markdown stringlengths 0 1.02M | code stringlengths 0 832k | output stringlengths 0 1.02M | license stringlengths 3 36 | path stringlengths 6 265 | repo_name stringlengths 6 127 |
|---|---|---|---|---|---|
Convert from Radians to Degrees The above answers will be in radians, use the following code to convert to degrees. | #substitute x_c and x_i as needed
z=c.phase(x_c)
m.degrees(z)
print("Angle in degrees =",m.degrees(z)) | Angle in degrees = -81.07294513104007
| MIT | Applications/ENGR 202 Solver.ipynb | smithrockmaker/ENGR213 |
Simple Circuit in Series | # For following three cells, if reactance is already given, replace "xind" or"xcap" with corresponding j value
# Resistor value is overwritten from previous cells when changed here
# Not all simple circuits will have all three components. Modify as needed.
# Original formula - series_comb = r + ind + cap
r = 100 ... | Series Rectangular Form = (100+6919.805079547168j)
| MIT | Applications/ENGR 202 Solver.ipynb | smithrockmaker/ENGR213 |
Simple Parallel Circuit - Product/Sum | # Product sum rule works only with 2 components
# Original Formula - prod_sum = res*cap/(res + cap)
ind = 0 + xind*1j
cap = 0 + xcap*1j
res = 100
prod_sum = res*cap/(res + cap)
print("Product/sum Rectangular Form =",prod_sum) | Product/sum Rectangular Form = (97.59201358307332-15.329717646080926j)
| MIT | Applications/ENGR 202 Solver.ipynb | smithrockmaker/ENGR213 |
Simple Parallel Circuit | # Use as many components as necessary
# Original formula - parallel_comb = 1/(1/res + 1/ind + 1/cap)
ind = 0 + xind*1j
cap = 0 + xcap*1j
res = 100
parallel_comb = 1/(1/res + 1/ind + 1/cap)
print("Parallel Rectangular Form =",parallel_comb) | Parallel Rectangular Form = (98.04620253923555-13.840607701931122j)
| MIT | Applications/ENGR 202 Solver.ipynb | smithrockmaker/ENGR213 |
Current Solver | # Make sure to use the parallel cell that IS NOT product/sum
# Copy and paste cur_ind or cur_cap sections as necessary to account for all components. Some code modifaction/addition may be required.
# This cell useful as is for one of each component.
# Once previous cells are complete, this will populate automatically E... | Z Polar = (99.01828242260899, -0.14023751838586943)
Z Rectangular = (98.04620253923555-13.840607701931122j)
Source Current = (0.1+0.014116413837030016j)
Source Current, Polar = (0.1009914508244054, 0.14023751838586943)
Angle = 8.035017932898603
Capacitor Current = (-0+0.015707963267948967j)
Capacitor Cur... | MIT | Applications/ENGR 202 Solver.ipynb | smithrockmaker/ENGR213 |
Series-Parallel Circuits | # Organization cell for component values
# Inductors
z1 = 200*1j
# Resistors
z2 = 300
z3 = 270
#Capacitors
z4 = -1500*1j
# This cell is ambiguous with just z values to make it easy to modify. Keep track of z values.
# Original Form of equation - parallel_react = 1/(1/z1+1/z2+1/(z3+z4))
parallel_react = 1/(1/z1+1/z... | Z Rectangular = (111.7846057266418+141.10138457782585j)
Z Polar = (180.01499606210666, 0.9008118071374078)
| MIT | Applications/ENGR 202 Solver.ipynb | smithrockmaker/ENGR213 |
TPU | AUTO = tf.data.experimental.AUTOTUNE
# Detect hardware, return appropriate distribution strategy
try:
tpu = tf.distribute.cluster_resolver.TPUClusterResolver() # TPU detection. No parameters necessary if TPU_NAME environment variable is set. On Kaggle this is always the case.
print('Running on TPU ', tpu.maste... | _____no_output_____ | MIT | Image Classification/CGIAR Computer Vision for Crop Disease/recongratulationsyoure3iniclrworkshopchallenge1/model4.ipynb | ZindiAfrica/Computer-Vision |
This notebook was prepared by [Donne Martin](https://github.com/donnemartin). Source and license info is on [GitHub](https://github.com/donnemartin/interactive-coding-challenges). Solution Notebook Problem: Find the highest product of three numbers in a list.* [Constraints](Constraints)* [Test Cases](Test-Cases)* [Al... | class Solution(object):
def max_prod_three_nlogn(self, array):
if array is None:
raise TypeError('array cannot be None')
if len(array) < 3:
raise ValueError('array must have 3 or more ints')
array.sort()
product = 1
for item in array[-3:]:
... | _____no_output_____ | Apache-2.0 | online_judges/prod_three/prod_three_solution.ipynb | sophomore99/PythonInterective |
Unit Test | %%writefile test_prod_three.py
from nose.tools import assert_equal, assert_raises
class TestProdThree(object):
def test_prod_three(self):
solution = Solution()
assert_raises(TypeError, solution.max_prod_three, None)
assert_raises(ValueError, solution.max_prod_three, [1, 2])
assert... | Success: test_prod_three
| Apache-2.0 | online_judges/prod_three/prod_three_solution.ipynb | sophomore99/PythonInterective |
Unzipping and Zipping FilesAs you are probably aware, files can be compressed to a zip format. Often people use special programs on their computer to unzip these files, luckily for us, Python can do the same task with just a few simple lines of code. Create Files to Compress | # slashes may need to change for MacOS or Linux
f = open("new_file.txt",'w+')
f.write("Here is some text")
f.close()
# slashes may need to change for MacOS or Linux
f = open("new_file2.txt",'w+')
f.write("Here is some text")
f.close() | _____no_output_____ | MIT | 4-assets/BOOKS/Jupyter-Notebooks/06-Unzipping-and-Zipping-Files-checkpoint.ipynb | impastasyndrome/Lambda-Resource-Static-Assets |
Zipping FilesThe [zipfile library](https://docs.python.org/3/library/zipfile.html) is built in to Python, we can use it to compress folders or files. To compress all files in a folder, just use the os.walk() method to iterate this process for all the files in a directory. | import zipfile | _____no_output_____ | MIT | 4-assets/BOOKS/Jupyter-Notebooks/06-Unzipping-and-Zipping-Files-checkpoint.ipynb | impastasyndrome/Lambda-Resource-Static-Assets |
Create Zip file first , then write to it (the write step compresses the files.) | comp_file = zipfile.ZipFile('comp_file.zip','w')
comp_file.write("new_file.txt",compress_type=zipfile.ZIP_DEFLATED)
comp_file.write('new_file2.txt',compress_type=zipfile.ZIP_DEFLATED)
comp_file.close() | _____no_output_____ | MIT | 4-assets/BOOKS/Jupyter-Notebooks/06-Unzipping-and-Zipping-Files-checkpoint.ipynb | impastasyndrome/Lambda-Resource-Static-Assets |
