plateform
stringclasses
1 value
repo_name
stringlengths
13
113
name
stringlengths
3
74
ext
stringclasses
1 value
path
stringlengths
12
229
size
int64
23
843k
source_encoding
stringclasses
9 values
md5
stringlengths
32
32
text
stringlengths
23
843k
github
alcu/sms-master
unpadn.m
.m
sms-master/irt/utilities/unpadn.m
818
utf_8
efcd912abfb2bb3205ef2bc3c72ff01c
function out = unpadn(mat, newdim) %|function out = unpadn(mat, newdim) %| pad old to newdim, preserving 'center point' if nargin == 1 && streq(mat, 'test'), unpadn_test, return, end if nargin < 2, error(mfilename), help(mfilename), end olddim = size(mat); if any(newdim > olddim), error('must be smaller'), end idx ...
github
alcu/sms-master
isvar.m
.m
sms-master/irt/utilities/isvar.m
1,046
utf_8
1c4ab9cfce881a1a785ede660f1a49a3
%|function tf = isvar(name) %| %| determine if "name" is a variable in the caller's workspace. %| %| if argument is of the form 'name.field' or 'name.field1.field2' etc. %| then this uses isfield(st, 'field') recursively as needed %| %| Copyright 2000-01-01, Jeff Fessler, University of Michigan %| modified 2010-04-21 t...
github
alcu/sms-master
dft_sym_check.m
.m
sms-master/irt/utilities/dft_sym_check.m
1,133
utf_8
0c88b2192a03b4a602be0b1e654d48c9
function dft_sym_check(xk, varargin) %|function dft_sym_check(xk) %| see if array xk (usually containing DFT coefficients) %| obeys the conjugate symmetry conditions expected %| for its (inverse) DFT to be real. if nargin < 1, help(mfilename), error(mfilename), end if nargin == 1 && streq(xk, 'test'), dft_sym_check_...
github
alcu/sms-master
dsingle.m
.m
sms-master/irt/utilities/dsingle.m
320
utf_8
131f14ecc3813385227fdb653e79b58b
function [varargout] = dsingle(varargin) %function y = dsingle(x) % truncate a (possibly) double precision value to single precision, % then convert back to "double" using double6: % y = double6(single(x)); for ii=1:nargin varargout{ii} = dsingle1(varargin{ii}); end function y = dsingle1(x) y = double6(single(x));
github
alcu/sms-master
ir_bspline_basis.m
.m
sms-master/irt/utilities/ir_bspline_basis.m
1,613
utf_8
71dcabbbdc1bd0ec7eef85f60eea1750
function out = ir_bspline_basis(t, k, ti) %function out = ir_bspline_basis(t, k, ti) %| %| Return B-spline basis of order k >= 0, for knots ti, at locations t. %| This is useful for didactic plotting and small problems %| but is not efficient (compute or memory) for large problems. %| %| in %| t [M 1] values at which...
github
alcu/sms-master
jf_equal.m
.m
sms-master/irt/utilities/jf_equal.m
4,616
utf_8
7ce82f61e30ea40f34cec3863cecd53c
function yn = jf_equal(a, b, varargin) %|function yn = jf_equal(a, b, varargin) %| %| verify that the two arguments are equal. %| if not, print informative error message. %| See also: equivs %| note: to compare objects, use jf_equal(struct(a), struct(b)) %| %| option %| 'what' 'warn' 'fail' 'silent' %| default: ...
github
alcu/sms-master
hist_equal.m
.m
sms-master/irt/utilities/hist_equal.m
2,486
utf_8
6c2519a67e007b658b934393abf3cc25
function [nk, center] = hist_equal(data, ncent, varargin) %|function [nk, center] = hist_equal(data, ncent, varargin) %| %| fast histogram of multidimensional data into equally-spaced bins %| todo: accumarray? %| %| in %| data [N M] data values to be binned (M-dimensional) %| ncent [1 M] # of centroids for each dim...
github
alcu/sms-master
fwhm1.m
.m
sms-master/irt/utilities/fwhm1.m
2,232
utf_8
6c1ec1b60ba8ca217fcc0a6eaa5c6ff0
function [fw, hr, hl] = fwhm1(psfs, varargin) %|function [fw, hr, hl] = fwhm1(psfs, options) %| %| compute fwhm of point-spread function centered at pixel imid %| and half-right width and half-left width (fw = hr + hl) %| %| in %| psfs [np nc] column psf(s) %| %| option %| imid [1] which pixel is the middle (default:...
github
alcu/sms-master
ndgrid_jf.m
.m
sms-master/irt/utilities/ndgrid_jf.m
1,825
utf_8
81141c55ae5b933826ba00a65b23fcc3
function out = ndgrid_jf(otype, varargin) %function out = ndgrid_jf('cell', varargin) %function out = ndgrid_jf('mat', varargin) %| version of ndgrid where output is cell array (default) or large matrix %| think: %| [out{1} ... out{M}] = ndgrid_cell(in{1}, ..., in{M}); %| also supported is ndgrid_jf('mat'|'cell', {in1...
github
alcu/sms-master
kde_pmf_width.m
.m
sms-master/irt/utilities/kde_pmf_width.m
1,562
utf_8
dd7297e5a902121f9b1a4d8ec9a62a79
function dx = kde_pmf_width(x, varargin) %|function dx = kde_pmf_width(x, varargin) %| %| Determine bin width for kde_pmf1,2 %| %| in %| x [N 1] iid realizations of random variable pairs (x,y) %| %| option %| 'type' 'silverman' to choose dx based on rule-of-thumb %| using inter-quartile range (p48 of 1986 book) %|...
github
alcu/sms-master
arg_get.m
.m
sms-master/irt/utilities/arg_get.m
954
utf_8
59f03a8b702fe06e405e10e515fb5741
function [arg, ii] = arg_get(list, what, default) %function [arg, ii] = arg_get(list, what, default) % extract an argument (e.g., nx or pixel_size, etc.) % from a list (vertically concatenated strings) % e.g., as generated by arg_pair() or strvcat() % the *first* match is used. % Copyright 2005, Jeff Fessler, The Univ...
github
alcu/sms-master
filtmat.m
.m
sms-master/irt/utilities/filtmat.m
3,762
utf_8
f0327a52412e81316e1dac6e5babedd8
function [mat, nexp] = filtmat(arg0, arg1, arg2, arg3, arg4) %function mat = filtmat('1d', kernx, nx) % 1d nonperiodic %function mat = filtmat({'1d' 'per'}, kernx, nx) % 1d periodic %function mat = filtmat('2d,mul,sep', kernx, kerny, [nx ny]) % 2d multiplicatively separable nonperiodic %function mat = filtmat('3...
