tra-analysis/analysis-master/tra_analysis/analysis.py

1487 lines
43 KiB
Python
Raw Normal View History

2020-03-10 03:58:51 +00:00
# Titan Robotics Team 2022: Data Analysis Module
# Written by Arthur Lu & Jacob Levine
# Notes:
# this should be imported as a python module using 'from tra_analysis import analysis'
2020-03-10 03:58:51 +00:00
# this should be included in the local directory or environment variable
# this module has been optimized for multhreaded computing
# current benchmark of optimization: 1.33 times faster
# setup:
__version__ = "1.2.2.000"
2020-03-10 03:58:51 +00:00
# changelog should be viewed using print(analysis.__changelog__)
__changelog__ = """changelog:
1.2.2.000:
- added Sort class
- added several array sorting functions to Sort class including:
- quick sort
- merge sort
- intro(spective) sort
- heap sort
- insertion sort
- tim sort
- selection sort
- bubble sort
- cycle sort
- cocktail sort
- tested all sorting algorithms with both lists and numpy arrays
- depreciated sort function from Array class
- added warnings as an import
1.2.1.004:
- added sort and search functions to Array class
1.2.1.003:
Merge service-dev changes with master (#24) * added config.json removed old config files Signed-off-by: Arthur <learthurgo@gmail.com> * superscript.py v 0.0.6.000 Signed-off-by: Arthur <learthurgo@gmail.com> * changed data.py Signed-off-by: Arthur <learthurgo@gmail.com> * changes to config.json Signed-off-by: Arthur <learthurgo@gmail.com> * removed cells from visualize_pit.py Signed-off-by: Arthur <learthurgo@gmail.com> * more changes to visualize_pit.py Signed-off-by: Arthur <learthurgo@gmail.com> * added analysis-master/metrics/__pycache__ to git ignore moved pit configs in config.json to the borrom superscript.py v 0.0.6.001 Signed-off-by: Arthur <learthurgo@gmail.com> * removed old database key Signed-off-by: Arthur <learthurgo@gmail.com> * adjusted config files Signed-off-by: Arthur <learthurgo@gmail.com> * Delete config-pop.json * fixed .gitignore Signed-off-by: Arthur <learthurgo@gmail.com> * analysis.py 1.2.1.003 added team kv pair to config.json Signed-off-by: Arthur <learthurgo@gmail.com> * superscript.py v 0.0.6.002 Signed-off-by: Arthur <learthurgo@gmail.com> * finished app.py API made minute changes to parentheses use in various packages Signed-off-by: Arthur Lu <learthurgo@gmail.com> * bug fixes in app.py Signed-off-by: Arthur Lu <learthurgo@gmail.com> * bug fixes in app.py Signed-off-by: Arthur Lu <learthurgo@gmail.com> * made changes to .gitignore Signed-off-by: Arthur Lu <learthurgo@gmail.com> * made changes to .gitignore Signed-off-by: Arthur Lu <learthurgo@gmail.com> * deleted a __pycache__ folder from metrics Signed-off-by: Arthur Lu <learthurgo@gmail.com> * more changes to .gitignore Signed-off-by: Arthur Lu <learthurgo@gmail.com> * additions to app.py Signed-off-by: Arthur Lu <learthurgo@gmail.com> * renamed app.py to api.py Signed-off-by: Arthur Lu <learthurgo@gmail.com> * removed extranneous files Signed-off-by: Arthur Lu <learthurgo@gmail.com> * renamed api.py to tra.py removed rest api calls from tra.py * renamed api.py to tra.py removed rest api calls from tra.py Signed-off-by: Arthur Lu <learthurgo@gmail.com> * removed flask import from tra.py Signed-off-by: Arthur Lu <learthurgo@gmail.com> * changes to devcontainer.json Signed-off-by: Arthur Lu <learthurgo@gmail.com> * fixed unit tests to be correct removed some tests regressions because of potential function overflow removed trueskill unit test because of slight deviation chance Signed-off-by: Arthur Lu <learthurgo@gmail.com>
2020-05-20 13:52:38 +00:00
- changed output of basic_stats and histo_analysis to libraries
- fixed __all__
1.2.1.002:
- renamed ArrayTest class to Array
2020-05-13 01:19:58 +00:00
1.2.1.001:
- added add, mul, neg, and inv functions to ArrayTest class
- added normalize function to ArrayTest class
- added dot and cross functions to ArrayTest class
2020-05-04 19:50:36 +00:00
1.2.1.000:
- added ArrayTest class
- added elementwise mean, median, standard deviation, variance, min, max functions to ArrayTest class
- added elementwise_stats to ArrayTest which encapsulates elementwise statistics
- appended to __all__ to reflect changes
2020-05-04 16:59:25 +00:00
1.2.0.006:
- renamed func functions in regression to lin, log, exp, and sig
2020-05-02 03:59:54 +00:00
1.2.0.005:
- moved random_forrest_regressor and random_forrest_classifier to RandomForrest class
- renamed Metrics to Metric
- renamed RegressionMetrics to RegressionMetric
- renamed ClassificationMetrics to ClassificationMetric
- renamed CorrelationTests to CorrelationTest
- renamed StatisticalTests to StatisticalTest
- reflected rafactoring to all mentions of above classes/functions
1.2.0.004:
- fixed __all__ to reflected the correct functions and classes
- fixed CorrelationTests and StatisticalTests class functions to require self invocation
- added missing math import
- fixed KNN class functions to require self invocation
- fixed Metrics class functions to require self invocation
- various spelling fixes in CorrelationTests and StatisticalTests
1.2.0.003:
- bug fixes with CorrelationTests and StatisticalTests
- moved glicko2 and trueskill to the metrics subpackage
- moved elo to a new metrics subpackage
1.2.0.002:
- fixed docs
1.2.0.001:
- fixed docs
1.2.0.000:
- cleaned up wild card imports with scipy and sklearn
- added CorrelationTests class
- added StatisticalTests class
- added several correlation tests to CorrelationTests