Extracting from Zip FilesWe can easily extract files with either the extractall() method to get all the files, or just using the extract() method to only grab individual files. | zip_obj = zipfile.ZipFile('comp_file.zip','r')
zip_obj.extractall("extracted_content") | _____no_output_____ | MIT | 4-assets/BOOKS/Jupyter-Notebooks/06-Unzipping-and-Zipping-Files-checkpoint.ipynb | impastasyndrome/Lambda-Resource-Static-Assets |
________ Using shutil libraryOften you don't want to extract or archive individual files from a .zip, but instead archive everything at once. The shutil library that is built in to python has easy to use commands for this: | import shutil | _____no_output_____ | MIT | 4-assets/BOOKS/Jupyter-Notebooks/06-Unzipping-and-Zipping-Files-checkpoint.ipynb | impastasyndrome/Lambda-Resource-Static-Assets |
The shutil library can accept a format parameter, `format` is the archive format: one of "zip", "tar", "gztar", "bztar",or "xztar". | pwd
directory_to_zip='C:\\Users\\Marcial\\Pierian-Data-Courses\\Complete-Python-3-Bootcamp\\12-Advanced Python Modules'
# Creating a zip archive
output_filename = 'example'
# Just fill in the output_filename and the directory to zip
# Note this won't run as is because the variable are undefined
shutil.make_archive(outp... | _____no_output_____ | MIT | 4-assets/BOOKS/Jupyter-Notebooks/06-Unzipping-and-Zipping-Files-checkpoint.ipynb | impastasyndrome/Lambda-Resource-Static-Assets |
Causal Inference In Statistics - A Primer 3.1 Interventions Bruno Gonçalves www.data4sci.com @bgoncalves, @data4sci | from collections import Counter
from pprint import pprint
import pandas as pd
import numpy as np
import matplotlib
import matplotlib.pyplot as plt
from CausalModel import CausalModel
import watermark
%load_ext watermark
%matplotlib inline | _____no_output_____ | MIT | 3.1 - Interventions.ipynb | m5l14i11/Causality |
We start by print out the versions of the libraries we're using for future reference | %watermark -n -v -m -g -iv | watermark 2.0.2
json 2.0.9
pandas 1.0.1
matplotlib 3.1.3
numpy 1.18.1
autopep8 1.5
Sun Oct 18 2020
CPython 3.7.3
IPython 6.2.1
compiler : Clang 4.0.1 (tags/RELEASE_401/final)
system : Darwin
release : 19.6.0
machine : x86_64
processor : i386
CPU cores : 8
interpreter: 64bit
Git hash ... | MIT | 3.1 - Interventions.ipynb | m5l14i11/Causality |
Load default figure style | plt.style.use('./d4sci.mplstyle')
colors = plt.rcParams['axes.prop_cycle'].by_key()['color'] | _____no_output_____ | MIT | 3.1 - Interventions.ipynb | m5l14i11/Causality |
Graph Surgery | G = CausalModel()
G.add_causation('Ux', 'X')
G.add_causation('Uy', 'Y')
G.add_causation('Uz', 'Z')
G.add_causation('Z', 'X')
G.add_causation('Z', 'Y')
G.pos = {'Z': (0, 1), 'X': (-1, 0), 'Y':(1, 0), 'Uz':(0, 2), 'Ux':(-1, 1), 'Uy': (1,1)}
fig, ax = plt.subplots(1, figsize=(3, 2.5))
G.plot(ax=ax)
G.save_model('dags/Pri... | _____no_output_____ | MIT | 3.1 - Interventions.ipynb | m5l14i11/Causality |
Ex - GroupBy Introduction:GroupBy can be summarized as Split-Apply-Combine.Special thanks to: https://github.com/justmarkham for sharing the dataset and materials.Check out this [Diagram](http://i.imgur.com/yjNkiwL.png) Step 1. Import the necessary libraries | import pandas as pd | _____no_output_____ | BSD-3-Clause | 03_Grouping/Alcohol_Consumption/Exercise.ipynb | coderhh/pandas_exercises |
Step 2. Import the dataset from this [address](https://raw.githubusercontent.com/justmarkham/DAT8/master/data/drinks.csv). Step 3. Assign it to a variable called drinks. | url = 'https://raw.githubusercontent.com/justmarkham/DAT8/master/data/drinks.csv'
drinks = pd.read_csv(url)
drinks | _____no_output_____ | BSD-3-Clause | 03_Grouping/Alcohol_Consumption/Exercise.ipynb | coderhh/pandas_exercises |
Step 4. Which continent drinks more beer on average? | drinks.groupby('continent').beer_servings.mean() | _____no_output_____ | BSD-3-Clause | 03_Grouping/Alcohol_Consumption/Exercise.ipynb | coderhh/pandas_exercises |
Step 5. For each continent print the statistics for wine consumption. | drinks.groupby('continent').wine_servings.describe() | _____no_output_____ | BSD-3-Clause | 03_Grouping/Alcohol_Consumption/Exercise.ipynb | coderhh/pandas_exercises |
Step 6. Print the mean alcohol consumption per continent for every column | drinks.groupby('continent').mean() | _____no_output_____ | BSD-3-Clause | 03_Grouping/Alcohol_Consumption/Exercise.ipynb | coderhh/pandas_exercises |
Step 7. Print the median alcohol consumption per continent for every column | drinks.groupby('continent').median() | _____no_output_____ | BSD-3-Clause | 03_Grouping/Alcohol_Consumption/Exercise.ipynb | coderhh/pandas_exercises |
Step 8. Print the mean, min and max values for spirit consumption. This time output a DataFrame | drinks.groupby('continent').spirit_servings.agg(['mean', 'min', 'max']) | _____no_output_____ | BSD-3-Clause | 03_Grouping/Alcohol_Consumption/Exercise.ipynb | coderhh/pandas_exercises |
Week 3, Day 1 (Dataset Preparation and Arrangement)> Welcome to first day (Week 3) of the McE-51069 course.- sticky_rank: 7- toc: true- badges: false- comments: false- categories: [deep_learning, computer_vision] You can download resources for today from this [link](https://github.com/ytu-cvlab/mce-51069-week3-day1/ar... | import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import cv2
import math
%matplotlib inline | _____no_output_____ | Apache-2.0 | _notebooks/2020-12-23-week3-day1.ipynb | ytu-cvlab/Resource-Blog |
Pandas stores data in dataframe objects. We can assign columns to each to numpy array (or) list to create a dataframe. | #Create a dataframe
names = ['Jack','Jean','Jennifer','Jimmy']
ages = np.array([23,22,24,21])
# print(type(names))
# print(type(ages))
df = pd.DataFrame({'name': names,
'age': ages,
'city': ['London', 'Berlin', 'New York', 'Sydney']},index=None)
df.head()
# df.style.hide_index(... | _____no_output_____ | Apache-2.0 | _notebooks/2020-12-23-week3-day1.ipynb | ytu-cvlab/Resource-Blog |
Now, let's see some handy dataframe tricks. | df[['name','city']]
df.info()
# print(df.columns)
# print(df.age) | <class 'pandas.core.frame.DataFrame'>
RangeIndex: 4 entries, 0 to 3