github
alcu/sms-master
remove_spaces.m
.m
sms-master/irt/utilities/remove_spaces.m
793
utf_8
44dcf5c3a35ebc4ce3ddd6e590aab8a4
function arg = remove_spaces(arg) %|function arg = remove_spaces(arg) %| replace extra spaces at ends of matrix string array with zeros. %| also add extra '0' at end to be sure to null terminate string %| Copyright May 2000, Jeff Fessler, University of Michigan if nargin < 1, help(mfilename), error(mfilename), end ...
github
alcu/sms-master
interp1_lagrange.m
.m
sms-master/irt/utilities/interp1_lagrange.m
1,201
utf_8
09cf3435bbf5c9b0a2e2c60d2ea18220
function y = interp1_lagrange(xi, yi, x) %|function y = interp1_lagrange(xi, yi, x) %| %| Lagrange (1D) interpolation: polynomial of order length(xi) %| %| in %| xi [n 1], yi [n 1], x [m 1] %| out %| y [m 1] if yi is empty, then "y" is the m,n interpolation matrix %| %| Copyright 2003-5-13, Jeff Fessler, University ...
github
alcu/sms-master
arg_pair.m
.m
sms-master/irt/utilities/arg_pair.m
1,519
utf_8
b0dfe2defa2a10ffaf1901c3a4bb22e6
function arg = arg_pair(varargin) %|function arg = arg_pair(varargin) %| %| construct AsciiArg pairs from input arguments %| for building arguments akin to .dsc file %| example: arg = arg_pair('system', 2, 'nx', 128, ...) %| %| Copyright May 2000, Jeff Fessler, University of Michigan if nargin == 1 && streq(varargi...
github
alcu/sms-master
upsample_rep.m
.m
sms-master/irt/utilities/upsample_rep.m
629
utf_8
3db6e5ee1beac6a80b0a8880a17d5c3d
function y = upsample_rep(x, m) %function y = upsample_rep(x, m) % upsample a 2D image a factor of m by simple replication if nargin == 1 && streq(x, 'test'), upsample_rep_test, return, end if nargin < 1, help(mfilename), error(mfilename), end if nargin < 2, m = [2 2]; end y = upsample1_rep(x, m(1)); y = upsample1_r...
github
alcu/sms-master
max_percent_diff.m
.m
sms-master/irt/utilities/max_percent_diff.m
1,971
utf_8
0b4ab863976402c87de0914aa0e5a0a0
function d = max_percent_diff(s1, s2, varargin) %|function d = max_percent_diff(s1, s2, [options]) %| %| compute the "maximum percent difference" between two signals: s1, s2 %| options %| 1 use both arguments as the normalizer %| string print this %| %| Copyright 2000-9-16, Jeff Fessler, University of Michigan if ...
github
alcu/sms-master
bspline_1d_coef.m
.m
sms-master/irt/utilities/bspline_1d_coef.m
5,198
utf_8
9bc2d9c0359293a99564442103415254
function coef = bspline_1d_coef(fn, varargin) %|function coef = bspline_1d_coef(fn, varargin) %| %| Given f[n], uniformly spaced samples of a 1D signal f(t), %| compute the bspline coefficients so that the following model interpolates: %| f(t) = \sum_{k=0}^{N-1} coef[k] b(t - k) %| %| in %| fn [N,(L)] 1d signal(s) of...
github
alcu/sms-master
ir_dctmtx.m
.m
sms-master/irt/utilities/ir_dctmtx.m
802
utf_8
44653be06318cf836a96ab9d73447c6b
function out = ir_dctmtx(siz) %|function out = ir_dctmtx(siz) %| %| 1D or 2D DCT orthonormal matrix %| %| Reference: %| Jain, Fundamentals of Digital Image Processing, p. 150. if nargin ~= 1, help(mfilename), error(mfilename), end if streq(siz, 'test'), ir_dctmtx_test, return, end ndim = numel(siz); switch(ndim) ca...
github
alcu/sms-master
downsample1.m
.m
sms-master/irt/utilities/downsample1.m
1,243
utf_8
369f7422aa823211d70adf0860f3cf47
function y = downsample1(x, down, varargin) %|function y = downsample1(x, down, varargin) %| downsample by factor m along first dimension by averaging %| %| in %| x [n1 (Nd)] %| down integer downsampling factor %| option %| 'warn' 0|1 warn if noninteger multiple (default: 1) %| out %| y [n1/down (Nd)] %| %| Copyri...
github
alcu/sms-master
ir_display_struct.m
.m
sms-master/irt/utilities/ir_display_struct.m
1,779
utf_8
7c7455f2167bace79f65a6f53bb34375
function ir_display_struct(ob, varargin) %function ir_display_struct(ob, varargin) %| %| display a structure concisely %| %| option %| 'pre' prefix %| %| if nargin < 1, help(mfilename), error(mfilename), end if streq(ob, 'test'), ir_display_struct_test, return, end arg.pre = ' '; arg = vararg_pair(arg, varargin); ...
github
alcu/sms-master
fwhm2.m
.m
sms-master/irt/utilities/fwhm2.m
3,909
utf_8
331d73b8b740a9f57509efcb1a44f4b1
function [fw, angle, rad] = fwhm2(psf, varargin) %|function [fw, angle, rad] = fwhm2(psf, [options]) %| %| compute 2d fwhm of point-spread function using contourc() %| %| in %| psf [nx ny] 2D psf image %| %| option %| 'dx' pixel size (default 1) %| 'dy' pixel size (default dx) %| 'level' default 0.5 (for "half" ma...
github
alcu/sms-master
poisson.m
.m
sms-master/irt/utilities/poisson.m
1,181
utf_8
8e0d9705c577cbea395c4ff374a6db08
function data = poisson(xm, seed, varargin) %|function data = poisson(xm, seed, [options]) %| %| option %| 'factor' see poisson2.m %| %| Generate Poisson random vector with mean xm. %| For small, use poisson1.m %| For large, use poisson2.m %| see num. rec. C, P. 222 %| %| Copyright 1997-4-29, Jeff Fessler, University ...
github
alcu/sms-master
nrms.m
.m
sms-master/irt/utilities/nrms.m
1,109
utf_8
042ba79752bcb9662b4db291ca79144b
function n = nrms(x, xtrue, arg) %|function n = nrms(x, xtrue, arg) %| normalized rms error %| in %| x [[Nd] Nrep] multiple x's for one xtrue %| xtrue [[Nd]] %| arg use '2' to do along 2nd dim for same size %| out %| scalar %| vector if x and xtrue have same size and arg = 2 if nargin == 1 && streq(x, 'test'), nrms...
github
alcu/sms-master
fld_read.m
.m
sms-master/irt/utilities/fld_read.m
12,384
utf_8
949270fc5eea294da1196f45881314b2
function [data, coord, dims] = fld_read(file, varargin) %|function [data, coord, dims] = fld_read(file, options) %| Read data from AVS format .fld file. %| %| options %| 'raw' 0|1 1: return raw data class (default), 0: return doubles %| 'slice' int specify which slice to read from 3D file. (0 ... nz-1) %| 'chat' 0|1 ...
