 4] @types def _loadedges(self) -> typing.Tuple[np.ndarray, np.ndarray, np.ndarray, float, np.ndarray]: """ Attempts to intelligently load the .mat file and take average of left and right edges :return: left and right averages :return: times for each column :return: accept/reject for each column :return: pixel-inch ratio """ data = sco.loadmat(self.filename) datakeys = [k for k in data.keys() if ('right' in k) or ('left' in k) or ('edge' in k)] averagedata = ((data[datakeys[0]] + data[datakeys[1]]) / 2) try: times = (data['times'] - data['times'].min())[0] except KeyError: times = np.arange(len(data[datakeys[0]][0])) try: accept = data['accept'] except KeyError: accept = np.zeros(len(times)) try: ratio = data['ratio'] except KeyError: ratio = 1 try: viscosity = data['viscosity'] except KeyError: viscosity = np.ones(len(times)) return averagedata, times, accept, ratio, viscosity def plot_file(self, name: str=None, time: int=None) -> None: """ Plot specific time for provided datafile. If no time provided, will plot middle. :param: savefile name :param: time/data column """ if not time: time = int(len(self.times) / 2) if not name: name = './img/' + self.filename + '.png' yhat, residuals, residual_mean, noise = self._get_fit(time) plt.figure() plt.scatter(self.domain, self.averagedata[:, time], alpha=0.2) plt.plot(yhat) plt.savefig(name) @staticmethod @types def ddiff(arr: np.ndarray) -> np.ndarray: """ Helper Function: Divided Differences input: array """ return arr[:-1] - arr[1:] @types def _gaussian_function(self, datalength: int, values: np.ndarray, height: int, index: int) -> np.ndarray: """ i'th Regression Model Gaussian :param: len(x) :param: x values :param: height of gaussian :param: position of gaussian :return: gaussian bumps over domain """ return height * np.exp(-(1 / (self.spread_number * datalength)) * (values - ((datalength / self.function_number) * index)) ** 2) @types @types def _get_noise(self, residuals: np.ndarray) -> float: """ Determine Noise of Residuals. :param: residuals :return: noise """ return np.mean(np.abs(residuals)) @types def analyze(self, time: int=None) -> typing.Tuple[float, float, int, int]: """ Determine noise, curvature, range, and domain of specified array. :param: pixel to inch ratio :param: time (column) to use. :return: curvature :return: noise :return: range :return: domain """ if not time: time = int(len(self.times) / 2) if self.domains[time] == 0: yhat, residuals, mean_residual, error = self._get_fit(time) yhat_p = self.ddiff(yhat) yhat_pp = self.ddiff(yhat_p) noise = self._get_noise(residuals) curvature = (1 / self.ratio) * (1 / len(yhat_pp)) * np.sqrt(si.simps(yhat_pp ** 2)) rng = (self.ratio * (np.max(self.averagedata[:, time]) - np.min(self.averagedata[:, time]))) dmn = self.ratio * len(self.averagedata[:, time]) self.noises[time] = np.log10(noise) self.curves[time] = np.log10(curvature) self.ranges[time] = np.log10(rng) self.domains[time] = np.log10(dmn) return self.noises[time], self.curves[time], self.ranges[time], self.domains[time] @types def analyze_full(self) -> typing.Tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray]: """ Determine noise, curvature, range, and domain of specified data. Like analyze, except examines the entire file. :param: float->pixel to inch ratio :return: array->curvatures :return: array->noises :return: array->ranges :return: array->domains """ if self.noises[0] == 0: timelength = len(self.times) for i in tqdm(range(timelength)): self.analyze(time=i) return self.noises, self.curves, self.ranges, self.domains
Dispersive-Hydrodynamics-Lab/PACE
PACE/PACE.py
LineFit._get_noise
python
def _get_noise(self, residuals: np.ndarray) -> float: return np.mean(np.abs(residuals))
Determine Noise of Residuals. :param: residuals :return: noise
train
https://github.com/Dispersive-Hydrodynamics-Lab/PACE/blob/4ce27d5fc9b02cc2ce55f6fea7fc8d6015317e1f/PACE/PACE.py#L381-L389
null