Data columns (total 3 columns):
# Column Non-Null Count Dtype
--- ------ -------------- -----
0 name 4 non-null object
1 age 4 non-null int32
2 city 4 non-null object
dtypes: int32(1), object(2)
memory usage... | Apache-2.0 | _notebooks/2020-12-23-week3-day1.ipynb | ytu-cvlab/Resource-Blog |
Now that we know how to create a dataframe, we can save the dataframe we created. | df.to_csv('Ages_and_cities.csv',index=False,header=True)
df = pd.read_csv('Ages_and_cities.csv')
df.head() | _____no_output_____ | Apache-2.0 | _notebooks/2020-12-23-week3-day1.ipynb | ytu-cvlab/Resource-Blog |
Understanding your dataset In this section, we used [Iris flowers dataset](https://en.wikipedia.org/wiki/Iris_flower_data_set), which contains petal and sepal measurements of three species of Iris flowers. Three species of Iris flowers from the dataset  Sepal vs Petal This dataset was introduced ... | df = pd.read_csv('iris_data.csv')
df.head()
# df.head(3)
df.tail() | _____no_output_____ | Apache-2.0 | _notebooks/2020-12-23-week3-day1.ipynb | ytu-cvlab/Resource-Blog |
Slicing data Now that we understand our dataset, let's prepare to seperate our data based on labels for unique visualization. | # df.loc[:3]
df.loc[80:85,("sepal_length","variety")]
# df.iloc[146:]
# df.iloc[80:85,2:5]
df.iloc[80:85,[0,4]]
Se= df.loc[df.variety =='Setosa', :]
Vc= df.loc[df.variety =='Versicolor', :]
Vi= df.loc[df.variety =='Virginica', :]
Vi.head() | _____no_output_____ | Apache-2.0 | _notebooks/2020-12-23-week3-day1.ipynb | ytu-cvlab/Resource-Blog |
Feature visualization | df = pd.read_csv('iris_data.csv')
# df.dtypes | _____no_output_____ | Apache-2.0 | _notebooks/2020-12-23-week3-day1.ipynb | ytu-cvlab/Resource-Blog |
First, we will visualize each measurement with histograms to observe the output distribution for each class. | # df.hist("sepal.length",bins=15,edgecolor='black')
plt.figure(figsize=(15,15))
plt.subplot(2, 2, 1)
plt.hist(Se.sepal_length,bins=15,color="steelblue",edgecolor='black',alpha =0.4, label="Setosa")
plt.hist(Vc.sepal_length,bins=15,color='red',edgecolor='black', alpha =0.3, label="Versicolor")
plt.hist(Vi.sepal_length... | _____no_output_____ | Apache-2.0 | _notebooks/2020-12-23-week3-day1.ipynb | ytu-cvlab/Resource-Blog |
Now, we will visualize multiple features with scatter plots to gain some more insights. | plt.figure(figsize=(15,15))
area = np.pi*20
plt.subplot(2, 2, 1)
plt.scatter(Se.sepal_length,Se.sepal_width, s=area, c="steelblue", alpha=0.6, label="Setosa")
plt.scatter(Vc.sepal_length,Vc.sepal_width, s=area, c="red", alpha=0.6, label="Versicolor")
plt.scatter(Vi.sepal_length,Vi.sepal_width, s=area, c="blue", alpha... | _____no_output_____ | Apache-2.0 | _notebooks/2020-12-23-week3-day1.ipynb | ytu-cvlab/Resource-Blog |
We can definitely see some blobs forming from these visualizations. "Setosa" class unsally stands out from the other two classes but the sepal width vs sepal length plot shows "versicolor" and "virginica" classes will more challenging to classify compared to "setosa" class. Training the model [Scikit-learn](https://sc... | from sklearn.model_selection import train_test_split
from sklearn.tree import DecisionTreeClassifier
from sklearn.ensemble import RandomForestClassifier
from sklearn import metrics
import seaborn as sns
df = pd.read_csv('iris_data.csv')
# df.dtypes
df.tail()
train_X, test_X, train_y, test_y = train_test_split(df[df.col... | _____no_output_____ | Apache-2.0 | _notebooks/2020-12-23-week3-day1.ipynb | ytu-cvlab/Resource-Blog |
Model Evaluation Decision Tree classifier | print(metrics.classification_report(DT_predicted, test_y))
mat = metrics.confusion_matrix(test_y, DT_predicted)
sns.heatmap(mat.T, square=True, annot=True, fmt='d', cbar=False)
plt.xlabel('true label')
plt.ylabel('predicted label'); | _____no_output_____ | Apache-2.0 | _notebooks/2020-12-23-week3-day1.ipynb | ytu-cvlab/Resource-Blog |
Ramdom Forest Classifier | print(metrics.classification_report(RF_predicted, test_y))
from sklearn.metrics import confusion_matrix
import seaborn as sns
mat = confusion_matrix(test_y, RF_predicted)
sns.heatmap(mat.T, square=True, annot=True,fmt='d', cbar=False)
plt.xlabel('true label')
plt.ylabel('predicted label'); | _____no_output_____ | Apache-2.0 | _notebooks/2020-12-23-week3-day1.ipynb | ytu-cvlab/Resource-Blog |
[colab notebook](https://colab.research.google.com/github/jakevdp/PythonDataScienceHandbook/blob/master/notebooks/05.08-Random-Forests.ipynb) Feature Engineering When generating new features, the product between two features is usually not recommended to engineer unless it makes a magnification of the situation. Here,... | #Generate new features
df = pd.read_csv('iris_data.csv')
df['petal_hypotenuse'] = np.sqrt(df["petal_length"]**2+df["petal_width"]**2)
df['petal_product']=df["petal_length"]*df["petal_width"]
df.tail()
Se= df.loc[df.variety =='Setosa', :]
Vc= df.loc[df.variety =='Versicolor', :]
Vi= df.loc[df.variety =='Virginica', :]
... | _____no_output_____ | Apache-2.0 | _notebooks/2020-12-23-week3-day1.ipynb | ytu-cvlab/Resource-Blog |
Train with Engineered features Now, let's replace two petal features with two new features we generated. | df.head()
df2 = df.loc[:,["sepal_length","sepal_width","petal_hypotenuse","petal_product","variety"]]
df2.dtypes
train_X, test_X, train_y, test_y = train_test_split(df2[df2.columns[0:4]].values,
df2.variety.values, test_size=0.25)
from sklearn.tree import DecisionTre... | _____no_output_____ | Apache-2.0 | _notebooks/2020-12-23-week3-day1.ipynb | ytu-cvlab/Resource-Blog |
Data PreprocessingThis notebook shows | import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import re
import json
import nltk
from nltk.corpus import wordnet
import sklearn
import seaborn as sns
import unicodedata
import inflect
nltk.download('punkt')
nltk.download('wordnet')
nltk.download('averaged_perceptron_tagger')
nltk.download('stop... | [nltk_data] Downloading package punkt to /home/jovyan/nltk_data...
[nltk_data] Package punkt is already up-to-date!
[nltk_data] Downloading package wordnet to /home/jovyan/nltk_data...
[nltk_data] Package wordnet is already up-to-date!