github
alcu/sms-master
outer_sum.m
.m
sms-master/irt/utilities/outer_sum.m
1,359
utf_8
69c4a2a90d18ddf41881a913a8b870b9
function ss = outer_sum(xx,yy) %|function ss = outer_sum(xx,yy) %| %| compute an "outer sum" x + y' %| that is analogous to the "outer product" x * y' %| %| in %| xx [nx 1] %| yy [1 ny] %| more generally: xx [(dim)] + yy [L,1] -> xx [(dim) LL] %| out %| ss [nx ny] ss(i,j) = xx(i) + yy(j) %| %| Copyright 2001, Jeff F...
github
alcu/sms-master
hist_bin_int.m
.m
sms-master/irt/utilities/hist_bin_int.m
3,676
utf_8
5ef626bda38ef55053a8c12220fe8ca5
function [nk center] = hist_bin_int(data, varargin) %|function [nk center] = hist_bin_int(data, varargin) %| %| fast histogram of multidimensional data into equally-spaced bins %| todo: time this vs accumarray %| %| in %| data [N M] data values to be binned (M-dimensional) %| must be integer valued >= 1 %| %| opt...
github
alcu/sms-master
stackpick.m
.m
sms-master/irt/utilities/stackpick.m
1,295
utf_8
d556fae93130a2ff478217d92ef038c3
function y = stackpick(x, ii) %|function y = stackpick(x, ii) %| pick one (or more) of the last "slices" of a multi-dimensional array %| think x(:,:,...,:,ii) %| This is useful in conjunction with stackup(). %| Copyright 2005-6-18, Jeff Fessler, University of Michiga if nargin == 1 && streq(x, 'test'), stackpick_test...
github
alcu/sms-master
bspline_1d_interp.m
.m
sms-master/irt/utilities/bspline_1d_interp.m
4,024
utf_8
aeff151b8ab4888a0d126ef7666f70d5
function ft = bspline_1d_interp(fn, ti, varargin) %|function ft = bspline_1d_interp(fn, ti, varargin) %| %| given f[n], the "unit spaced" samples of a continuous signal f(t), %| perform b-spline interpolation at sample locations t_i using the model: %| f(t) = \sum_{k=0}^{N-1} coef_k b(t - k) %| where the coefficients...
github
alcu/sms-master
ticker.m
.m
sms-master/irt/utilities/ticker.m
3,188
utf_8
d37224bdcbb9b5731efdb04eb3ef9d42
function ticker(varargin) %|function ticker(varargin) %| %| This utility is used in functions that might take a long time %| to provide a status update so the user knows it is working. %| "ticker print" to make output print to command window (default) %| "ticker waitbar" to display using waitbar %| ticker(waittime) to...
github
alcu/sms-master
downsample3.m
.m
sms-master/irt/utilities/downsample3.m
1,726
utf_8
80f0c157a98c90470017f7acf690c3dc
function y = downsample3(x, m) %|function y = downsample3(x, m) %| downsample by averaging by integer factors %| m can be a scalar (same factor for both dimensions) %| or a 3-vector %| in %| x [nx ny nz] %| m [1] or [1 3] %| out %| y [nx/m ny/m nz/m] %| function modified to 3D space by: Taka Masuda if nargin == 1 &&...
github
alcu/sms-master
embed.m
.m
sms-master/irt/utilities/embed.m
2,266
utf_8
108e1a6cc7480bee8cb65d3ab148f405
function ff = embed(x, mask, varargin) %|function ff = embed(x, mask, varargin) %| embed x in nonzero elements of (logical) mask %| in %| x [np (L)] the "nonzero" pixels (lexicographically stacked) %| mask [(Nd)] logical array, np = sum(mask) %| option %| '*dim' {0|1} 0: [(N) (L)] (default); 1: return [(N) *L] %| o...
github
alcu/sms-master
lloyd_max_hist.m
.m
sms-master/irt/utilities/lloyd_max_hist.m
4,356
utf_8
2557953b284dd10d47d6d87fbd822e28
function centers = lloyd_max_hist(data, centers, MM, tol, max_iter, chat) %|function centers = lloyd_max_hist(data, centers, MM, tol, max_iter, chat) %| %| "improved" version of the lloyd-max algorithm for scalar quantizer design %| that saves computation by histogramming the data first. %| Before using this, try usi...
github
alcu/sms-master
num2list.m
.m
sms-master/irt/utilities/num2list.m
390
utf_8
0076ebb4ee93b833462e21640099a15d
function varargout = num2list(x) %function varargout = num2list(x) % % convert a numeric vector input into a comma separated list of output values % DOES NOT WORK! if nargin < 1, num2list_test, help(mfilename), error(mfilename), end %nargout %nargout = numel(x) %varargout = cell(size(x)); for ii=1:numel(x) varargou...
github
alcu/sms-master
cpu.m
.m
sms-master/irt/utilities/cpu.m
1,725
utf_8
27808bbfc5ecd102b70225051d68de73
function out = cpu(arg, varargin) %|function out = cpu(arg, varargin) %| %| cpu etic %| cpu etoc printarg %| tic/toc elapsed time %| %| cpu etic %| cpu tic %| cpu toc %| cpu toc printarg %| work like tic/toc except they use cpu time instead of wall time. %| %| also: %| If printarg begins with a ':' then caller_name p...
github
alcu/sms-master
kde_pmf1.m
.m
sms-master/irt/utilities/kde_pmf1.m
4,795
utf_8
3f239349a0b6c199cab6908bd89d81e7
function [pmf xs] = kde_pmf1(x, varargin) %|function [pmf xs] = kde_pmf1(x, varargin) %| %| Compute an empirical probability mass function (PMF) using interpolation %| that is similar to a kernel density estimate. %| This routine is useful for making PMFs (interpolated / smoothed) histograms %| for computing entropy ...
github
alcu/sms-master
poly_string.m
.m
sms-master/irt/utilities/poly_string.m
1,452
utf_8
5efc7525a6d221ad41a8e377311fd06c
function [str, d1, d2] = poly_string(order, varargin) %function [str, d1, d2] = poly_string(order, [options]) % make a string with all the terms in a 2d polynomial % up to given order, e.g. [1+0*x x y ...] % also return partial derivatives w.r.t. x and y % in % order % options % maxdegree maximum of sum of degrees (de...
github
alcu/sms-master
run_mfile_local.m
.m
sms-master/irt/utilities/run_mfile_local.m
2,046
utf_8
f7e03aaa6ba17509f0d9ad9c7d4cf7d6
function run_mfile_local(arg, varargin) %|function run_mfile_local(arg) %| run an mfile in a local environment so that workspace variables %| are untouched. useful for tests. %| arg can be a cell with many mfiles to run %| options %| 'draw' 1|0 1 to draw after each test (default: 0) %| 'pause' 1|0 1 to pause after e...