- added several statistical tests to StatisticalTests
1.1.13.009:
- moved elo, glicko2, trueskill functions under class Metrics
1.1.13.008:
- moved Glicko2 to a seperate package
1.1.13.007:
- fixed bug with trueskill
1.1.13.006:
- cleaned up imports
1.1.13.005:
- cleaned up package
1.1.13.004:
- small fixes to regression to improve performance
1.1.13.003:
- filtered nans from regression
1.1.13.002:
- removed torch requirement, and moved Regression back to regression.py
1.1.13.001:
- bug fix with linear regression not returning a proper value
- cleaned up regression
- fixed bug with polynomial regressions
1.1.13.000:
- fixed all regressions to now properly work
1.1.12.006:
- fixed bg with a division by zero in histo_analysis
1.1.12.005:
- fixed numba issues by removing numba from elo, glicko2 and trueskill
1.1.12.004:
- renamed gliko to glicko
1.1.12.003:
- removed depreciated code
1.1.12.002:
- removed team first time trueskill instantiation in favor of integration in superscript.py
1.1.12.001:
- improved readibility of regression outputs by stripping tensor data
- used map with lambda to acheive the improved readibility
- lost numba jit support with regression, and generated_jit hangs at execution
- TODO: reimplement correct numba integration in regression
1.1.12.000:
- temporarily fixed polynomial regressions by using sklearn's PolynomialFeatures
1.1.11.010:
- alphabeticaly ordered import lists
1.1.11.009:
- bug fixes
1.1.11.008:
- bug fixes
1.1.11.007:
- bug fixes
1.1.11.006:
- tested min and max
- bug fixes
1.1.11.005:
- added min and max in basic_stats
1.1.11.004:
- bug fixes
1.1.11.003:
- bug fixes
1.1.11.002:
- consolidated metrics
- fixed __all__
1.1.11.001:
- added test/train split to RandomForestClassifier and RandomForestRegressor
1.1.11.000:
- added RandomForestClassifier and RandomForestRegressor
- note: untested
1.1.10.000:
- added numba.jit to remaining functions
1.1.9.002:
- kernelized PCA and KNN
1.1.9.001:
- fixed bugs with SVM and NaiveBayes
1.1.9.000:
- added SVM class, subclasses, and functions
- note: untested
1.1.8.000:
- added NaiveBayes classification engine
- note: untested
1.1.7.000:
- added knn()
- added confusion matrix to decisiontree()
1.1.6.002:
- changed layout of __changelog to be vscode friendly
1.1.6.001:
- added additional hyperparameters to decisiontree()
1.1.6.000:
- fixed __version__
- fixed __all__ order
- added decisiontree()
1.1.5.003:
- added pca
1.1.5.002:
- reduced import list
- added kmeans clustering engine
1.1.5.001:
- simplified regression by using .to(device)
1.1.5.000:
- added polynomial regression to regression(); untested
1.1.4.000:
- added trueskill()
1.1.3.002:
- renamed regression class to Regression, regression_engine() to regression gliko2_engine class to Gliko2
1.1.3.001:
- changed glicko2() to return tuple instead of array
1.1.3.000:
- added glicko2_engine class and glicko()
- verified glicko2() accuracy
1.1.2.003:
- fixed elo()
1.1.2.002:
- added elo()
- elo() has bugs to be fixed
1.1.2.001:
- readded regrression import
1.1.2.000:
- integrated regression.py as regression class
- removed regression import
- fixed metadata for regression class
- fixed metadata for analysis class
1.1.1.001:
- regression_engine() bug fixes, now actaully regresses
1.1.1.000:
- added regression_engine()
- added all regressions except polynomial
1.1.0.007:
- updated _init_device()
1.1.0.006:
- removed useless try statements
1.1.0.005:
- removed impossible outcomes
1.1.0.004:
- added performance metrics (r^2, mse, rms)
1.1.0.003:
- resolved nopython mode for mean, median, stdev, variance
1.1.0.002:
- snapped (removed) majority of uneeded imports
- forced object mode (bad) on all jit
- TODO: stop numba complaining about not being able to compile in nopython mode
1.1.0.001:
- removed from sklearn import * to resolve uneeded wildcard imports
1.1.0.000:
- removed c_entities,nc_entities,obstacles,objectives from __all__
- applied numba.jit to all functions
- depreciated and removed stdev_z_split
- cleaned up histo_analysis to include numpy and numba.jit optimizations
- depreciated and removed all regression functions in favor of future pytorch optimizer
- depreciated and removed all nonessential functions (basic_analysis, benchmark, strip_data)
- optimized z_normalize using sklearn.preprocessing.normalize
- TODO: implement kernel/function based pytorch regression optimizer
1.0.9.000:
- refactored
- numpyed everything
- removed stats in favor of numpy functions
1.0.8.005:
- minor fixes
1.0.8.004:
- removed a few unused dependencies
1.0.8.003:
- added p_value function
1.0.8.002:
- updated __all__ correctly to contain changes made in v 1.0.8.000 and v 1.0.8.001
1.0.8.001:
- refactors
- bugfixes
1.0.8.000:
- depreciated histo_analysis_old
- depreciated debug
- altered basic_analysis to take array data instead of filepath
- refactor
- optimization
1.0.7.002:
- bug fixes
1.0.7.001:
- bug fixes
1.0.7.000:
- added tanh_regression (logistical regression)
- bug fixes
1.0.6.005:
- added z_normalize function to normalize dataset
- bug fixes
1.0.6.004:
- bug fixes
1.0.6.003:
- bug fixes
1.0.6.002:
- bug fixes
1.0.6.001:
- corrected __all__ to contain all of the functions
1.0.6.000:
- added calc_overfit, which calculates two measures of overfit, error and performance
- added calculating overfit to optimize_regression
1.0.5.000:
- added optimize_regression function, which is a sample function to find the optimal regressions
- optimize_regression function filters out some overfit funtions (functions with r^2 = 1)