[nltk_data] Downloading package averaged_perceptron_tagger to
[nltk_data] /h... | MIT | notebooks/data_preprocessing.ipynb | SudoHead/movie-classifier |
Load the data: | path = '../data/movies_metadata.csv'
df = pd.read_csv(path)
df = pd.concat([df['release_date'], df['title'], df['overview'], df['genres']], axis=1)
# remove duplicates
duplicate_rows = df[df.duplicated()]
df.drop(duplicate_rows.index, inplace=True) | _____no_output_____ | MIT | notebooks/data_preprocessing.ipynb | SudoHead/movie-classifier |
Drop the NaN rows where either title or overview is NaN | # convert empty string to NaN
df['overview'].replace('', np.nan, inplace=True)
df.dropna(subset=['release_date', 'title', 'overview'], inplace=True)
# the release date is no longer necessary, because NaN are cleared
del df['release_date'] | _____no_output_____ | MIT | notebooks/data_preprocessing.ipynb | SudoHead/movie-classifier |
Drop rows with no overview info or blank | reg_404 = "^not available|^no overview"
overview_not_found = df['overview'].str.contains(reg_404, regex=True, flags=re.IGNORECASE)
overview_blank = df['overview'].str.isspace()
df.drop(df[overview_not_found].index, inplace=True)
df.drop(df[overview_blank].index, inplace=True)
df.head() | /opt/conda/lib/python3.7/site-packages/ipykernel_launcher.py:6: UserWarning: Boolean Series key will be reindexed to match DataFrame index.
| MIT | notebooks/data_preprocessing.ipynb | SudoHead/movie-classifier |
Transform column genre | def extract_genres(genres_str):
genres_str = genres_str.replace("'", '\"')
genres_json = json.loads(genres_str)
genres_list = []
for elem in genres_json:
genres_list.append(elem['name'])
return genres_list
# remove rows with no genres, since they don't provide any information
df.drop(df[df['... | _____no_output_____ | MIT | notebooks/data_preprocessing.ipynb | SudoHead/movie-classifier |
Visualise movie genres' distribution | all_genres = sum(df['genres'], [])
genre_types = set(all_genres)
len(genre_types)
all_genres = nltk.FreqDist(all_genres)
# create dataframe
all_genres_df = pd.DataFrame({'Genre': list(all_genres.keys()),
'Count': list(all_genres.values())})
g = all_genres_df.nlargest(columns="Count", n... | _____no_output_____ | MIT | notebooks/data_preprocessing.ipynb | SudoHead/movie-classifier |
Text Preprocessing | def to_lower(text):
return text.lower()
def remove_specials(sentence):
sentence = sentence.replace('-', ' ')
sentence = re.sub(r'[^\w\s]', '', sentence)
return sentence
def remove_stopwords(tokens):
words = []
for word in tokens:
if word not in nltk.corpus.stopwords.words('english'):
... | _____no_output_____ | MIT | notebooks/data_preprocessing.ipynb | SudoHead/movie-classifier |
Save the processed data: | # new_df.to_csv("../data/movies_data_ready.csv", index=False) | _____no_output_____ | MIT | notebooks/data_preprocessing.ipynb | SudoHead/movie-classifier |
Model Training | for a in alpha:
for i in range(5000):
for j in range(x.size):
h[j]=theta[0]*1 + theta[1]*x[j] + theta[2]*np.power(x[j],2) + theta[3]*np.power(x[j],3)
theta=theta+a*(y[j]-h[j])*np.array([1, x[j], np.power(x[j],2), np.power(x[j],3)])
if a==0.03:
error[i]=np.sum(np.s... | _____no_output_____ | MIT | Results/Jupyter/ML/ex1.ipynb | in2dblue/interactive-rl |
Dataset Bug DetectionIn this example, we will demonstrate how to detect bugs in a data set using the public Airlines data set. | # Since we use the category_encoders library to perform binary encoding on some of the features in this demo,
# we'll need to install it.
!pip install category_encoders
import pandas
pandas.options.display.max_rows=5 # restrict to 5 rows on display
df = pandas.read_csv("https://raw.githubusercontent.com/Devvrat53/Fli... | _____no_output_____ | Apache-2.0 | docs/notebooks/divergence/BugDetection.ipynb | FINRAOS/model-validation-toolkit |
Prepare daily dataLet's assume that we run new data each day through our model. For simplicity we will just look at the last 10 days of data. | df_daily = df[df['month'] > 11]
df_daily = df_daily[df_daily['day'] > 20]
df_daily | _____no_output_____ | Apache-2.0 | docs/notebooks/divergence/BugDetection.ipynb | FINRAOS/model-validation-toolkit |
Bug DetectionNow we want to find any bugs in any of our daily sets of data that we feed to our model.Note that we are performing binary encoding on the categorical columns (carrier, origin, and dest) so that we can pass the data to the variational estimation function directly. We are doing this for performance reasons... | import category_encoders as ce
from mvtk.supervisor.utils import compute_divergence_crosstabs
from mvtk.supervisor.divergence import calc_tv_knn
columns = ['dep_time', 'sched_dep_time', 'dep_delay', 'arr_time', 'sched_arr_time', 'arr_delay', 'air_time', 'distance', 'hour', 'minute', 'carrier', 'origin', 'dest']
encod... | _____no_output_____ | Apache-2.0 | docs/notebooks/divergence/BugDetection.ipynb | FINRAOS/model-validation-toolkit |
As you can see from the heatmap above, although there are some divergences between the days, there is nothing that is too alarming.Let's now update our data set to contain a "bug" in the "sched_dep_time" feature. For day 30, all of the values of that feature are null (which we are then translating to 0). | df_daily.loc[df_daily['day'] == 30, ['sched_dep_time']] = None | _____no_output_____ | Apache-2.0 | docs/notebooks/divergence/BugDetection.ipynb | FINRAOS/model-validation-toolkit |
Below is the percentage of scheduled departure times that are empty per day in our updated daily data set | day = 21
for df_day in df_daily.groupby('day'):
day_pct = df_day[1]['sched_dep_time'].value_counts(normalize=True, dropna=False) * 100
pct = day_pct.loc[day_pct.index.isnull()].values
if (len(pct) == 0):
pct = 0
else:
pct = pct[0]
print('Day ' + str(day) + ': ' + str(round(pct)) + '%... | _____no_output_____ | Apache-2.0 | docs/notebooks/divergence/BugDetection.ipynb | FINRAOS/model-validation-toolkit |
Think BayesThis notebook presents example code and exercise solutions for Think Bayes.Copyright 2018 Allen B. DowneyMIT License: https://opensource.org/licenses/MIT | # Configure Jupyter so figures appear in the notebook
%matplotlib inline
# Configure Jupyter to display the assigned value after an assignment
%config InteractiveShell.ast_node_interactivity='last_expr_or_assign'
# import classes from thinkbayes2
from thinkbayes2 import Hist, Pmf, Suite, Beta
import thinkplot
import... | _____no_output_____ | MIT | solutions/world_cup_soln.ipynb | chwebster/ThinkBayes2 |
Controllo di un braccio robotico con giunto flessibileUn collegamento di un braccio robotico è azionato da un motore elettrico tramite un giunto flessibile che si comporta come una molla torsionale. La dinamica del sistema può essere approssimata con un sistema lineare tempo invariante del terzo ordine in cui gli stat... | A = numpy.matrix('0 1 -1; -0.9 -0.09 0.09; 0.1 0.01 -0.01')
B = numpy.matrix('0; 90; 0')
C = numpy.matrix('0 0 1')
D = numpy.matrix('0')
sys_tf = control.tf(sss(A,B,C,D))
print(sys_tf) |
3.886e-16 s^2 + 0.9 s + 9
-----------------------------
s^3 + 0.1 s^2 + s - 8.674e-19
| BSD-3-Clause | ICCT_it/examples/04/.ipynb_checkpoints/SS-40-Controllo_di_un_braccio_robotico_con_giunto_flessibile-checkpoint.ipynb | ICCTerasmus/ICCT |
con poli | import warnings
# In order to suppress the warning BadCoefficient
warnings.filterwarnings("ignore")
print(numpy.round(sys_tf.pole(),3)) | [-0.05+0.999j -0.05-0.999j 0. +0.j ]
| BSD-3-Clause | ICCT_it/examples/04/.ipynb_checkpoints/SS-40-Controllo_di_un_braccio_robotico_con_giunto_flessibile-checkpoint.ipynb | ICCTerasmus/ICCT |
e zeri | print(numpy.round(sys_tf.zero(),3),'.') | [-2.31613695e+15 -1.00000000e+01] .