class LineFit(object): def __init__(self, filename: str, data: np.ndarray=None, function_number: int=16, spread_number: int=22): """ Main class for line fitting and parameter determination :param: filename :param: data for fitting :param: number of functions :param: gaussian spread number """ self.filename = filename (self.averagedata, self.times, self.accepts, self.ratio, self.viscosity) = self._loadedges() self.domain = np.arange(len(self.averagedata[:, 0])) self.function_number = function_number self.spread_number = spread_number if data is None: self.noises = np.zeros(len(self.times)) self.curves = np.zeros(len(self.times)) self.ranges = np.zeros(len(self.times)) self.domains = np.zeros(len(self.times)) else: self.noises = data[:, 1] self.curves = data[:, 2] self.ranges = data[:, 3] self.domains = data[:, 4] @types def _loadedges(self) -> typing.Tuple[np.ndarray, np.ndarray, np.ndarray, float, np.ndarray]: """ Attempts to intelligently load the .mat file and take average of left and right edges :return: left and right averages :return: times for each column :return: accept/reject for each column :return: pixel-inch ratio """ data = sco.loadmat(self.filename) datakeys = [k for k in data.keys() if ('right' in k) or ('left' in k) or ('edge' in k)] averagedata = ((data[datakeys[0]] + data[datakeys[1]]) / 2) try: times = (data['times'] - data['times'].min())[0] except KeyError: times = np.arange(len(data[datakeys[0]][0])) try: accept = data['accept'] except KeyError: accept = np.zeros(len(times)) try: ratio = data['ratio'] except KeyError: ratio = 1 try: viscosity = data['viscosity'] except KeyError: viscosity = np.ones(len(times)) return averagedata, times, accept, ratio, viscosity def plot_file(self, name: str=None, time: int=None) -> None: """ Plot specific time for provided datafile. If no time provided, will plot middle. :param: savefile name :param: time/data column """ if not time: time = int(len(self.times) / 2) if not name: name = './img/' + self.filename + '.png' yhat, residuals, residual_mean, noise = self._get_fit(time) plt.figure() plt.scatter(self.domain, self.averagedata[:, time], alpha=0.2) plt.plot(yhat) plt.savefig(name) @staticmethod @types def ddiff(arr: np.ndarray) -> np.ndarray: """ Helper Function: Divided Differences input: array """ return arr[:-1] - arr[1:] @types def _gaussian_function(self, datalength: int, values: np.ndarray, height: int, index: int) -> np.ndarray: """ i'th Regression Model Gaussian :param: len(x) :param: x values :param: height of gaussian :param: position of gaussian :return: gaussian bumps over domain """ return height * np.exp(-(1 / (self.spread_number * datalength)) * (values - ((datalength / self.function_number) * index)) ** 2) @types def _get_fit(self, time: int) -> typing.Tuple[np.ndarray, np.ndarray, float, float]: """ Fit regression model to data :param: time (column of data) :return: predicted points :return: residuals :return: mean residual :return: error """ rawdata = self.averagedata[:, time] domain = np.arange(len(rawdata)) datalength = len(domain) coefficients = np.zeros((datalength, self.function_number + 2)) coefficients[:, 0] = 1 coefficients[:, 1] = domain for i in range(self.function_number): coefficients[:, 2 + i] = self._gaussian_function(datalength, domain, 1, i) betas = linalg.inv(coefficients.transpose().dot(coefficients)).dot(coefficients.transpose().dot(rawdata)) predicted_values = coefficients.dot(betas) residuals = rawdata - predicted_values error = np.sqrt(residuals.transpose().dot(residuals) / (datalength - (self.function_number + 2))) return predicted_values, residuals, residuals.mean(), error @types @types def analyze(self, time: int=None) -> typing.Tuple[float, float, int, int]: """ Determine noise, curvature, range, and domain of specified array. :param: pixel to inch ratio :param: time (column) to use. :return: curvature :return: noise :return: range :return: domain """ if not time: time = int(len(self.times) / 2) if self.domains[time] == 0: yhat, residuals, mean_residual, error = self._get_fit(time) yhat_p = self.ddiff(yhat) yhat_pp = self.ddiff(yhat_p) noise = self._get_noise(residuals) curvature = (1 / self.ratio) * (1 / len(yhat_pp)) * np.sqrt(si.simps(yhat_pp ** 2)) rng = (self.ratio * (np.max(self.averagedata[:, time]) - np.min(self.averagedata[:, time]))) dmn = self.ratio * len(self.averagedata[:, time]) self.noises[time] = np.log10(noise) self.curves[time] = np.log10(curvature) self.ranges[time] = np.log10(rng) self.domains[time] = np.log10(dmn) return self.noises[time], self.curves[time], self.ranges[time], self.domains[time] @types def analyze_full(self) -> typing.Tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray]: """ Determine noise, curvature, range, and domain of specified data. Like analyze, except examines the entire file. :param: float->pixel to inch ratio :return: array->curvatures :return: array->noises :return: array->ranges :return: array->domains """ if self.noises[0] == 0: timelength = len(self.times) for i in tqdm(range(timelength)): self.analyze(time=i) return self.noises, self.curves, self.ranges, self.domains