- planned addition: overfit detection in the optimize_regression function
1.0.4.002:
- added __changelog__
- updated debug function with log and exponential regressions
1.0.4.001:
- added log regressions
- added exponential regressions
- added log_regression and exp_regression to __all__
1.0.3.008:
- added debug function to further consolidate functions
1.0.3.007:
- added builtin benchmark function
- added builtin random (linear) data generation function
- added device initialization (_init_device)
1.0.3.006:
- reorganized the imports list to be in alphabetical order
- added search and regurgitate functions to c_entities, nc_entities, obstacles, objectives
1.0.3.005:
- major bug fixes
- updated historical analysis
- depreciated old historical analysis
1.0.3.004:
- added __version__, __author__, __all__
- added polynomial regression
- added root mean squared function
- added r squared function
1.0.3.003:
- bug fixes
- added c_entities
1.0.3.002:
- bug fixes
- added nc_entities, obstacles, objectives
- consolidated statistics.py to analysis.py
1.0.3.001:
- compiled 1d, column, and row basic stats into basic stats function
1.0.3.000:
- added historical analysis function
1.0.2.xxx:
- added z score test
1.0.1.xxx:
- major bug fixes
1.0.0.xxx:
- added loading csv
- added 1d, column, row basic stats
2020-03-10 03:58:51 +00:00
"""
__author__ = (
"Arthur Lu <learthurgo@gmail.com>",
"Jacob Levine <jlevine@imsa.edu>",
2020-03-10 03:58:51 +00:00
)
__all__ = [
'load_csv',
'basic_stats',
'z_score',
'z_normalize',
'histo_analysis',
'regression',
2020-05-02 03:59:54 +00:00
'Metric',
'RegressionMetric',
'ClassificationMetric',
'kmeans',
'pca',
'decisiontree',
'KNN',
'NaiveBayes',
'SVM',
2020-05-02 03:59:54 +00:00
'RandomForrest',
'CorrelationTest',
'StatisticalTest',
'Array',
# all statistics functions left out due to integration in other functions
2020-03-10 03:58:51 +00:00
]
# now back to your regularly scheduled programming:
# imports (now in alphabetical order! v 1.0.3.006):
import csv
from tra_analysis.metrics import elo as Elo
from tra_analysis.metrics import glicko2 as Glicko2
import math
2020-03-10 03:58:51 +00:00
import numba
from numba import jit
import numpy as np
import scipy
2020-04-13 19:58:04 +00:00
from scipy import optimize, stats
2020-03-10 03:58:51 +00:00
import sklearn
2020-04-13 19:58:04 +00:00
from sklearn import preprocessing, pipeline, linear_model, metrics, cluster, decomposition, tree, neighbors, naive_bayes, svm, model_selection, ensemble
from tra_analysis.metrics import trueskill as Trueskill
import warnings
2020-03-10 03:58:51 +00:00
class error(ValueError):
pass
2020-03-10 03:58:51 +00:00
def load_csv(filepath):
with open(filepath, newline='') as csvfile:
file_array = np.array(list(csv.reader(csvfile)))
csvfile.close()
return file_array
2020-03-10 03:58:51 +00:00
# expects 1d array
@jit(forceobj=True)
def basic_stats(data):
data_t = np.array(data).astype(float)
2020-03-10 03:58:51 +00:00
_mean = mean(data_t)
_median = median(data_t)
_stdev = stdev(data_t)
_variance = variance(data_t)
_min = npmin(data_t)
_max = npmax(data_t)
2020-03-10 03:58:51 +00:00
Merge service-dev changes with master (#24) * added config.json removed old config files Signed-off-by: Arthur <learthurgo@gmail.com> * superscript.py v 0.0.6.000 Signed-off-by: Arthur <learthurgo@gmail.com> * changed data.py Signed-off-by: Arthur <learthurgo@gmail.com> * changes to config.json Signed-off-by: Arthur <learthurgo@gmail.com> * removed cells from visualize_pit.py Signed-off-by: Arthur <learthurgo@gmail.com> * more changes to visualize_pit.py Signed-off-by: Arthur <learthurgo@gmail.com> * added analysis-master/metrics/__pycache__ to git ignore moved pit configs in config.json to the borrom superscript.py v 0.0.6.001 Signed-off-by: Arthur <learthurgo@gmail.com> * removed old database key Signed-off-by: Arthur <learthurgo@gmail.com> * adjusted config files Signed-off-by: Arthur <learthurgo@gmail.com> * Delete config-pop.json * fixed .gitignore Signed-off-by: Arthur <learthurgo@gmail.com> * analysis.py 1.2.1.003 added team kv pair to config.json Signed-off-by: Arthur <learthurgo@gmail.com> * superscript.py v 0.0.6.002 Signed-off-by: Arthur <learthurgo@gmail.com> * finished app.py API made minute changes to parentheses use in various packages Signed-off-by: Arthur Lu <learthurgo@gmail.com> * bug fixes in app.py Signed-off-by: Arthur Lu <learthurgo@gmail.com> * bug fixes in app.py Signed-off-by: Arthur Lu <learthurgo@gmail.com> * made changes to .gitignore Signed-off-by: Arthur Lu <learthurgo@gmail.com> * made changes to .gitignore Signed-off-by: Arthur Lu <learthurgo@gmail.com> * deleted a __pycache__ folder from metrics Signed-off-by: Arthur Lu <learthurgo@gmail.com> * more changes to .gitignore Signed-off-by: Arthur Lu <learthurgo@gmail.com> * additions to app.py Signed-off-by: Arthur Lu <learthurgo@gmail.com> * renamed app.py to api.py Signed-off-by: Arthur Lu <learthurgo@gmail.com> * removed extranneous files Signed-off-by: Arthur Lu <learthurgo@gmail.com> * renamed api.py to tra.py removed rest api calls from tra.py * renamed api.py to tra.py removed rest api calls from tra.py Signed-off-by: Arthur Lu <learthurgo@gmail.com> * removed flask import from tra.py Signed-off-by: Arthur Lu <learthurgo@gmail.com> * changes to devcontainer.json Signed-off-by: Arthur Lu <learthurgo@gmail.com> * fixed unit tests to be correct removed some tests regressions because of potential function overflow removed trueskill unit test because of slight deviation chance Signed-off-by: Arthur Lu <learthurgo@gmail.com>