| BSD-3-Clause | ICCT_it/examples/04/.ipynb_checkpoints/SS-40-Controllo_di_un_braccio_robotico_con_giunto_flessibile-checkpoint.ipynb | ICCTerasmus/ICCT |
Innanzitutto, si analizza il sistema per verificare se è controllabile e osservabile. La matrice di controllabilità $\mathcal{C}$ è | Ctrb = control.ctrb(A,B)
display(Markdown(bmatrix(Ctrb)))
# print(numpy.linalg.matrix_rank(Ctrb)) | _____no_output_____ | BSD-3-Clause | ICCT_it/examples/04/.ipynb_checkpoints/SS-40-Controllo_di_un_braccio_robotico_con_giunto_flessibile-checkpoint.ipynb | ICCTerasmus/ICCT |
e ha rango pari a 3 quindi il sistema è controllabile. La matrice di osservabilità $\mathcal{O}$ è | Obsv = control.obsv(A,C)
display(Markdown(bmatrix(Obsv)))
# print(numpy.linalg.matrix_rank(Obsv)) | _____no_output_____ | BSD-3-Clause | ICCT_it/examples/04/.ipynb_checkpoints/SS-40-Controllo_di_un_braccio_robotico_con_giunto_flessibile-checkpoint.ipynb | ICCTerasmus/ICCT |
e ha rango pari a 3 quindi il sistema è osservabile.Ciò potrebbe essere effettivamente dedotto dal fatto che il denominatore della funzione di trasferimento è del terzo ordine (uguale alla dimensione del vettore nello spazio degli stati). Design del regolatore Design del controllerDati i requisiti, si sa che si devono ... | # Preparatory cell
X0 = numpy.matrix('0.0; 0.0; 0.0')
K = numpy.matrix([8/15,-4.4,-4])
L = numpy.matrix([[23],[66],[107/3]])
Aw = matrixWidget(3,3)
Aw.setM(A)
Bw = matrixWidget(3,1)
Bw.setM(B)
Cw = matrixWidget(1,3)
Cw.setM(C)
X0w = matrixWidget(3,1)
X0w.setM(X0)
Kw = matrixWidget(1,3)
Kw.setM(K)
Lw = matrixWidget(3,... | _____no_output_____ | BSD-3-Clause | ICCT_it/examples/04/.ipynb_checkpoints/SS-40-Controllo_di_un_braccio_robotico_con_giunto_flessibile-checkpoint.ipynb | ICCTerasmus/ICCT |
Data Science and Business Analytics Intern @ The Sparks Foundation Author : Aniket M. Wazarkar Task 2 : Prediction using Unsupervised ML Dataset : Iris.csv (https://bit.ly/3kXTdox) Algorithm used here : K-Means Clustering Import Libraries | %matplotlib inline
import matplotlib.pyplot as plt
import pandas as pd
import numpy as np
from sklearn.datasets import load_iris
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
from sklearn.cluster import KMeans
from sklearn.decomposition import PCA
| _____no_output_____ | Apache-2.0 | Task #2 Prediction using Unsupervised ML.ipynb | aniketspeaks/Task-2-Prediction-using-Unsupervised-ML |
Load Dataset | df=pd.read_csv('Iris.csv')
df | _____no_output_____ | Apache-2.0 | Task #2 Prediction using Unsupervised ML.ipynb | aniketspeaks/Task-2-Prediction-using-Unsupervised-ML |
K-Means is considered an unsupervised learning algorthm. This means you only need a features matrix. In the iris dataset, there are four features. In this notebook, the features matrix will only be two features as it is easier to visualize clusters in two dimensions. KMeans is a popular clustering algorithm that we can... |
df.info()
df.Species.unique()
df["Species"].value_counts() | _____no_output_____ | Apache-2.0 | Task #2 Prediction using Unsupervised ML.ipynb | aniketspeaks/Task-2-Prediction-using-Unsupervised-ML |
Arrange Data into Feature Matrix Use DataFrame.loc attribute to access a particular cell in the given Dataframe using the index and column labels. | features = ['PetalLengthCm','PetalWidthCm']
# Create features matrix
x = df.loc[:, features].values
x | _____no_output_____ | Apache-2.0 | Task #2 Prediction using Unsupervised ML.ipynb | aniketspeaks/Task-2-Prediction-using-Unsupervised-ML |
class sklearn.preprocessing.LabelEncoder.Encode target labels with value between 0 and n_classes-1. | from sklearn import preprocessing
le=preprocessing.LabelEncoder()
df.Species=le.fit_transform(df.Species.values)
df.Species
y=df.Species
y | _____no_output_____ | Apache-2.0 | Task #2 Prediction using Unsupervised ML.ipynb | aniketspeaks/Task-2-Prediction-using-Unsupervised-ML |
Standardize the data Standardize features by removing the mean and scaling to unit varianceThe standard score of a sample x is calculated as:z = (x - u) / swhere u is the mean of the training samples or zero if with_mean=False, and s is the standard deviation of the training samples or one if with_std=False. | x=StandardScaler().fit_transform(x) | _____no_output_____ | Apache-2.0 | Task #2 Prediction using Unsupervised ML.ipynb | aniketspeaks/Task-2-Prediction-using-Unsupervised-ML |
Plot data to estimate number of clusters | X=pd.DataFrame(x,columns=features)
plt.figure(figsize=(6,5))
plt.scatter(X['PetalLengthCm'], X['PetalWidthCm'])
plt.xlabel('petal length (cm)')
plt.ylabel('petal width (cm)');
plt.title('K-Means Clustering')
| _____no_output_____ | Apache-2.0 | Task #2 Prediction using Unsupervised ML.ipynb | aniketspeaks/Task-2-Prediction-using-Unsupervised-ML |
Finding the optimum number of clusters for K-means clustering | # Finding the optimum number of clusters for k-means classification
wcss = []
for i in range(1, 11):
kmeans = KMeans(n_clusters = i, init = 'k-means++',
max_iter = 300, n_init = 10, random_state = 0)
kmeans.fit(x)
wcss.append(kmeans.inertia_)
# Plotting the results onto a line graph,
... | _____no_output_____ | Apache-2.0 | Task #2 Prediction using Unsupervised ML.ipynb | aniketspeaks/Task-2-Prediction-using-Unsupervised-ML |
It is called 'The elbow method' from the above graph, the optimum clusters is where the elbow occurs. This is when the within cluster sum of squares (WCSS) doesn't decrease significantly with every iteration.From this we choose the number of clusters as **'3'**. K-Means Clustering | # Make an instance of KMeans with 3 clusters
kmeans = KMeans(n_clusters=3, random_state=1)
# Fit only on a features matrix
kmeans.fit(x)
# Get labels and cluster centroids
labels = kmeans.labels_
centroids = kmeans.cluster_centers_
labels
centroids | _____no_output_____ | Apache-2.0 | Task #2 Prediction using Unsupervised ML.ipynb | aniketspeaks/Task-2-Prediction-using-Unsupervised-ML |
Visually Evaluate the clusters | colormap = np.array(['r', 'g', 'b'])