github
alcu/sms-master
minmax.m
.m
sms-master/irt/utilities/minmax.m
2,326
utf_8
cdf69beeced912f2887f4a8f16933b53
function r = minmax(x, varargin) %function r = minmax(x, dim) % Return mininum and maximum of values of input x. % Default is to examine the entire array. % Specify a dimension dim otherwise. % If dim is a string, print... % Copyright 2003, Jeff Fessler, University of Michigan if nargin < 1, help(mfilename), error(mf...
github
alcu/sms-master
masker.m
.m
sms-master/irt/utilities/masker.m
884
utf_8
de844e9fcd7044d558d79c5644951dc9
function y = masker(x, mask) %|function y = masker(x, mask) %| extract from x the nonzero elements within (logical) mask %| in %| x [(Nd) (L)] image(s) %| mask [(Nd)] logical array, np = sum(mask) %| out %| y [np (L)] concise columns %| %| See also: embed %| %| Copyright 2004-9-28, Jeff Fessler, University of Michig...
github
alcu/sms-master
truncate_precision.m
.m
sms-master/irt/utilities/truncate_precision.m
924
utf_8
0f0d50a4e0af0fded1105d0ee262b6d8
function y = truncate_precision(x, digits, varargin) %|function y = truncate_precision(x, digits, [option]) %| %| truncate x to "digits" signficant digits of precision %| option %| 'type' floor (default) | round | ceil %| %| Copyright 2008-11-05, Jeff Fessler, University of Michigan if nargin == 1 && streq(x, 'test...
github
alcu/sms-master
jf.m
.m
sms-master/irt/utilities/jf.m
4,442
utf_8
1d7f14c6111af255339f61d684d3f5be
function out = jf(varargin) %|function out = jf(varargin) %| various personalization routines %| 'mv var_old var_new' %| 'ncore' # of cores this machine has %| add optional 2nd argument to over-ride, e.g., jf('ncore', 4) %| 'whos' %| 'clf' %| 'nobar' preclude toolbar and menubar from figures to render faster! %| 'no...
github
alcu/sms-master
ir_project_simplex.m
.m
sms-master/irt/utilities/ir_project_simplex.m
1,197
utf_8
7893825d0f47f9fb58091d2bc0db2bf0
function x = ir_project_simplex(y) %|function x = ir_project_simplex(y) %| %| project an n-dim vector y to the simplex Dn %| Dn = { x : x n-dim, 1 >= x >= 0, sum(x) = 1} %| %| (c) Xiaojing Ye %| xyex19@gmail.com %| %| Algorithm is explained as in the linked document %| http://arxiv.org/abs/1101.6081 %| or %| http://u...
github
alcu/sms-master
ir_odwt2.m
.m
sms-master/irt/utilities/ir_odwt2.m
6,502
utf_8
a130ae050bcc9904f3e37aaa550652ef
function [coef codes] = ir_odwt2(x, varargin) %|function [coef codes] = ir_odwt2(x, varargin) %| %| 2D orthonormal discrete wavelet transform (ODWT) of 2D image x. %| Also provides the adjoint, which is the inverse ODWT. %| %| in %| x [N M] image %| %| option %| 'level' int default: 1 %| 'wname' char default: 'haar' ...
github
alcu/sms-master
streq.m
.m
sms-master/irt/utilities/streq.m
686
utf_8
450bda5cc228936f2f8f88c03d1c5513
function tf = streq(a, b, n) %|function tf = streq(a, b [,n]) %| %| return 1 if two strings "a" and "b" are equal %| (optionally checking only up to 1st n chars) %| caution: whereas strcmp allows comparisons of cell arrays, %| this routine allows only two strings. %| %| Jeff Fessler if nargin == 1 && strcmp(a, 'test...
github
alcu/sms-master
vararg_pair.m
.m
sms-master/irt/utilities/vararg_pair.m
6,074
utf_8
4e8a72331a91a7486d098a62afb2b108
function [opt, extra] = vararg_pair(opt, varargs, varargin) %|function [opt, extra] = vararg_pair(opt, varargs, [options]) %| %| Process name / value pairs, replacing the "default" field values %| of the opt structure with the user-specified values. %| This allows flexible argument order and "named arguments" somewha...
github
alcu/sms-master
reshapee.m
.m
sms-master/irt/utilities/reshapee.m
1,255
utf_8
203da0142afea972cf8d708e085ea2d1
function y = reshapee(x, varargin) %|function y = reshapee(x, varargin) %| %| reshape function that allows possibly one null argument, and all %| other arguments can be vectors, unlike matlab that requires scalars. %| example: reshape(rand(2*3*5,7), [2 3], [], 7) will become [2 3 5 7] %| %| in %| x [(*dim)] %| vararg...
github
alcu/sms-master
sinc_periodic.m
.m
sms-master/irt/utilities/sinc_periodic.m
1,401
utf_8
6ded04b3754f67a5d4372c05ed651386
function x = sinc_periodic(t, K) %|function x = sinc_periodic(t, K) %| periodic sinc function, obtained by replicates of a sinc: %| x(t) = \sum_l sinc(t - l K) = ... = sin(pi*t) / tan(pi*t/K) / K for K even. %| This function is bandlimited and its samples are an impulse train. %| It is closely related to the Dirichle...
github
alcu/sms-master
poisson2.m
.m
sms-master/irt/utilities/poisson2.m
1,861
utf_8
5445163bb4f713ca1acc3b929698dc69
function data = poisson2(xm, varargin) %|function data = poisson2(xm, [options]) %| %| in %| 'xm' float %| %| option %| 'seed' int seed for rng() %| 'factor' double a value < 1 for rejection method %| %| Generate Poisson random column vector with mean xm. %| Uses rejection method - good for large values of xm. ...
github
alcu/sms-master
min_cos_quad.m
.m
sms-master/irt/utilities/min_cos_quad.m
3,179
utf_8
27cc84604461ab9b4aeb93bca086ac8f
function t = min_cos_quad(m, p, b, c, niter) %|function t = min_cos_quad(m, p, b, c, niter) %| %| find the minimizer of the sinusoid + quadratic form: %| f(t) = m * (1-cos(t-p)) + b*t + c/2 t^2 %| %| This is useful for certain phase-related optimization transfer problems. %| %| Copyright 2003-11-9, Jeff Fessler, Univ...
github
alcu/sms-master
ir_struct_find_field.m
.m
sms-master/irt/utilities/ir_struct_find_field.m
1,254
utf_8
931cc673b7e57430dcd0758c6d1d0b3a
function [out ok] = ir_struct_find_field(ss, field) %|function [out ok] = ir_struct_find_field(ss, field) %| %| Look recursively through the fields of a structure for a given field. %| Objects like Fatrix and fatrix2 are converted to structs too for this. %| %| Returns ss.(field) or ss.?.(field) or ss.?.?.(field) etc...