2020-05-20 13:52:38 +00:00
return {"mean": _mean, "median": _median, "standard-deviation": _stdev, "variance": _variance, "minimum": _min, "maximum": _max}
2020-03-10 03:58:51 +00:00
# returns z score with inputs of point, mean and standard deviation of spread
@jit(forceobj=True)
def z_score(point, mean, stdev):
score = (point - mean) / stdev
return score
2020-03-10 03:58:51 +00:00
# expects 2d array, normalizes across all axes
@jit(forceobj=True)
def z_normalize(array, *args):
2020-05-02 04:14:19 +00:00
array = np.array(array)
for arg in args:
array = sklearn.preprocessing.normalize(array, axis = arg)
2020-03-10 03:58:51 +00:00
2020-05-02 04:16:32 +00:00
return array
2020-03-10 03:58:51 +00:00
@jit(forceobj=True)
# expects 2d array of [x,y]
def histo_analysis(hist_data):
Merge service-dev changes with master (#24) * added config.json removed old config files Signed-off-by: Arthur <learthurgo@gmail.com> * superscript.py v 0.0.6.000 Signed-off-by: Arthur <learthurgo@gmail.com> * changed data.py Signed-off-by: Arthur <learthurgo@gmail.com> * changes to config.json Signed-off-by: Arthur <learthurgo@gmail.com> * removed cells from visualize_pit.py Signed-off-by: Arthur <learthurgo@gmail.com> * more changes to visualize_pit.py Signed-off-by: Arthur <learthurgo@gmail.com> * added analysis-master/metrics/__pycache__ to git ignore moved pit configs in config.json to the borrom superscript.py v 0.0.6.001 Signed-off-by: Arthur <learthurgo@gmail.com> * removed old database key Signed-off-by: Arthur <learthurgo@gmail.com> * adjusted config files Signed-off-by: Arthur <learthurgo@gmail.com> * Delete config-pop.json * fixed .gitignore Signed-off-by: Arthur <learthurgo@gmail.com> * analysis.py 1.2.1.003 added team kv pair to config.json Signed-off-by: Arthur <learthurgo@gmail.com> * superscript.py v 0.0.6.002 Signed-off-by: Arthur <learthurgo@gmail.com> * finished app.py API made minute changes to parentheses use in various packages Signed-off-by: Arthur Lu <learthurgo@gmail.com> * bug fixes in app.py Signed-off-by: Arthur Lu <learthurgo@gmail.com> * bug fixes in app.py Signed-off-by: Arthur Lu <learthurgo@gmail.com> * made changes to .gitignore Signed-off-by: Arthur Lu <learthurgo@gmail.com> * made changes to .gitignore Signed-off-by: Arthur Lu <learthurgo@gmail.com> * deleted a __pycache__ folder from metrics Signed-off-by: Arthur Lu <learthurgo@gmail.com> * more changes to .gitignore Signed-off-by: Arthur Lu <learthurgo@gmail.com> * additions to app.py Signed-off-by: Arthur Lu <learthurgo@gmail.com> * renamed app.py to api.py Signed-off-by: Arthur Lu <learthurgo@gmail.com> * removed extranneous files Signed-off-by: Arthur Lu <learthurgo@gmail.com> * renamed api.py to tra.py removed rest api calls from tra.py * renamed api.py to tra.py removed rest api calls from tra.py Signed-off-by: Arthur Lu <learthurgo@gmail.com> * removed flask import from tra.py Signed-off-by: Arthur Lu <learthurgo@gmail.com> * changes to devcontainer.json Signed-off-by: Arthur Lu <learthurgo@gmail.com> * fixed unit tests to be correct removed some tests regressions because of potential function overflow removed trueskill unit test because of slight deviation chance Signed-off-by: Arthur Lu <learthurgo@gmail.com>
2020-05-20 13:52:38 +00:00
if len(hist_data[0]) > 2:
2020-03-10 03:58:51 +00:00
hist_data = np.array(hist_data)
derivative = np.array(len(hist_data) - 1, dtype = float)
t = np.diff(hist_data)
derivative = t[1] / t[0]
np.sort(derivative)
2020-03-10 03:58:51 +00:00
Merge service-dev changes with master (#24) * added config.json removed old config files Signed-off-by: Arthur <learthurgo@gmail.com> * superscript.py v 0.0.6.000 Signed-off-by: Arthur <learthurgo@gmail.com> * changed data.py Signed-off-by: Arthur <learthurgo@gmail.com> * changes to config.json Signed-off-by: Arthur <learthurgo@gmail.com> * removed cells from visualize_pit.py Signed-off-by: Arthur <learthurgo@gmail.com> * more changes to visualize_pit.py Signed-off-by: Arthur <learthurgo@gmail.com> * added analysis-master/metrics/__pycache__ to git ignore moved pit configs in config.json to the borrom superscript.py v 0.0.6.001 Signed-off-by: Arthur <learthurgo@gmail.com> * removed old database key Signed-off-by: Arthur <learthurgo@gmail.com> * adjusted config files Signed-off-by: Arthur <learthurgo@gmail.com> * Delete config-pop.json * fixed .gitignore Signed-off-by: Arthur <learthurgo@gmail.com> * analysis.py 1.2.1.003 added team kv pair to config.json Signed-off-by: Arthur <learthurgo@gmail.com> * superscript.py v 0.0.6.002 Signed-off-by: Arthur <learthurgo@gmail.com> * finished app.py API made minute changes to parentheses use in various packages Signed-off-by: Arthur Lu <learthurgo@gmail.com> * bug fixes in app.py Signed-off-by: Arthur Lu <learthurgo@gmail.com> * bug fixes in app.py Signed-off-by: Arthur Lu <learthurgo@gmail.com> * made changes to .gitignore Signed-off-by: Arthur Lu <learthurgo@gmail.com> * made changes to .gitignore Signed-off-by: Arthur Lu <learthurgo@gmail.com> * deleted a __pycache__ folder from metrics Signed-off-by: Arthur Lu <learthurgo@gmail.com> * more changes to .gitignore Signed-off-by: Arthur Lu <learthurgo@gmail.com> * additions to app.py Signed-off-by: Arthur Lu <learthurgo@gmail.com> * renamed app.py to api.py Signed-off-by: Arthur Lu <learthurgo@gmail.com> * removed extranneous files Signed-off-by: Arthur Lu <learthurgo@gmail.com> * renamed api.py to tra.py removed rest api calls from tra.py * renamed api.py to tra.py removed rest api calls from tra.py Signed-off-by: Arthur Lu <learthurgo@gmail.com> * removed flask import from tra.py Signed-off-by: Arthur Lu <learthurgo@gmail.com> * changes to devcontainer.json Signed-off-by: Arthur Lu <learthurgo@gmail.com> * fixed unit tests to be correct removed some tests regressions because of potential function overflow removed trueskill unit test because of slight deviation chance Signed-off-by: Arthur Lu <learthurgo@gmail.com>