Dispersive-Hydrodynamics-Lab/PACE
PACE/PACE.py
LineFit.analyze
python
def analyze(self, time: int=None) -> typing.Tuple[float, float, int, int]: if not time: time = int(len(self.times) / 2) if self.domains[time] == 0: yhat, residuals, mean_residual, error = self._get_fit(time) yhat_p = self.ddiff(yhat) yhat_pp = self.ddiff(yhat_p) noise = self._get_noise(residuals) curvature = (1 / self.ratio) * (1 / len(yhat_pp)) * np.sqrt(si.simps(yhat_pp ** 2)) rng = (self.ratio * (np.max(self.averagedata[:, time]) - np.min(self.averagedata[:, time]))) dmn = self.ratio * len(self.averagedata[:, time]) self.noises[time] = np.log10(noise) self.curves[time] = np.log10(curvature) self.ranges[time] = np.log10(rng) self.domains[time] = np.log10(dmn) return self.noises[time], self.curves[time], self.ranges[time], self.domains[time]
Determine noise, curvature, range, and domain of specified array. :param: pixel to inch ratio :param: time (column) to use. :return: curvature :return: noise :return: range :return: domain
train
https://github.com/Dispersive-Hydrodynamics-Lab/PACE/blob/4ce27d5fc9b02cc2ce55f6fea7fc8d6015317e1f/PACE/PACE.py#L392-L420
null
class LineFit(object): def __init__(self, filename: str, data: np.ndarray=None, function_number: int=16, spread_number: int=22): """ Main class for line fitting and parameter determination :param: filename :param: data for fitting :param: number of functions :param: gaussian spread number """ self.filename = filename (self.averagedata, self.times, self.accepts, self.ratio, self.viscosity) = self._loadedges() self.domain = np.arange(len(self.averagedata[:, 0])) self.function_number = function_number self.spread_number = spread_number if data is None: self.noises = np.zeros(len(self.times)) self.curves = np.zeros(len(self.times)) self.ranges = np.zeros(len(self.times)) self.domains = np.zeros(len(self.times)) else: self.noises = data[:, 1] self.curves = data[:, 2] self.ranges = data[:, 3] self.domains = data[:, 4] @types def _loadedges(self) -> typing.Tuple[np.ndarray, np.ndarray, np.ndarray, float, np.ndarray]: """ Attempts to intelligently load the .mat file and take average of left and right edges :return: left and right averages :return: times for each column :return: accept/reject for each column :return: pixel-inch ratio """ data = sco.loadmat(self.filename) datakeys = [k for k in data.keys() if ('right' in k) or ('left' in k) or ('edge' in k)] averagedata = ((data[datakeys[0]] + data[datakeys[1]]) / 2) try: times = (data['times'] - data['times'].min())[0] except KeyError: times = np.arange(len(data[datakeys[0]][0])) try: accept = data['accept'] except KeyError: accept = np.zeros(len(times)) try: ratio = data['ratio'] except KeyError: ratio = 1 try: viscosity = data['viscosity'] except KeyError: viscosity = np.ones(len(times)) return averagedata, times, accept, ratio, viscosity def plot_file(self, name: str=None, time: int=None) -> None: """ Plot specific time for provided datafile. If no time provided, will plot middle. :param: savefile name :param: time/data column """ if not time: time = int(len(self.times) / 2) if not name: name = './img/' + self.filename + '.png' yhat, residuals, residual_mean, noise = self._get_fit(time) plt.figure() plt.scatter(self.domain, self.averagedata[:, time], alpha=0.2) plt.plot(yhat) plt.savefig(name) @staticmethod @types def ddiff(arr: np.ndarray) -> np.ndarray: """ Helper Function: Divided Differences input: array """ return arr[:-1] - arr[1:] @types def _gaussian_function(self, datalength: int, values: np.ndarray, height: int, index: int) -> np.ndarray: """ i'th Regression Model Gaussian :param: len(x) :param: x values :param: height of gaussian :param: position of gaussian :return: gaussian bumps over domain """ return height * np.exp(-(1 / (self.spread_number * datalength)) * (values - ((datalength / self.function_number) * index)) ** 2) @types def _get_fit(self, time: int) -> typing.Tuple[np.ndarray, np.ndarray, float, float]: """ Fit regression model to data :param: time (column of data) :return: predicted points :return: residuals :return: mean residual :return: error """ rawdata = self.averagedata[:, time] domain = np.arange(len(rawdata)) datalength = len(domain) coefficients = np.zeros((datalength, self.function_number + 2)) coefficients[:, 0] = 1 coefficients[:, 1] = domain for i in range(self.function_number): coefficients[:, 2 + i] = self._gaussian_function(datalength, domain, 1, i) betas = linalg.inv(coefficients.transpose().dot(coefficients)).dot(coefficients.transpose().dot(rawdata)) predicted_values = coefficients.dot(betas) residuals = rawdata - predicted_values error = np.sqrt(residuals.transpose().dot(residuals) / (datalength - (self.function_number + 2))) return predicted_values, residuals, residuals.mean(), error @types def _get_noise(self, residuals: np.ndarray) -> float: """ Determine Noise of Residuals. :param: residuals :return: noise """ return np.mean(np.abs(residuals)) @types @types def analyze_full(self) -> typing.Tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray]: """ Determine noise, curvature, range, and domain of specified data. Like analyze, except examines the entire file. :param: float->pixel to inch ratio :return: array->curvatures :return: array->noises :return: array->ranges :return: array->domains """ if self.noises[0] == 0: timelength = len(self.times) for i in tqdm(range(timelength)): self.analyze(time=i) return self.noises, self.curves, self.ranges, self.domains
Dispersive-Hydrodynamics-Lab/PACE
PACE/PACE.py
LineFit.analyze_full
python
def analyze_full(self) -> typing.Tuple[np.ndarray, np.ndarray, np.ndarray, np.ndarray]: if self.noises[0] == 0: timelength = len(self.times) for i in tqdm(range(timelength)): self.analyze(time=i) return self.noises, self.curves, self.ranges, self.domains
Determine noise, curvature, range, and domain of specified data. Like analyze, except examines the entire file. :param: float->pixel to inch ratio :return: array->curvatures :return: array->noises :return: array->ranges :return: array->domains
train
https://github.com/Dispersive-Hydrodynamics-Lab/PACE/blob/4ce27d5fc9b02cc2ce55f6fea7fc8d6015317e1f/PACE/PACE.py#L423-L439
null
class LineFit(object): def __init__(self, filename: str, data: np.ndarray=None, function_number: int=16, spread_number: int=22): """ Main class for line fitting and parameter determination :param: filename :param: data for fitting :param: number of functions :param: gaussian spread number """ self.filename = filename (self.averagedata, self.times, self.accepts, self.ratio, self.viscosity) = self._loadedges() self.domain = np.arange(len(self.averagedata[:, 0])) self.function_number = function_number self.spread_number = spread_number if data is None: self.noises = np.zeros(len(self.times)) self.curves = np.zeros(len(self.times)) self.ranges = np.zeros(len(self.times)) self.domains = np.zeros(len(self.times)) else: self.noises = data[:, 1] self.curves = data[:, 2] self.ranges = data[:, 3] self.domains = data[:, 4] @types def _loadedges(self) -> typing.Tuple[np.ndarray, np.ndarray, np.ndarray, float, np.ndarray]: """ Attempts to intelligently load the .mat file and take average of left and right edges :return: left and right averages :return: times for each column :return: accept/reject for each column :return: pixel-inch ratio """ data = sco.loadmat(self.filename) datakeys = [k for k in data.keys() if ('right' in k) or ('left' in k) or ('edge' in k)] averagedata = ((data[datakeys[0]] + data[datakeys[1]]) / 2) try: times = (data['times'] - data['times'].min())[0] except KeyError: times = np.arange(len(data[datakeys[0]][0])) try: accept = data['accept'] except KeyError: accept = np.zeros(len(times)) try: ratio = data['ratio'] except KeyError: ratio = 1 try: viscosity = data['viscosity'] except KeyError: viscosity = np.ones(len(times)) return averagedata, times, accept, ratio, viscosity def plot_file(self, name: str=None, time: int=None) -> None: """ Plot specific time for provided datafile. If no time provided, will plot middle. :param: savefile name :param: time/data column """ if not time: time = int(len(self.times) / 2) if not name: name = './img/' + self.filename + '.png' yhat, residuals, residual_mean, noise = self._get_fit(time) plt.figure() plt.scatter(self.domain, self.averagedata[:, time], alpha=0.2) plt.plot(yhat) plt.savefig(name) @staticmethod @types def ddiff(arr: np.ndarray) -> np.ndarray: """ Helper Function: Divided Differences input: array """ return arr[:-1] - arr[1:] @types def _gaussian_function(self, datalength: int, values: np.ndarray, height: int, index: int) -> np.ndarray: """ i'th Regression Model Gaussian :param: len(x) :param: x values :param: height of gaussian :param: position of gaussian :return: gaussian bumps over domain """ return height * np.exp(-(1 / (self.spread_number * datalength)) * (values - ((datalength / self.function_number) * index)) ** 2) @types def _get_fit(self, time: int) -> typing.Tuple[np.ndarray, np.ndarray, float, float]: """ Fit regression model to data :param: time (column of data) :return: predicted points :return: residuals :return: mean residual :return: error """ rawdata = self.averagedata[:, time] domain = np.arange(len(rawdata)) datalength = len(domain) coefficients = np.zeros((datalength, self.function_number + 2)) coefficients[:, 0] = 1 coefficients[:, 1] = domain for i in range(self.function_number): coefficients[:, 2 + i] = self._gaussian_function(datalength, domain, 1, i) betas = linalg.inv(coefficients.transpose().dot(coefficients)).dot(coefficients.transpose().dot(rawdata)) predicted_values = coefficients.dot(betas) residuals = rawdata - predicted_values error = np.sqrt(residuals.transpose().dot(residuals) / (datalength - (self.function_number + 2))) return predicted_values, residuals, residuals.mean(), error @types def _get_noise(self, residuals: np.ndarray) -> float: """ Determine Noise of Residuals. :param: residuals :return: noise """ return np.mean(np.abs(residuals)) @types def analyze(self, time: int=None) -> typing.Tuple[float, float, int, int]: """ Determine noise, curvature, range, and domain of specified array. :param: pixel to inch ratio :param: time (column) to use. :return: curvature :return: noise :return: range :return: domain """ if not time: time = int(len(self.times) / 2) if self.domains[time] == 0: yhat, residuals, mean_residual, error = self._get_fit(time) yhat_p = self.ddiff(yhat) yhat_pp = self.ddiff(yhat_p) noise = self._get_noise(residuals) curvature = (1 / self.ratio) * (1 / len(yhat_pp)) * np.sqrt(si.simps(yhat_pp ** 2)) rng = (self.ratio * (np.max(self.averagedata[:, time]) - np.min(self.averagedata[:, time]))) dmn = self.ratio * len(self.averagedata[:, time]) self.noises[time] = np.log10(noise) self.curves[time] = np.log10(curvature) self.ranges[time] = np.log10(rng) self.domains[time] = np.log10(dmn) return self.noises[time], self.curves[time], self.ranges[time], self.domains[time] @types
Dispersive-Hydrodynamics-Lab/PACE
PACE/PACE.py
ML.get_algo
python
def get_algo(self, args: argparse.Namespace, algo: str) -> object: if algo == 'nn': return NearestNeighbor(args.nnk)
Returns machine learning algorithm based on arguments
train
https://github.com/Dispersive-Hydrodynamics-Lab/PACE/blob/4ce27d5fc9b02cc2ce55f6fea7fc8d6015317e1f/PACE/PACE.py#L449-L452
null
class ML(object): def __init__(self, args: argparse.Namespace, algo: str='nn'): """ Machine Learning to determine usability of data.... """ self.algo = self.get_algo(args, algo) def train(self) -> None: """ Trains specified algorithm """ traindata = self.get_data() self.algo.train(traindata) def get_data(self) -> np.ndarray: """ Gets data for training We use the domain column to determine what fields have been filled out If the domain is zero (i.e. not in error) than we should probably ignore it anyway """ traindata = data.get_traindata() return traindata def plot_fitspace(self, name: str, X: np.ndarray, y: np.ndarray, clf: object) -> None: """ Plot 2dplane of fitspace """ cmap_light = ListedColormap(['#FFAAAA', '#AAFFAA', '#AAAAFF']) cmap_bold = ListedColormap(['#FF0000', '#00FF00', '#0000FF']) # Plot the decision boundary. For that, we will assign a color to each # point in the mesh [x_min, m_max]x[y_min, y_max]. h = 0.01 # Mesh step size x_min, x_max = X[:, 0].min() - 1, X[:, 0].max() + 1 y_min, y_max = X[:, 1].min() - 1, X[:, 1].max() + 1 xx, yy = np.meshgrid(np.arange(x_min, x_max, h), np.arange(y_min, y_max, h)) Z = clf.predict(np.c_[xx.ravel(), yy.ravel()]) # Put the result into a color plot Z = Z.reshape(xx.shape) plt.figure() plt.pcolormesh(xx, yy, Z, cmap=cmap_light) # Plot also the training points plt.scatter(X[:, 0], X[:, 1], c=y, cmap=cmap_bold) plt.xlim(xx.min(), xx.max()) plt.ylim(yy.min(), yy.max()) plt.xlabel(r'$\log_{10}$ Noise') plt.ylabel(r'$\log_{10}$ Curvature') plt.savefig(name)