github
alcu/sms-master
kde_pmf2.m
.m
sms-master/irt/utilities/kde_pmf2.m
6,300
utf_8
ab62e809502eea382a6f01bf2e37be4b
function [pmf xs ys] = kde_pmf2(x, y, varargin) %|function [pmf xs ys] = kde_pmf2(x, y, varargin) %| %| Compute a 2D empirical joint probability mass function (PMF) %| using interpolation that is similar to a kernel density estimate. %| This joint PMF (interpolated / smoothed histogram) is useful for %| computing joi...
github
alcu/sms-master
rms.m
.m
sms-master/irt/utilities/rms.m
779
utf_8
8e6f3ae3104ee31f7fea466a6a2d8f84
function [rhat, s] = rms(x) %|function [rhat, s] = rms(x) %| function [r, s] = rms(err) %| in %| x: NxM error vectors %| out %| r: Mx1 estimate of rms error %| s: Mx1 estimate of std. dev. of that estimate if nargin < 1, help(mfilename), error(mfilename), end if streq(x, 'test'), rms_test, return, end N = nrow(x)...
github
alcu/sms-master
stackup.m
.m
sms-master/irt/utilities/stackup.m
1,653
utf_8
258aada4c6220aa3ba3d2b64cdd35b72
function ss = stackup(varargin) %|function ss = stackup(x1, x2, ...) %| %| stack up 2D arrays to make 3D array %| or stack up 3D arrays to make 4D array %| This generalizes how [a; b] "stacks up" 1D vectors to make a 2D array. %| This is useful in conjunction with stackpick(). %| It is akin to cat(ndims(x1)+1, x1, x2...
github
alcu/sms-master
bspline_1d_synth.m
.m
sms-master/irt/utilities/bspline_1d_synth.m
7,643
utf_8
9211b4f0a910dff26ec64ab9859a9c5c
function ft = bspline_1d_synth(coef, ti, varargin) %|function ft = bspline_1d_synth(coef, ti, varargin) %| %| Given c[n], the b-spline coefficients for the following model: %| f(t) = \sum_{k=0}^{N-1} coef_k b(t - k), %| and given sorted sample locations ti, synthesize signal samples f(t_i). %| %| in %| coef [N (L)] ...
github
alcu/sms-master
strum_test.m
.m
sms-master/irt/utilities/strum_test.m
2,235
utf_8
c462d966b83de02e9b682fa0ead74f11
function strum_test %|function strum_test %| test strum object %| Copyright 2006-1-19, Jeff Fessler, University of Michigan dat.a = 'x'; dat.b = 'y'; imp10 = @(st) st.b; imp11 = @(st, arg) [st.a arg]; imp12 = @(st, arg1, arg2) sum(arg1) + sum(arg2); list = { ... 'fun10', @strum_test_fun10, '()'; ... 'fun11', @st...
github
alcu/sms-master
highrate_centers.m
.m
sms-master/irt/utilities/highrate_centers.m
1,963
utf_8
97ce42064c1088889bda22226347fa87
function centers = highrate_centers(data, L, M) %|function centers = highrate_centers(data, L, M) %| According to high-rate scalar quantization theory (Gersho, T-IT, 1979), %| the density of MMSE quantization cell centroids should be f^{k/(k+2)}, %| for k-dimensional data distributed with pdf f. %| This m-file design...
github
alcu/sms-master
padn.m
.m
sms-master/irt/utilities/padn.m
1,368
utf_8
0e6e93516aba2748ee63b838b85f26da
function out = padn(mat, newdim) %|function out = padn(mat, newdim) %| pad input matrix to newdim, preserving 'center point' %| using this avoids matlab's padarray.m which is in image proc toolbox if nargin < 1, help(mfilename), error(mfilename), end if streq(mat, 'test'), padn_test, return, end % default: round to...
github
alcu/sms-master
reale.m
.m
sms-master/irt/utilities/reale.m
1,974
utf_8
f3caf767d930e1c501d34ea2da29e880
function y = reale(x, arg2, arg3) %| %| y = reale(x) %| y = reale(x, tol) %| y = reale(x, 'warn', 'message') %| y = reale(x, 'error') %| y = reale(x, 'report') %| y = reale(x, 'prompt') %| y = reale(x, 'disp') %| %| return real part of complex data (with error checking). %| checks that imaginary part is negligible (or ...
github
alcu/sms-master
equivs.m
.m
sms-master/irt/utilities/equivs.m
2,276
utf_8
7558c811be366947ee50f7e6155b9ea8
function out = equivs(var1, var2, varargin) %|function out = equivs(var1, var2, command) %| verify that var1 and var2 are equivalent to within single precision accuracy %| if not, print error message. an alternative to isequal(). %| See also: jf_equal %| option %| 'thresh' threshold (default: 1e-6) %| 'format' forma...
github
alcu/sms-master
ir_dwt_filters.m
.m
sms-master/irt/utilities/ir_dwt_filters.m
1,879
utf_8
796732b13b21d924af0060dda4d70904
function [lo hi] = ir_dwt_filters(wname, varargin) %|function [coef codes] = ir_odwt1(x, varargin) %| %| filters for discrete wavelet transform DWT %| %| in %| 'wname' char default: 'haar' %| %| option %| 'usemat' 0|1 default: 0. (if 1 then use matlab wfilters) %| 'ortho' 0|1 default: 1 (make normalized) %| 'abs' 0|1...
github
alcu/sms-master
jf_rm_toolbox.m
.m
sms-master/irt/utilities/private/jf_rm_toolbox.m
892
utf_8
81c39e54ce0f7b93c85f4faa5c889d2b
function jf_rm_toolbox % remove matlab toolboxes from path % for testing my toolbox with vanilla matlab rm aero rm aeroblks rm bioinfo rm comm rm commblks rm compiler rm control rm curvefit rm database rm des rm distcomp rm dspblks rm eml rm emlcoder rm filterdesign rm finance rm fixedpoint rm fuzzy rm geoweb rm ident...
github
alcu/sms-master
jf_prctile.m
.m
sms-master/irt/utilities/private/jf_prctile.m
1,066
utf_8
4f9221f4120fbff5aa6add943c5a1ac9
function y = jf_prctile(pn, x, p, dim) % mimic matlab's prctile which is in the "stats" toolbox if streq(x, 'test'), jf_prctile_test, y = []; return, end if nargin < 3, help(mfilename), error(mfilename), end % work with permuted dimensions if needed if nargin > 3 && ~isempty(dim) order = [dim:ndims(x) 1:dim-1]; x =...