2020-05-20 13:52:38 +00:00
return {"mean": basic_stats(derivative)["mean"], "deviation": basic_stats(derivative)["standard-deviation"]}
2020-03-10 03:58:51 +00:00
else:
2020-03-10 03:58:51 +00:00
return None
2020-03-10 03:58:51 +00:00
def regression(inputs, outputs, args): # inputs, outputs expects N-D array
X = np.array(inputs)
y = np.array(outputs)
2020-03-10 03:58:51 +00:00
regressions = []
2020-03-10 03:58:51 +00:00
if 'lin' in args: # formula: ax + b
2020-03-10 03:58:51 +00:00
try:
2020-03-10 03:58:51 +00:00
2020-05-04 16:59:25 +00:00
def lin(x, a, b):
2020-03-10 03:58:51 +00:00
return a * x + b
2020-03-10 03:58:51 +00:00
2020-05-04 16:59:25 +00:00
popt, pcov = scipy.optimize.curve_fit(lin, X, y)
2020-03-10 03:58:51 +00:00
coeffs = popt.flatten().tolist()
regressions.append(str(coeffs[0]) + "*x+" + str(coeffs[1]))
2020-03-10 03:58:51 +00:00
except Exception as e:
2020-03-10 03:58:51 +00:00
pass
2020-03-10 03:58:51 +00:00
if 'log' in args: # formula: a log (b(x + c)) + d
2020-03-10 03:58:51 +00:00
try:
2020-03-10 03:58:51 +00:00
2020-05-04 16:59:25 +00:00
def log(x, a, b, c, d):
2020-03-10 03:58:51 +00:00
return a * np.log(b*(x + c)) + d
2020-03-10 03:58:51 +00:00
2020-05-04 16:59:25 +00:00
popt, pcov = scipy.optimize.curve_fit(log, X, y)
2020-03-10 03:58:51 +00:00
coeffs = popt.flatten().tolist()
regressions.append(str(coeffs[0]) + "*log(" + str(coeffs[1]) + "*(x+" + str(coeffs[2]) + "))+" + str(coeffs[3]))
2020-03-10 03:58:51 +00:00
except Exception as e:
pass
2020-03-10 03:58:51 +00:00
if 'exp' in args: # formula: a e ^ (b(x + c)) + d
2020-03-10 03:58:51 +00:00
try:
2020-03-10 03:58:51 +00:00
2020-05-04 16:59:25 +00:00
def exp(x, a, b, c, d):
2020-03-10 03:58:51 +00:00
return a * np.exp(b*(x + c)) + d
2020-03-10 03:58:51 +00:00
2020-05-04 16:59:25 +00:00
popt, pcov = scipy.optimize.curve_fit(exp, X, y)
2020-03-10 03:58:51 +00:00
coeffs = popt.flatten().tolist()
regressions.append(str(coeffs[0]) + "*e^(" + str(coeffs[1]) + "*(x+" + str(coeffs[2]) + "))+" + str(coeffs[3]))
2020-03-10 03:58:51 +00:00
except Exception as e:
2020-03-10 03:58:51 +00:00
pass
2020-03-10 03:58:51 +00:00
if 'ply' in args: # formula: a + bx^1 + cx^2 + dx^3 + ...
inputs = np.array([inputs])
outputs = np.array([outputs])
2020-03-10 03:58:51 +00:00
plys = []
limit = len(outputs[0])
2020-03-10 03:58:51 +00:00
for i in range(2, limit):
2020-03-10 03:58:51 +00:00
model = sklearn.preprocessing.PolynomialFeatures(degree = i)
model = sklearn.pipeline.make_pipeline(model, sklearn.linear_model.LinearRegression())
model = model.fit(np.rot90(inputs), np.rot90(outputs))
2020-03-10 03:58:51 +00:00
params = model.steps[1][1].intercept_.tolist()
params = np.append(params, model.steps[1][1].coef_[0].tolist()[1::])
params = params.flatten().tolist()
temp = ""
counter = 0
for param in params:
temp += "(" + str(param) + "*x^" + str(counter) + ")"
counter += 1
plys.append(temp)
2020-03-10 03:58:51 +00:00
regressions.append(plys)
2020-03-10 03:58:51 +00:00
if 'sig' in args: # formula: a tanh (b(x + c)) + d
2020-03-10 03:58:51 +00:00
try:
2020-03-10 03:58:51 +00:00
2020-05-04 16:59:25 +00:00
def sig(x, a, b, c, d):
2020-03-10 03:58:51 +00:00
return a * np.tanh(b*(x + c)) + d
2020-03-10 03:58:51 +00:00
2020-05-04 16:59:25 +00:00
popt, pcov = scipy.optimize.curve_fit(sig, X, y)
2020-03-10 03:58:51 +00:00
coeffs = popt.flatten().tolist()
regressions.append(str(coeffs[0]) + "*tanh(" + str(coeffs[1]) + "*(x+" + str(coeffs[2]) + "))+" + str(coeffs[3]))
2020-03-10 03:58:51 +00:00
except Exception as e:
pass
2020-03-10 03:58:51 +00:00
return regressions
2020-03-10 03:58:51 +00:00
2020-05-02 03:59:54 +00:00
class Metric:
2020-03-10 03:58:51 +00:00
def elo(self, starting_score, opposing_score, observed, N, K):
2020-03-10 03:58:51 +00:00
return Elo.calculate(starting_score, opposing_score, observed, N, K)
2020-03-10 03:58:51 +00:00
def glicko2(self, starting_score, starting_rd, starting_vol, opposing_score, opposing_rd, observations):
2020-03-10 03:58:51 +00:00
player = Glicko2.Glicko2(rating = starting_score, rd = starting_rd, vol = starting_vol)
2020-03-10 03:58:51 +00:00
player.update_player([x for x in opposing_score], [x for x in opposing_rd], observations)
2020-03-10 03:58:51 +00:00
return (player.rating, player.rd, player.vol)
2020-03-10 03:58:51 +00:00
def trueskill(self, teams_data, observations): # teams_data is array of array of tuples ie. [[(mu, sigma), (mu, sigma), (mu, sigma)], [(mu, sigma), (mu, sigma), (mu, sigma)]]
2020-03-10 03:58:51 +00:00
team_ratings = []
2020-03-10 03:58:51 +00:00
for team in teams_data:
team_temp = ()
for player in team:
player = Trueskill.Rating(player[0], player[1])
team_temp = team_temp + (player,)
team_ratings.append(team_temp)
return Trueskill.rate(team_ratings, ranks=observations)
2020-03-10 03:58:51 +00:00
2020-05-02 03:59:54 +00:00
class RegressionMetric():
2020-03-10 03:58:51 +00:00
def __new__(cls, predictions, targets):
2020-03-10 03:58:51 +00:00
return cls.r_squared(cls, predictions, targets), cls.mse(cls, predictions, targets), cls.rms(cls, predictions, targets)