plt.scatter(X['PetalLengthCm'], X['PetalWidthCm'], c=colormap[labels])
plt.scatter(centroids[:,0], centroids[:,1], s = 300, marker = 'x', c = 'k')
plt.xlabel('petal length (cm)')
plt.ylabel('petal width (cm)'); | _____no_output_____ | Apache-2.0 | Task #2 Prediction using Unsupervised ML.ipynb | aniketspeaks/Task-2-Prediction-using-Unsupervised-ML |
Visually Evaluate the clusters and compare the species | plt.figure(figsize=(10,4))
plt.subplot(1, 2, 1)
plt.scatter(X['PetalLengthCm'], X['PetalWidthCm'], c=colormap[labels])
plt.scatter(centroids[:,0], centroids[:,1], s = 300, marker = 'x', c = 'k')
plt.xlabel('petal length (cm)')
plt.ylabel('petal width (cm)');
plt.title('K-Means Clustering (k = 3)')
plt.subplot(1, 2, ... | _____no_output_____ | Apache-2.0 | Task #2 Prediction using Unsupervised ML.ipynb | aniketspeaks/Task-2-Prediction-using-Unsupervised-ML |
They look pretty similar. Looks like KMeans picked up flower differences with only two features and not the labels. The colors are different in the two graphs simply because KMeans gives out a arbitrary cluster number and the iris dataset has an arbitrary number in the target column. PCA Projection in 2D The original... | pca = PCA(n_components=2)
# Fit and transform the data
principalComponents = pca.fit_transform(x)
principalDf = pd.DataFrame(data = principalComponents, columns = ['principal component 1', 'principal component 2'])
df=pd.read_csv('Iris.csv') | _____no_output_____ | Apache-2.0 | Task #2 Prediction using Unsupervised ML.ipynb | aniketspeaks/Task-2-Prediction-using-Unsupervised-ML |
2D Projection | finalDf = pd.concat([principalDf, df[['Species']]], axis = 1)
finalDf
fig, ax = plt.subplots(nrows = 1, ncols = 1, figsize = (8,8));
targets = df.loc[:, 'Species'].unique()
colors = ['r', 'g', 'b']
for target, color in zip(targets,colors):
indicesToKeep = finalDf['Species'] == target
ax.scatter(finalDf.loc[ind... | _____no_output_____ | Apache-2.0 | Task #2 Prediction using Unsupervised ML.ipynb | aniketspeaks/Task-2-Prediction-using-Unsupervised-ML |
From the graph, it looks like the setosa class is well separated from the versicolor and virginica classes. Explained Varience The explained variance tells us how much information (variance) can be attributed to each of the principal components. This is important as while you can convert 4 dimensional space to 2 dimen... | pca.explained_variance_ratio_
sum(pca.explained_variance_ratio_) | _____no_output_____ | Apache-2.0 | Task #2 Prediction using Unsupervised ML.ipynb | aniketspeaks/Task-2-Prediction-using-Unsupervised-ML |
Dynamics 365 Business Central Trouble Shooting Guide (TSG) - Performance analysis (overview)
This notebook contains Kusto queries that can help getting to the root cause of a performance issue for an environment. Each section in the notebook contains links to the performance tuning guide on docs [aka.ms/bcperformance... | # load the KQLmagic module
%reload_ext Kqlmagic
# Connect to the Application Insights API
%kql appinsights://appid='<add app id from the Application Insights portal>';appkey='<add API key from the Application Insights portal>' | _____no_output_____ | MIT | samples/AppInsights/TroubleShootingGuides/Performance-overview-TSG.ipynb | dmc-dk/BCTech |
2. Define filters
This workbook is designed for troubleshooting a single environment. Please provide values for aadTenantId and environmentName: | aadTenantId = "<Add AAD tenant id here>"
environmentName = "<add environment name here>" | _____no_output_____ | MIT | samples/AppInsights/TroubleShootingGuides/Performance-overview-TSG.ipynb | dmc-dk/BCTech |
Analyze performance
Now you can run Kusto queries to look for possible root causes for performance issues.
Either click **Run All** above to run all sections, or scroll down to the type of analysis you want to do and manually run queries Sessions
Performance tuning guide: https://docs.microsoft.com/en-us/dynamics36... | %%kql
let _aadTenantId = aadTenantId;
let _environmentName = environmentName;
traces
| where 1==1
and customDimensions.aadTenantId == _aadTenantId
and customDimensions.environmentName == _environmentName
and customDimensions.eventId == 'RT0004'
and timestamp > ago(7d)
| extend clientType = tos... | _____no_output_____ | MIT | samples/AppInsights/TroubleShootingGuides/Performance-overview-TSG.ipynb | dmc-dk/BCTech |
Web service requests
Performance tuning guide: https://docs.microsoft.com/en-us/dynamics365/business-central/dev-itpro/performance/performance-developerwriting-efficient-web-services
Web service telemetry docs: https://docs.microsoft.com/en-us/dynamics365/business-central/dev-itpro/administration/telemetry-webservic... | %%kql
let _aadTenantId = aadTenantId;
let _environmentName = environmentName;
traces
| where 1==1
and customDimensions.aadTenantId == _aadTenantId
and customDimensions.environmentName == _environmentName
and customDimensions.eventId == 'RT0008'
and timestamp > ago(7d)
| extend category = tostr... | _____no_output_____ | MIT | samples/AppInsights/TroubleShootingGuides/Performance-overview-TSG.ipynb | dmc-dk/BCTech |
Data related
Performance tuning guide:
* [Efficient data access](https://docs.microsoft.com/en-us/dynamics365/business-central/dev-itpro/performance/performance-developerefficient-data-access)
* [Avoid locking](https://docs.microsoft.com/en-us/dynamics365/business-central/dev-itpro/performance/performance-applicatio... | %%kql
let _aadTenantId = aadTenantId;
let _environmentName = environmentName;
traces
| where 1==1
and customDimensions.aadTenantId == _aadTenantId
and customDimensions.environmentName == _environmentName
and customDimensions.eventId == 'RT0005'
and timestamp > ago(7d)
| summarize count() by bi... | _____no_output_____ | MIT | samples/AppInsights/TroubleShootingGuides/Performance-overview-TSG.ipynb | dmc-dk/BCTech |
Company management
Operations such as "copy company" can cause performance degradations if they are done when users are logged into the system.