github
alcu/sms-master
jf_color_order.m
.m
sms-master/irt/utilities/private/jf_color_order.m
984
utf_8
009a5177b58c46e4e5ed781f43bdf930
function out = jf_color_order(pn, varargin) %function out = jf_color_order(pn, varargin) % set color order to start with yellow then cyan etc., for black background if ~length(varargin) out = jf_color_order_bw; return end out = []; switch varargin{1} case 'revert' set(0, 'DefaultAxesColorOrder', 'default') retu...
github
alcu/sms-master
jf_struct_recurse.m
.m
sms-master/irt/utilities/private/jf_struct_recurse.m
1,824
utf_8
d7849c16991e5a735db1c20c9bfe3a79
function out = jf_struct_recurse(pn, st, varargin) %|function out = jf_struct_recurse(pn, st, varargin) %| %| Recursively descend a structure and examine its (numeric) members %| % Copyright 2010-04-05, Jeff Fessler, University of Michigan if nargin < 1, help(mfilename), error(mfilename), end arg.type = 'minmax_nan...
github
alcu/sms-master
display.m
.m
sms-master/irt/utilities/@strum/display.m
1,333
utf_8
6dc919bfa1a95d0a08d0e95ab1232ed0
function display(ob) %function display(ob) % "display" method for this class name = inputname(1); printm('"%s" is an object of class "%s":', name, class(ob)) ob = struct(ob); ir_display_struct(ob); if ~isempty(ob.meth) && ~isempty(ob.docs) printf('\n methods with comments:') display_method(ob.meth, ob.docs) end re...
github
alcu/sms-master
strum.m
.m
sms-master/irt/utilities/@strum/strum.m
3,362
utf_8
0cb6f8e9273c021dec2ada612358ef36
function ob = strum(data, methods, varargin) %|function ob = strum(data, methods, [options]) %| %| Construct a "strum" object. This acts like an ordinary matlab "struct" %| except that it also can contain user-defined "methods." %| For example, if x is a fieldname of the structure, then ob.x will return %| the corre...
github
alcu/sms-master
subsref_mm.m
.m
sms-master/irt/utilities/@strum/subsref_mm.m
3,299
utf_8
0759e96b1c0200abb8e0d65139da7015
function varargout = subsref(ob, args) %function varargout = subsref(ob, args) % handle subscript references like ob.ref or ob.ref(arg[s]) % Copyright 2006-1-19, Jeff Fessler, University of Michigan % 2011-10-25 modified by Madison McGaffin to support multiple outputs st = struct(ob); varargout = cell(nargout,1); if...
github
alcu/sms-master
mat_write.m
.m
sms-master/irt/utilities/arch/mat_write.m
2,632
utf_8
a893a9a496d886773d54df7231831e59
function mat_write(file, data, varargin) %function mat_write(file, data, % ['-check', '-nocheck', '-type', datatype, '-name', varname]) % % Save matlab "data" to "file", first checking if "file" exists! % If file suffix is .fld, then save as AVS .fld file in xdr_float. % % datatype can be 'xdr_float' 'xdr_int' ... % ...
github
alcu/sms-master
qpwls_precon.m
.m
sms-master/irt/wls/qpwls_precon.m
5,123
utf_8
908045fe64996174938f92f41d941ca6
function M = qpwls_precon(type, sys, C, mask, varargin) %|function M = qpwls_precon(type, sys, C, mask, varargin) %| usage: %| M = qpwls_precon('circ0', {T}, C, mask); %| M = qpwls_precon('circ0', {A, W}, C, mask); %| %| build a preconditioner for QPWLS problems; %| approximations to inv([A'WA + C'C]) %| %| in %| typ...
github
alcu/sms-master
pwls_art.m
.m
sms-master/irt/wls/pwls_art.m
3,522
utf_8
faea06efabbaae283fd280a75757ad99
From smalone@engin.umich.edu Wed Jul 15 23:51:01 1998 To: "Prof. Fessler" <fessler@eecs.umich.edu> Subject: MATLAB ART Algorithm Prof. Fessler, I have two ART algorithms for you to look at. Like the two PWLS+SOR algorithms, one MATLAB function has G'*W*G as a precomputed input argument and the other is without th...
github
alcu/sms-master
ir_hct_dgrad.m
.m
sms-master/irt/wls/ir_hct_dgrad.m
2,773
utf_8
6e959f8bea5db75de1d4f2cbc20bd86c
function [grad out com] = ir_hct_dgrad(x, A, yi, wi, varargin) %|function [grad out com] = ir_hct_dgrad(x, A, yi, wi, varargin) %| %| compute gradient of WLS data-fit term using hct binary %| for UM testing only %| %| in %| option %| (many - see below) %| %| Jeff Fessler, 2012-06-17 if nargin == 1 && streq(x, 'test'...
github
alcu/sms-master
qpwls_pcg2.m
.m
sms-master/irt/wls/qpwls_pcg2.m
4,102
utf_8
aa687f6d60490c2994b5796f74c175fd
function [xs, info] = qpwls_pcg2(x, T, bb, C, varargin) %|function [xs, info] = qpwls_pcg2(x, T, bb, C, [options]) %| %| quadratic penalized weighted least squares via %| preconditioned conjugate gradients (PCG) algorithm %| cost: Psi(x) = (y-Ax)'W(y-Ax)/2 + x'C'Cx/2 %| = -x'b + x'Tx/2 + x'C'Cx/2, where T = A'WA, b ...
github
alcu/sms-master
qpwls_qn.m
.m
sms-master/irt/wls/qpwls_qn.m
2,199
utf_8
950d9e08b6b6dfb0553136aa8429a095
function [xs, Ps] = qpwls_qn(x, G, W, yy, C, P, niter) %function [xs, Ps] = qpwls_qn(x, G, W, yy, C, P, niter) % % quadratic penalized weighted least squares (QPWLS) via % (preconditioned) quasi-Newton (QN) algorithm % cost(x) = (y-Gx)'W(y-Gx) / 2 + x'C'Cx / 2 % in % x [np,1] initial estimate % G [nn,np] system mat...
github
alcu/sms-master
qpwls_pcg1.m
.m
sms-master/irt/wls/qpwls_pcg1.m
6,494
utf_8
ec1a47614605fc4671c56339c97f5484
function [xs, info] = qpwls_pcg1(x, A, W, yi, C, varargin) %|function [xs, info] = qpwls_pcg1(x, A, W, yi, C, [options]) %| %| quadratic penalized weighted least squares (QPWLS) via %| preconditioned conjugate gradients (PCG) algorithm %| cost(x) = (y-Ax)'W(y-Ax) / 2 + x'C'Cx / 2 %| %| in %| x [np 1] initial estimat...
github
alcu/sms-master
pwls_sqs_os.m
.m
sms-master/irt/wls/pwls_sqs_os.m
4,191
utf_8
6b197dad846005c3379cede039c1d1a3
function [xs, info] = pwls_sqs_os(x, Ab, yi, R, varargin) %|function [xs, info] = pwls_sqs_os(x, Ab, yi, R, [options]) %| %| penalized weighted least squares estimation / image reconstruction %| using separable quadratic surrogates algorithm with %| (optionally relaxed) ordered subsets. (relaxation ensures convergen...