2020-03-10 03:58:51 +00:00
def r_squared(self, predictions, targets): # assumes equal size inputs
2020-03-10 03:58:51 +00:00
return sklearn.metrics.r2_score(targets, predictions)
2020-03-10 03:58:51 +00:00
def mse(self, predictions, targets):
2020-03-10 03:58:51 +00:00
return sklearn.metrics.mean_squared_error(targets, predictions)
2020-03-10 03:58:51 +00:00
def rms(self, predictions, targets):
2020-03-10 03:58:51 +00:00
return math.sqrt(sklearn.metrics.mean_squared_error(targets, predictions))
2020-03-10 03:58:51 +00:00
2020-05-02 03:59:54 +00:00
class ClassificationMetric():
2020-03-10 03:58:51 +00:00
def __new__(cls, predictions, targets):
2020-03-10 03:58:51 +00:00
return cls.cm(cls, predictions, targets), cls.cr(cls, predictions, targets)
2020-03-10 03:58:51 +00:00
def cm(self, predictions, targets):
2020-03-10 03:58:51 +00:00
return sklearn.metrics.confusion_matrix(targets, predictions)
2020-03-10 03:58:51 +00:00
def cr(self, predictions, targets):
2020-03-10 03:58:51 +00:00
return sklearn.metrics.classification_report(targets, predictions)
2020-03-10 03:58:51 +00:00
@jit(nopython=True)
def mean(data):
return np.mean(data)
2020-03-10 03:58:51 +00:00
@jit(nopython=True)
def median(data):
return np.median(data)
2020-03-10 03:58:51 +00:00
@jit(nopython=True)
def stdev(data):
return np.std(data)
2020-03-10 03:58:51 +00:00
@jit(nopython=True)
def variance(data):
return np.var(data)
2020-03-10 03:58:51 +00:00
@jit(nopython=True)
def npmin(data):
return np.amin(data)
2020-03-10 03:58:51 +00:00
@jit(nopython=True)
def npmax(data):
return np.amax(data)
2020-03-10 03:58:51 +00:00
@jit(forceobj=True)
def kmeans(data, n_clusters=8, init="k-means++", n_init=10, max_iter=300, tol=0.0001, precompute_distances="auto", verbose=0, random_state=None, copy_x=True, n_jobs=None, algorithm="auto"):
kernel = sklearn.cluster.KMeans(n_clusters = n_clusters, init = init, n_init = n_init, max_iter = max_iter, tol = tol, precompute_distances = precompute_distances, verbose = verbose, random_state = random_state, copy_x = copy_x, n_jobs = n_jobs, algorithm = algorithm)
kernel.fit(data)
predictions = kernel.predict(data)
centers = kernel.cluster_centers_
2020-03-10 03:58:51 +00:00
return centers, predictions
2020-03-10 03:58:51 +00:00
@jit(forceobj=True)
def pca(data, n_components = None, copy = True, whiten = False, svd_solver = "auto", tol = 0.0, iterated_power = "auto", random_state = None):
kernel = sklearn.decomposition.PCA(n_components = n_components, copy = copy, whiten = whiten, svd_solver = svd_solver, tol = tol, iterated_power = iterated_power, random_state = random_state)
2020-03-10 03:58:51 +00:00
return kernel.fit_transform(data)
2020-03-10 03:58:51 +00:00
@jit(forceobj=True)
def decisiontree(data, labels, test_size = 0.3, criterion = "gini", splitter = "default", max_depth = None): #expects *2d data and 1d labels
data_train, data_test, labels_train, labels_test = sklearn.model_selection.train_test_split(data, labels, test_size=test_size, random_state=1)
model = sklearn.tree.DecisionTreeClassifier(criterion = criterion, splitter = splitter, max_depth = max_depth)
model = model.fit(data_train,labels_train)
predictions = model.predict(data_test)
2020-05-02 03:59:54 +00:00
metrics = ClassificationMetric(predictions, labels_test)
2020-03-10 03:58:51 +00:00
return model, metrics
2020-03-10 03:58:51 +00:00
class KNN:
2020-03-10 03:58:51 +00:00
def knn_classifier(self, data, labels, test_size = 0.3, algorithm='auto', leaf_size=30, metric='minkowski', metric_params=None, n_jobs=None, n_neighbors=5, p=2, weights='uniform'): #expects *2d data and 1d labels post-scaling
2020-03-10 03:58:51 +00:00
data_train, data_test, labels_train, labels_test = sklearn.model_selection.train_test_split(data, labels, test_size=test_size, random_state=1)
model = sklearn.neighbors.KNeighborsClassifier()
model.fit(data_train, labels_train)
predictions = model.predict(data_test)
2020-03-10 03:58:51 +00:00
2020-05-02 03:59:54 +00:00
return model, ClassificationMetric(predictions, labels_test)
2020-03-10 03:58:51 +00:00
def knn_regressor(self, data, outputs, test_size, n_neighbors = 5, weights = "uniform", algorithm = "auto", leaf_size = 30, p = 2, metric = "minkowski", metric_params = None, n_jobs = None):
data_train, data_test, outputs_train, outputs_test = sklearn.model_selection.train_test_split(data, outputs, test_size=test_size, random_state=1)
model = sklearn.neighbors.KNeighborsRegressor(n_neighbors = n_neighbors, weights = weights, algorithm = algorithm, leaf_size = leaf_size, p = p, metric = metric, metric_params = metric_params, n_jobs = n_jobs)
model.fit(data_train, outputs_train)
predictions = model.predict(data_test)
2020-03-10 03:58:51 +00:00
2020-05-02 03:59:54 +00:00
return model, RegressionMetric(predictions, outputs_test)
2020-03-10 03:58:51 +00:00
class NaiveBayes:
def guassian(self, data, labels, test_size = 0.3, priors = None, var_smoothing = 1e-09):
2020-03-10 03:58:51 +00:00
data_train, data_test, labels_train, labels_test = sklearn.model_selection.train_test_split(data, labels, test_size=test_size, random_state=1)
model = sklearn.naive_bayes.GaussianNB(priors = priors, var_smoothing = var_smoothing)
model.fit(data_train, labels_train)
predictions = model.predict(data_test)
2020-03-10 03:58:51 +00:00
2020-05-02 03:59:54 +00:00
return model, ClassificationMetric(predictions, labels_test)
2020-03-10 03:58:51 +00:00
def multinomial(self, data, labels, test_size = 0.3, alpha=1.0, fit_prior=True, class_prior=None):