Read more in the performance tuning guide here: https://docs.microsoft.com/en-us/dynamics365/business-central/dev-itpro/performance/performance-applicationbe-cautious-with-t... | %%kql
let _aadTenantId = aadTenantId;
let _environmentName = environmentName;
traces
| where 1==1
and customDimensions.aadTenantId == _aadTenantId
and customDimensions.environmentName == _environmentName
and customDimensions.eventId in ('LC0001')
and timestamp > ago(7d)
| extend operation_typ... | _____no_output_____ | MIT | samples/AppInsights/TroubleShootingGuides/Performance-overview-TSG.ipynb | dmc-dk/BCTech |
Reports
Learn more about how to write performant reports here in the performance tuning guide: https://docs.microsoft.com/en-us/dynamics365/business-central/dev-itpro/performance/performance-developerwriting-efficient-reports
Report telemetry docs: https://docs.microsoft.com/en-us/dynamics365/business-central/dev-it... | %%kql
let _aadTenantId = aadTenantId;
let _environmentName = environmentName;
traces
| where 1==1
and customDimensions.aadTenantId == _aadTenantId
and customDimensions.environmentName == _environmentName
and customDimensions.eventId == 'RT0006'
and timestamp > ago(7d)
| extend clientType = tos... | _____no_output_____ | MIT | samples/AppInsights/TroubleShootingGuides/Performance-overview-TSG.ipynb | dmc-dk/BCTech |
Page views
Page view telemetry docs: https://docs.microsoft.com/en-us/dynamics365/business-central/dev-itpro/administration/telemetry-page-view-trace
KQL samples
* https://github.com/microsoft/BCTech/blob/master/samples/AppInsights/KQL/RawData/PageViews.kql
* https://github.com/microsoft/BCTech/blob/master/samples/A... | %%kql
// Top 10 longest page times
//
let _aadTenantId = aadTenantId;
let _environmentName = environmentName;
pageViews
| where 1==1
and customDimensions.aadTenantId == _aadTenantId
// and customDimensions.environmentName == _environmentName
| where timestamp > ago(7d)
| extend objectId = tostring(c... | _____no_output_____ | MIT | samples/AppInsights/TroubleShootingGuides/Performance-overview-TSG.ipynb | dmc-dk/BCTech |
FloPy Creating a Complex MODFLOW 6 Model with FlopyThe purpose of this notebook is to demonstrate the Flopy capabilities for building a more complex MODFLOW 6 model from scratch. This notebook will demonstrate the capabilities by replicating the advgw_tidal model that is distributed with MODFLOW 6. Setup the Noteboo... | %matplotlib inline
import sys
import os
import platform
import numpy as np
import matplotlib as mpl
import matplotlib.pyplot as plt
# run installed version of flopy or add local path
try:
import flopy
except:
fpth = os.path.abspath(os.path.join('..', '..'))
sys.path.append(fpth)
import flopy
print(sys... | _____no_output_____ | CC0-1.0 | examples/Notebooks/flopy3_mf6_B_complex-model.ipynb | gyanz/flopy |
Create the MODFLOW 6 Input Files and Run the ModelOnce all the flopy objects are created, it is very easy to create all of the input files and run the model. | # change folder to save simulation
#sim.simulation_data.mfpath.set_sim_path(run_folder)
# write simulation to new location
sim.write_simulation()
# Print a list of the files that were created
# in workspace
print(os.listdir(workspace)) | ['intercell_flow_obs1.csv', 'riv_flowsA.csv', 'advgw_tidal.nam', 'advgw_tidal.ic', 'riv_flowsB.csv', 'advgw_tidal.dis.grb', 'recharge_rates_1.ts', 'advgw_tidal.sto', 'advgw_tidal.lst', 'tides.ts', 'advgw_tidal.cbb', 'advgw_tidal.ims', 'well-rates.ts', 'advgw_tidal.ghb', 'advgw_tidal.obs', 'advgw_tidal.riv.obs', 'ghb_fl... | CC0-1.0 | examples/Notebooks/flopy3_mf6_B_complex-model.ipynb | gyanz/flopy |
Run the SimulationWe can also run the simulation from the notebook, but only if the MODFLOW 6 executable is available. The executable can be made available by putting the executable in a folder that is listed in the system path variable. Another option is to just put a copy of the executable in the simulation folder... | # Run the simulation
success, buff = sim.run_simulation()
print('\nSuccess is: ', success) | FloPy is using the following executable to run the model: /Users/jdhughes/.local/bin/mf6
MODFLOW 6
U.S. GEOLOGICAL SURVEY MODULAR HYDROLOGIC MODEL
VERSION 6.0.3 08/09/2018
MODFLOW 6 compiled Sep 24 2018 16:09:01 with GFORTRAN compiler (v... | CC0-1.0 | examples/Notebooks/flopy3_mf6_B_complex-model.ipynb | gyanz/flopy |
Post-Process Head ResultsPost-processing MODFLOW 6 results is still a work in progress. There aren't any Flopy plotting functions built in yet, like they are for other MODFLOW versions. So we need to plot the results using general Flopy capabilities. We can also use some of the Flopy ModelMap capabilities for MODFL... | # Read the binary head file and plot the results
# We can use the existing Flopy HeadFile class because
# the format of the headfile for MODFLOW 6 is the same
# as for previous MODFLOW verions
headfile = '{}.hds'.format(model_name)
fname = os.path.join(workspace, headfile)
hds = flopy.utils.binaryfile.HeadFile(fname)
h... | _____no_output_____ | CC0-1.0 | examples/Notebooks/flopy3_mf6_B_complex-model.ipynb | gyanz/flopy |
Post-Process FlowsMODFLOW 6 writes a binary grid file, which contains information about the model grid. MODFLOW 6 also writes a binary budget file, which contains flow information. Both of these files can be read using Flopy capabilities. The MfGrdFile class in Flopy can be used to read the binary grid file. The C... | # read the binary grid file
fname = os.path.join(workspace, '{}.dis.grb'.format(model_name))
bgf = flopy.utils.mfgrdfile.MfGrdFile(fname)
# data read from the binary grid file is stored in a dictionary
bgf._datadict
# Information from the binary grid file is easily retrieved
ia = bgf._datadict['IA'] - 1
ja = bgf._data... | _____no_output_____ | CC0-1.0 | examples/Notebooks/flopy3_mf6_B_complex-model.ipynb | gyanz/flopy |
Clustering text documents using k-meansAs an example we'll be using the 20 newsgroups dataset consists of 18000+ newsgroup posts on 20 topics. You can learn more about the dataset at http://qwone.com/~jason/20Newsgroups/ | from sklearn.datasets import fetch_20newsgroups
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.preprocessing import Normalizer
from sklearn import metrics
import matplotlib.pyplot as plt
from sklearn.cluster import KMeans, MiniBatchKMeans
import numpy as np | _____no_output_____ | MIT | jupyter_notebooks/machine_learning/ebook_mastering_ml_in_6_steps/Chapter_5_Code/Code/Document_Clustering.ipynb | manual123/Nacho-Jupyter-Notebooks |
Load data | newsgroups_train = fetch_20newsgroups(subset='train')
print(list(newsgroups_train.target_names)) | ['alt.atheism', 'comp.graphics', 'comp.os.ms-windows.misc', 'comp.sys.ibm.pc.hardware', 'comp.sys.mac.hardware', 'comp.windows.x', 'misc.forsale', 'rec.autos', 'rec.motorcycles', 'rec.sport.baseball', 'rec.sport.hockey', 'sci.crypt', 'sci.electronics', 'sci.med', 'sci.space', 'soc.religion.christian', 'talk.politics.gu... | MIT | jupyter_notebooks/machine_learning/ebook_mastering_ml_in_6_steps/Chapter_5_Code/Code/Document_Clustering.ipynb | manual123/Nacho-Jupyter-Notebooks |