github
alcu/sms-master
qpwls_bb1.m
.m
sms-master/irt/wls/qpwls_bb1.m
5,870
utf_8
ec9a3eda5e8154f08cac635506779cfe
function [xs, info] = qpwls_bb1(x, A, W, yi, C, varargin) %|function [xs, info] = qpwls_bb1(x, A, W, yi, C, [options]) %| %| quadratic penalized weighted least squares (QPWLS) via %| preconditioned (todo) Barzilai and Borwein gradient method. %| cost(x) = (y-Ax)'W(y-Ax) / 2 + x'C'Cx / 2 %| %| in %| x [np 1] initial ...
github
alcu/sms-master
pwls_pcg1.m
.m
sms-master/irt/wls/pwls_pcg1.m
7,580
utf_8
a64997e6145be968fbbebe41db5f3d40
function [xs, info] = pwls_pcg1(x, A, W, yi, R, varargin) %|function [xs, info] = pwls_pcg1(x, A, W, yi, R, [options]) %| %| penalized weighted least squares (PWLS) %| with convex non-quadratic regularization, %| minimized via preconditioned conjugate gradient algorithm. %| cost(x) = (y-Ax)'W(y-Ax)/2 + R(x) %| %| in ...
github
alcu/sms-master
qpwls_psd.m
.m
sms-master/irt/wls/qpwls_psd.m
3,317
utf_8
d2983170e45b5a9604a3e1cb862e6b57
function [xs, info] = qpwls_psd(x, A, W, yi, C, varargin) %|function [xs, info] = qpwls_psd(x, A, W, yi, C, [options]) %| %| quadratic penalized weighted least squares (QPWLS) via %| preconditioned steepest descent (PSD) algorithm %| cost(x) = (y-Ax)'W(y-Ax) / 2 + x'C'Cx / 2 %| %| in %| x [np 1] initial estimate %| ...
github
alcu/sms-master
pwls_sor.m
.m
sms-master/irt/wls/pwls_sor.m
3,878
utf_8
edcdc3e411ad60eac6a1758e57fe4f86
From smalone@engin.umich.edu Wed Jul 15 13:37:52 1998 To: "Jeffrey A. Fessler" <fessler@eecs.umich.edu> Subject: Re: pwls_sor Prof. Fessler, I'm sending along two of my pwls_sor routines. Both compute one iteration of the routine. One uses precomputed diag[G'*W*G]. The other does not. I could combine the two, if y...
github
alcu/sms-master
qpwls_pcg1.m
.m
sms-master/irt/wls/arch/qpwls_pcg1.m
3,314
utf_8
fd3f754f50a8b368040d6e09fdf1914c
function [xs, info] = qpwls_pcg1(x, G, W, yi, C, varargin) %function [xs, info] = qpwls_pcg1(x, G, W, yi, C, [options]) % % quadratic penalized weighted least squares via % preconditioned conjugate gradients (PCG) algorithm % cost(x) = (y-Gx)'W(y-Gx)/2 + x'C'Cx/2 % in % x [np,1] initial estimate % G [nd,np] system m...
github
alcu/sms-master
pwls_sps.m
.m
sms-master/irt/wls/arch/pwls_sps.m
2,077
utf_8
5d799707ddebdbd6ef6a44aa3df5f359
function xs = pwls_sps(x, Gt, wi, yi, P, niter, pixmax, ldenom, gi) %function xs = pwls_sps(x, Gt, wi, yi, P, niter, pixmax, ldenom, gi) % % penalized weighted least squares image reconstruction % using separable paraboloidal surrogates algorithm % cost function: J(x) = (y-Gx) W (y-Gx) / 2 + x' C' C x / 2 % Inputs req...
github
alcu/sms-master
strvcat.m
.m
sms-master/irt/freemat/strvcat.m
272
utf_8
90fa0731122679f089828dfb7fbde41a
%| function out = strvcat(varargin) %| %| (for freemat) function out = strvcat(varargin) % for compatability with matlab, remove empty arguments tmp = {}; for ii=1:length(varargin) if ~isempty(varargin{ii}) tmp = {tmp{:}, varargin{ii}}; end end out = char(tmp{:});
github
alcu/sms-master
isequal.m
.m
sms-master/irt/freemat/isequal.m
915
utf_8
1ba0cf9dd50bfe69854faa7414ad7595
function out = isequal(a,b) %function out = isequal(a,b) % needed for freemat 4.0 if nargin < 2, help(mfilename), error(mfilename), end out = false; % check that class type matches if ~isnumeric(a) || ~isnumeric(b) if ~streq(class(a), class(b)), return, end end if ischar(a) out = streq(a, b); return end if isce...
github
alcu/sms-master
strmatch.m
.m
sms-master/irt/freemat/pre-4.0/strmatch.m
1,059
utf_8
0347fe3860ac2aebcfb853e5e41ad8bd
function out = strmatch(str, strs, type) %function out = strmatch(str, strs, type) % return matching row indices if nargin == 1 && streq(str, 'test'), strmatch_test, return, end if nargin < 3, help(mfilename), error(mfilename), end if ~streq(type, 'exact'), error('only "exact" done'), end if iscell(strs) nstr = len...
github
alcu/sms-master
subsasgn.m
.m
sms-master/irt/freemat/pre-3.6/subsasgn.m
787
utf_8
5f5fd80e8d836587f9c46393402aaae1
function out = subsasgn(a, s, b) %function out = subsasgn(a, s, b) if nargin == 1 && streq(a, 'test'), subsasgn_test, return, end if nargin < 3, help(mfilename), error(mfilename), end out = a; % out = builtin('subsasgn', varargin{:}); % seems to recurse % a.s = b; if length(s) == 1 && streq(s(1).type, '.') out.(s(...
github
alcu/sms-master
penalty_mex.m
.m
sms-master/irt/freemat/pre-3.5/penalty_mex.m
5,279
utf_8
4c050f4d3802c7d66298ca64a07c6839
function out = penalty_mex(arg, varargin) %function out = penalty_mex(arg, varargin) % m-file interface to penalty_mex.mex* % Usage for penalty_mex: % d = function('diff1,forw1', x, offsets, [lastdim]); % d = C * x % offsets is int32 array such as [1 nx nx-1 nx+1]. % Usually lastdim is absent, and x is treated as s...
github
alcu/sms-master
myiswt2.m
.m
sms-master/irt/penalty/myiswt2.m
2,208
utf_8
f29f4f2652fb13bafdd45302978ef924
function a = myiswt2(swc, lo_r, hi_r) % ISWT2 Inverse discrete stationary wavelet transform 2-D. % ISWT2 performs a multilevel 2-D stationary wavelet % reconstruction using either a specific orthogonal wavelet % ('wname', see WFILTERS for more information) or specific % reconstruction filters (Lo_r and Hi_r). % ...