2020-03-10 03:58:51 +00:00
data_train, data_test, labels_train, labels_test = sklearn.model_selection.train_test_split(data, labels, test_size=test_size, random_state=1)
model = sklearn.naive_bayes.MultinomialNB(alpha = alpha, fit_prior = fit_prior, class_prior = class_prior)
model.fit(data_train, labels_train)
predictions = model.predict(data_test)
2020-03-10 03:58:51 +00:00
2020-05-02 03:59:54 +00:00
return model, ClassificationMetric(predictions, labels_test)
2020-03-10 03:58:51 +00:00
def bernoulli(self, data, labels, test_size = 0.3, alpha=1.0, binarize=0.0, fit_prior=True, class_prior=None):
2020-03-10 03:58:51 +00:00
data_train, data_test, labels_train, labels_test = sklearn.model_selection.train_test_split(data, labels, test_size=test_size, random_state=1)
model = sklearn.naive_bayes.BernoulliNB(alpha = alpha, binarize = binarize, fit_prior = fit_prior, class_prior = class_prior)
model.fit(data_train, labels_train)
predictions = model.predict(data_test)
2020-03-10 03:58:51 +00:00
2020-05-02 03:59:54 +00:00
return model, ClassificationMetric(predictions, labels_test)
2020-03-10 03:58:51 +00:00
def complement(self, data, labels, test_size = 0.3, alpha=1.0, fit_prior=True, class_prior=None, norm=False):
2020-03-10 03:58:51 +00:00
data_train, data_test, labels_train, labels_test = sklearn.model_selection.train_test_split(data, labels, test_size=test_size, random_state=1)
model = sklearn.naive_bayes.ComplementNB(alpha = alpha, fit_prior = fit_prior, class_prior = class_prior, norm = norm)
model.fit(data_train, labels_train)
predictions = model.predict(data_test)
2020-03-10 03:58:51 +00:00
2020-05-02 03:59:54 +00:00
return model, ClassificationMetric(predictions, labels_test)
2020-03-10 03:58:51 +00:00
class SVM:
class CustomKernel:
2020-03-10 03:58:51 +00:00
def __new__(cls, C, kernel, degre, gamma, coef0, shrinking, probability, tol, cache_size, class_weight, verbose, max_iter, decision_function_shape, random_state):
2020-03-10 03:58:51 +00:00
return sklearn.svm.SVC(C = C, kernel = kernel, gamma = gamma, coef0 = coef0, shrinking = shrinking, probability = probability, tol = tol, cache_size = cache_size, class_weight = class_weight, verbose = verbose, max_iter = max_iter, decision_function_shape = decision_function_shape, random_state = random_state)
2020-03-10 03:58:51 +00:00
class StandardKernel:
2020-03-10 03:58:51 +00:00
def __new__(cls, kernel, C=1.0, degree=3, gamma='auto_deprecated', coef0=0.0, shrinking=True, probability=False, tol=0.001, cache_size=200, class_weight=None, verbose=False, max_iter=-1, decision_function_shape='ovr', random_state=None):
2020-03-10 03:58:51 +00:00
return sklearn.svm.SVC(C = C, kernel = kernel, gamma = gamma, coef0 = coef0, shrinking = shrinking, probability = probability, tol = tol, cache_size = cache_size, class_weight = class_weight, verbose = verbose, max_iter = max_iter, decision_function_shape = decision_function_shape, random_state = random_state)
2020-03-10 03:58:51 +00:00
class PrebuiltKernel:
2020-03-10 03:58:51 +00:00
class Linear:
2020-03-10 03:58:51 +00:00
def __new__(cls):
2020-03-10 03:58:51 +00:00
return sklearn.svm.SVC(kernel = 'linear')
2020-03-10 03:58:51 +00:00
class Polynomial:
2020-03-10 03:58:51 +00:00
def __new__(cls, power, r_bias):
2020-03-10 03:58:51 +00:00
return sklearn.svm.SVC(kernel = 'polynomial', degree = power, coef0 = r_bias)
2020-03-10 03:58:51 +00:00
class RBF:
2020-03-10 03:58:51 +00:00
def __new__(cls, gamma):
2020-03-10 03:58:51 +00:00
return sklearn.svm.SVC(kernel = 'rbf', gamma = gamma)
2020-03-10 03:58:51 +00:00
class Sigmoid:
2020-03-10 03:58:51 +00:00
def __new__(cls, r_bias):
2020-03-10 03:58:51 +00:00
return sklearn.svm.SVC(kernel = 'sigmoid', coef0 = r_bias)
2020-03-10 03:58:51 +00:00
def fit(self, kernel, train_data, train_outputs): # expects *2d data, 1d labels or outputs
2020-03-10 03:58:51 +00:00
return kernel.fit(train_data, train_outputs)
2020-03-10 03:58:51 +00:00
def eval_classification(self, kernel, test_data, test_outputs):
2020-03-10 03:58:51 +00:00
predictions = kernel.predict(test_data)
2020-03-10 03:58:51 +00:00
2020-05-02 03:59:54 +00:00
return ClassificationMetric(predictions, test_outputs)
2020-03-10 03:58:51 +00:00
def eval_regression(self, kernel, test_data, test_outputs):
2020-03-10 03:58:51 +00:00
predictions = kernel.predict(test_data)
2020-03-10 03:58:51 +00:00
2020-05-02 03:59:54 +00:00
return RegressionMetric(predictions, test_outputs)
2020-03-10 03:58:51 +00:00
2020-05-02 03:59:54 +00:00
class RandomForrest:
2020-03-10 03:58:51 +00:00
2020-05-02 03:59:54 +00:00
def random_forest_classifier(self, data, labels, test_size, n_estimators="warn", criterion="gini", max_depth=None, min_samples_split=2, min_samples_leaf=1, min_weight_fraction_leaf=0.0, max_features="auto", max_leaf_nodes=None, min_impurity_decrease=0.0, min_impurity_split=None, bootstrap=True, oob_score=False, n_jobs=None, random_state=None, verbose=0, warm_start=False, class_weight=None):
data_train, data_test, labels_train, labels_test = sklearn.model_selection.train_test_split(data, labels, test_size=test_size, random_state=1)
kernel = sklearn.ensemble.RandomForestClassifier(n_estimators = n_estimators, criterion = criterion, max_depth = max_depth, min_samples_split = min_samples_split, min_samples_leaf = min_samples_leaf, min_weight_fraction_leaf = min_weight_fraction_leaf, max_leaf_nodes = max_leaf_nodes, min_impurity_decrease = min_impurity_decrease, bootstrap = bootstrap, oob_score = oob_score, n_jobs = n_jobs, random_state = random_state, verbose = verbose, warm_start = warm_start, class_weight = class_weight)