To keep it simple, let's filter only 3 topics. Assume that we do not know the topics, let's run clustering algorithm and examine the keywords of each clusters | categories = ['alt.atheism', 'comp.graphics', 'rec.motorcycles']
dataset = fetch_20newsgroups(subset='all', categories=categories, shuffle=True, random_state=2017)
print("%d documents" % len(dataset.data))
print("%d categories" % len(dataset.target_names))
labels = dataset.target
print("Extracting features from the... | 2768 documents
3 categories
Extracting features from the dataset using a sparse vectorizer
n_samples: 2768, n_features: 35311
| MIT | jupyter_notebooks/machine_learning/ebook_mastering_ml_in_6_steps/Chapter_5_Code/Code/Document_Clustering.ipynb | manual123/Nacho-Jupyter-Notebooks |
LSA via SVDLatent Semantic Analysis (LSA) is a mathematical method that tries to bring out latent relationships within a collection of documents. Rather than looking at each document isolated from the others it looks at all the documents as a whole and the terms within them to identify relationships. Let's perform LSA... | from IPython.display import Image
Image(filename='../Chapter 5 Figures/SVD.png', width=500)
from sklearn.decomposition import TruncatedSVD
# Lets reduce the dimensionality to 2000
svd = TruncatedSVD(2000)
lsa = make_pipeline(svd, Normalizer(copy=False))
X = lsa.fit_transform(X)
explained_variance = svd.explained_var... | Explained variance of the SVD step: 95%
| MIT | jupyter_notebooks/machine_learning/ebook_mastering_ml_in_6_steps/Chapter_5_Code/Code/Document_Clustering.ipynb | manual123/Nacho-Jupyter-Notebooks |
k-means clustering | from __future__ import print_function
km = KMeans(n_clusters=3, init='k-means++', max_iter=100, n_init=1)
# Scikit learn provides MiniBatchKMeans to run k-means in batch mode suitable for a very large corpus
# km = MiniBatchKMeans(n_clusters=5, init='k-means++', n_init=1, init_size=1000, batch_size=1000)
print("Clus... | Clustering sparse data with KMeans(algorithm='auto', copy_x=True, init='k-means++', max_iter=100,
n_clusters=3, n_init=1, n_jobs=1, precompute_distances='auto',
random_state=None, tol=0.0001, verbose=0)
Top terms per cluster:
Cluster 0: edu graphics university god subject lines organization com posting uk
Clust... | MIT | jupyter_notebooks/machine_learning/ebook_mastering_ml_in_6_steps/Chapter_5_Code/Code/Document_Clustering.ipynb | manual123/Nacho-Jupyter-Notebooks |
Hierarchical clustering | from sklearn.metrics.pairwise import cosine_similarity
dist = 1 - cosine_similarity(X)
from scipy.cluster.hierarchy import ward, dendrogram
linkage_matrix = ward(dist) #define the linkage_matrix using ward clustering pre-computed distances
fig, ax = plt.subplots(figsize=(8, 8)) # set size
ax = dendrogram(linkage_matr... | _____no_output_____ | MIT | jupyter_notebooks/machine_learning/ebook_mastering_ml_in_6_steps/Chapter_5_Code/Code/Document_Clustering.ipynb | manual123/Nacho-Jupyter-Notebooks |
10 May 2017 - Lecture 2 JNB Code Along - WH Nixalo[Notebook](https://github.com/fastai/courses/blob/ed1fb08d86df277d2736972a1ff1ac39ea1ac733/deeplearning1/nbs/lesson2.ipynb) | Lecture[1:20:00](https://www.youtube.com/watch?v=e3aM6XTekJc) 1 Linear models with CNN features | # This is to point Python to my utils folder
import sys; import os
# DIR = %pwd
sys.path.insert(1, os.path.join('../utils'))
# Rather than importing everything manually, we'll make things easy
# and load them all in utils.py, and just import them from there.
import utils; reload(utils)
from utils import *
%matplotli... | Using Theano backend.
| MIT | FAI_old/lesson2/lesson2_codealong.ipynb | WNoxchi/Kawkasos |
1.1 IntroWe need to find a way to convert the imagenet predictions to a probability of being a cat or a dog, since that is what the Kaggle copmetition requires us to submit. We could use the imagenet hierarchy to download a list of all the imagenet categories in each of the dog and cat groups, and could then solve ou... | %matplotlib inline
from __future__ import division, print_function
import os, json
from glob import glob
import numpy as np
import scipy
from sklearn.preprocessing import OneHotEncoder
from sklearn.metrics import confusion_matrix
np.set_printoptions(precision=4, linewidth=100)
from matplotlib import pyplot as plt
impo... | _____no_output_____ | MIT | FAI_old/lesson2/lesson2_codealong.ipynb | WNoxchi/Kawkasos |
1.2 Linear models in kerasLet's forget the motivating example for a second a see how we can create a simple Linear model in Keras:Each of the ```Dense()``` layers is just a *linear* model, followed by a simple *activation function*.In a linear model each row is calculated as ```sum(row * weights)```, where weights nee... | # we'll create a random matrix w/ 2 columns; & do a MatMul to get our
# y value using a vector [2, 3] & adding a constant of 1.
x = random((30, 2))
y = np.dot(x, [2., 3.]) + 1.
x[:5]
y[:5] | _____no_output_____ | MIT | FAI_old/lesson2/lesson2_codealong.ipynb | WNoxchi/Kawkasos |
We can use kears to create a simple linear model (```Dense()``` - with no activation - in Keras) and optimize it using SGD to minimize mean squared error. | # Keras calls the Linear Model "Dense"; aka. "Fully-Connected" in other
# libraries.
# So when we go 'Dense' w/ an input of 2 columns, & output of 1 col,
# we're defining a linear model that can go from the 2 col array above, to
# the 1 col output of y above.
# Sequential() is a way of building multiple-layer network... | _____no_output_____ | MIT | FAI_old/lesson2/lesson2_codealong.ipynb | WNoxchi/Kawkasos |
Above is everything Keras is doing behind the scenes.So, if we pass multiple layers to Keras via ```Sequential(..)```, we can start to build & optimize Deep Neural Networks.Before that, we can still use the single-layer LM to create a pretty decent entry to the dogs-vs-cats Kaggle competition. 1.3 Train Linear Model o... | # setup the directories
os.mkdir('data')
os.mkdir('data/dogscats')
path = "data/dogscats/"
model_path = path + 'models/'
# if the path to our models DNE, make it
if not os.path.exists(model_path): os.mkdir(model_path)
# NOTE: os.mkdir(..) only works for a single folder
# Also will throw error if dir already exis... | _____no_output_____ | MIT | FAI_old/lesson2/lesson2_codealong.ipynb | WNoxchi/Kawkasos |
We'll process as many images at a time as we can. This is a case of T&E to find the max batch size that doesn't cause a memory error. | batch_size = 100 | _____no_output_____ | MIT | FAI_old/lesson2/lesson2_codealong.ipynb | WNoxchi/Kawkasos |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.