github
alcu/sms-master
Reg1.m
.m
sms-master/irt/penalty/Reg1.m
23,615
utf_8
75da01050a9ea5d61d2d880d6d7ae318
function R = Reg1(kappa, varargin) %|function R = Reg1(kappa, [options]) %| %| Build roughness penalty regularization "object" based on Cdiff1() objects, %| for regularized solutions to inverse problems. %| This version supercedes Robject() by providing its capabilities while %| also providing options that use less m...
github
alcu/sms-master
penalty_displace.m
.m
sms-master/irt/penalty/penalty_displace.m
3,585
utf_8
daae8a09db6bcc8839e0e437f74deb52
function displace = penalty_displace(offsets, sizes) %|function displace = penalty_displace(offsets, sizes) %| %| Convert scalar offsets to vector displacements, i.e., %| find d1,d2,... such that d1*1 + d2*n1 + d3*n1*n2 + ... = offset. %| Example: if offset = n1, then [d1 d2 d3] = [0 1 0] %| Assumes that d <= floor(n...
github
alcu/sms-master
Cdiff1_tune.m
.m
sms-master/irt/penalty/Cdiff1_tune.m
3,013
utf_8
3ecb396069b71348797f5bbf64a0bf07
function type_diff = Cdiff1_tune(isize, varargin) %function type_diff = Cdiff1_tune(isize, [options]) %| %| "Auto-tune" function for Cdiff1() that selects the fastest overall method %| %| in %| isize [] vector of object dimensions (N), e.g., [64 64] %| %| option: %| 'offset' [int] offset (default: 1) %| 'order' 1 or...
github
alcu/sms-master
huber_pot.m
.m
sms-master/irt/penalty/huber_pot.m
383
utf_8
7074ef88c04edbf2d8a674143a224881
function h = huber_pot(t, d) %| function h = huber_pot(t, d) %| huber potential function if nargin < 1, help(mfilename), error(mfilename), end if nargin < 2, huber_pot_test, return, end h = t.^2 / 2; ii = abs(t) > d; h(ii) = d * abs(t(ii)) - d.^2/2; function huber_pot_test t = linspace(-9,9,101); delta = 4; plot(t, ...
github
alcu/sms-master
ir_hct_rgrad.m
.m
sms-master/irt/penalty/ir_hct_rgrad.m
4,672
utf_8
328fac6e04b83f7ce893fd9b056ef097
function [grad out com] = ir_hct_rgrad(x, varargin) %|function [grad out com] = ir_hct_rgrad(x, varargin) %| %| compute 3D regularizer gradient using hct binary %| for UM testing only %| %| in %| x in xyz order %| option %| (many - see below) %| %| Notes: because hct2 uses 'zxy' order, it is essential to use Reg1 %| ...
github
alcu/sms-master
qpwls_psf.m
.m
sms-master/irt/penalty/qpwls_psf.m
13,114
utf_8
e610d68a9b49f03e6281558a0c2d46f4
function [psf, var, fwhm, mtf, con] = qpwls_psf(A, C, beta, mask, W, varargin) %|function [psf, var, fwhm, mtf, con] = qpwls_psf(A, C, beta, mask, W, [option]) %| %| Compute FFT-approximate PSF %| for quadratically-penalized shift-invariant unweighted least-squares %| uses psf = [A'WA + beta C'C]^{-1} A'W y, where us...
github
alcu/sms-master
C2sparse.m
.m
sms-master/irt/penalty/C2sparse.m
4,020
utf_8
bccb4c431c640ab2b5479d543ee2188c
function [C, wjk] = C2sparse(type, kappa, nbrs, chat, dist_power) %function [C, wjk] = C2sparse(type, kappa, nbrs, chat, dist_power) % % create 2D penalty matrix C with 2nd-order pixel neighborhood % that performs 1st-order finite differences. % R = beta * C'*D(wjk)*C is the penalty Hessian (in quadratic case) % in: %...
github
alcu/sms-master
potential_shift.m
.m
sms-master/irt/penalty/potential_shift.m
1,363
utf_8
99ba645443a4bfba020e125120157dd1
function pot_new = potential_shift(pot_orig, vector) %function pot_new = potential_shift(pot_orig, vector) % % Change pot_orig([Cx]_k) to shifted form: pot_new( [Cx]_k - vector_k ) % % out % inline functions: % pot.potk(pot, C*x) potential function value % pot.wpot(pot, C*x) potential 'weights' (aka half-quad. curvatu...
github
alcu/sms-master
reg_offset_xyz_to_zxy.m
.m
sms-master/irt/penalty/reg_offset_xyz_to_zxy.m
1,657
utf_8
9596eb6b6022420c7d50c90ab4e21aee
function offset_zxy = reg_offset_xyz_to_zxy(offset_xyz, dim_xyz) %|function offset_zxy = reg_offset_xyz_to_zxy(offset_xyz, dim_xyz) %| %| convert usual offsets for xyz image ordering %| to offsets appropriate for zxy image ordering %| %| in %| offset_xyz [1 noffset] %| dim_xyz [3] %| %| out %| offset_zxy [1 noffset]...
github
alcu/sms-master
Robject.m
.m
sms-master/irt/penalty/Robject.m
8,093
utf_8
b296c5415462af0d48b30953df27f765
function R = Robject(kappa, varargin) %function R = Robject(kappa, [options]) % % Build roughness penalty regularization "object" based on C = Cdiff() object, % for regularized solutions to inverse problems. % % General form of nonquadratic penalty function: % R(x) = \sumk w_k \pot([Cx]_k), where [Cx]_k = \sum_j c_{kj...
github
alcu/sms-master
Cdiff1.m
.m
sms-master/irt/penalty/Cdiff1.m
19,004
utf_8
e85e378261250a5e825e21d69604e79b
function ob = Cdiff1(isize, varargin) %|function C1 = Cdiff1(isize, [options]) %| %| Construct Cdiff1 object that can compute C1 * x and the adjoint C1' * d %| for a "finite differencing" matrix C1 for roughness penalty regularization. %| This object provides the finite difference in only a *single* direction, %| so ...
github
alcu/sms-master
Csparse.m
.m
sms-master/irt/penalty/Csparse.m
5,746
utf_8
d6e181cab286d0f68dcd0c0bcc173df6
function [C, ugibb] = Csparse(arg0, arg1, arg2, arg3) %function C = Csparse('b2info', kappa, mask, chat) %function C = Csparse('center', Cscale, mask, chat) %function C = Csparse('maskleak', mask, chat) %function C = Csparse('maskleak3', [beta_xy beta_z], mask, chat) %function C = Csparse('maskbar', mask, chat) %funct...