kernel.fit(data_train, labels_train)
predictions = kernel.predict(data_test)
2020-03-10 03:58:51 +00:00
2020-05-02 03:59:54 +00:00
return kernel, ClassificationMetric(predictions, labels_test)
2020-03-10 03:58:51 +00:00
2020-05-02 03:59:54 +00:00
def random_forest_regressor(self, data, outputs, test_size, n_estimators="warn", criterion="mse", max_depth=None, min_samples_split=2, min_samples_leaf=1, min_weight_fraction_leaf=0.0, max_features="auto", max_leaf_nodes=None, min_impurity_decrease=0.0, min_impurity_split=None, bootstrap=True, oob_score=False, n_jobs=None, random_state=None, verbose=0, warm_start=False):
2020-03-10 03:58:51 +00:00
2020-05-02 03:59:54 +00:00
data_train, data_test, outputs_train, outputs_test = sklearn.model_selection.train_test_split(data, outputs, test_size=test_size, random_state=1)
kernel = sklearn.ensemble.RandomForestRegressor(n_estimators = n_estimators, criterion = criterion, max_depth = max_depth, min_samples_split = min_samples_split, min_weight_fraction_leaf = min_weight_fraction_leaf, max_features = max_features, max_leaf_nodes = max_leaf_nodes, min_impurity_decrease = min_impurity_decrease, min_impurity_split = min_impurity_split, bootstrap = bootstrap, oob_score = oob_score, n_jobs = n_jobs, random_state = random_state, verbose = verbose, warm_start = warm_start)
kernel.fit(data_train, outputs_train)
predictions = kernel.predict(data_test)
2020-03-10 03:58:51 +00:00
2020-05-02 03:59:54 +00:00
return kernel, RegressionMetric(predictions, outputs_test)
2020-04-13 19:58:04 +00:00
2020-05-02 03:59:54 +00:00
class CorrelationTest:
2020-04-13 19:58:04 +00:00
def anova_oneway(self, *args): #expects arrays of samples
2020-04-13 19:58:04 +00:00
results = scipy.stats.f_oneway(*args)
return {"F-value": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def pearson(self, x, y):
2020-04-13 19:58:04 +00:00
results = scipy.stats.pearsonr(x, y)
return {"r-value": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def spearman(self, a, b = None, axis = 0, nan_policy = 'propagate'):
2020-04-13 19:58:04 +00:00
results = scipy.stats.spearmanr(a, b = b, axis = axis, nan_policy = nan_policy)
return {"r-value": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def point_biserial(self, x,y):
2020-04-13 19:58:04 +00:00
results = scipy.stats.pointbiserialr(x, y)
return {"r-value": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def kendall(self, x, y, initial_lexsort = None, nan_policy = 'propagate', method = 'auto'):
2020-04-13 19:58:04 +00:00
results = scipy.stats.kendalltau(x, y, initial_lexsort = initial_lexsort, nan_policy = nan_policy, method = method)
return {"tau": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def kendall_weighted(self, x, y, rank = True, weigher = None, additive = True):
2020-04-13 19:58:04 +00:00
results = scipy.stats.weightedtau(x, y, rank = rank, weigher = weigher, additive = additive)
return {"tau": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def mgc(self, x, y, compute_distance = None, reps = 1000, workers = 1, is_twosamp = False, random_state = None):
2020-04-13 19:58:04 +00:00
results = scipy.stats.multiscale_graphcorr(x, y, compute_distance = compute_distance, reps = reps, workers = workers, is_twosamp = is_twosamp, random_state = random_state)
return {"k-value": results[0], "p-value": results[1], "data": results[2]} # unsure if MGC test returns a k value
2020-04-13 19:58:04 +00:00
2020-05-02 03:59:54 +00:00
class StatisticalTest:
2020-04-13 19:58:04 +00:00
def ttest_onesample(self, a, popmean, axis = 0, nan_policy = 'propagate'):
2020-04-13 19:58:04 +00:00
results = scipy.stats.ttest_1samp(a, popmean, axis = axis, nan_policy = nan_policy)
return {"t-value": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def ttest_independent(self, a, b, equal = True, nan_policy = 'propagate'):
2020-04-13 19:58:04 +00:00
results = scipy.stats.ttest_ind(a, b, equal_var = equal, nan_policy = nan_policy)
return {"t-value": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def ttest_statistic(self, o1, o2, equal = True):
2020-04-13 19:58:04 +00:00
results = scipy.stats.ttest_ind_from_stats(o1["mean"], o1["std"], o1["nobs"], o2["mean"], o2["std"], o2["nobs"], equal_var = equal)
return {"t-value": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def ttest_related(self, a, b, axis = 0, nan_policy='propagate'):
2020-04-13 19:58:04 +00:00
results = scipy.stats.ttest_rel(a, b, axis = axis, nan_policy = nan_policy)
return {"t-value": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def ks_fitness(self, rvs, cdf, args = (), N = 20, alternative = 'two-sided', mode = 'approx'):
2020-04-13 19:58:04 +00:00
results = scipy.stats.kstest(rvs, cdf, args = args, N = N, alternative = alternative, mode = mode)
return {"ks-value": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def chisquare(self, f_obs, f_exp = None, ddof = None, axis = 0):
2020-04-13 19:58:04 +00:00
results = scipy.stats.chisquare(f_obs, f_exp = f_exp, ddof = ddof, axis = axis)
return {"chisquared-value": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def powerdivergence(self, f_obs, f_exp = None, ddof = None, axis = 0, lambda_ = None):
2020-04-13 19:58:04 +00:00
results = scipy.stats.power_divergence(f_obs, f_exp = f_exp, ddof = ddof, axis = axis, lambda_ = lambda_)
return {"powerdivergence-value": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def ks_twosample(self, x, y, alternative = 'two_sided', mode = 'auto'):
results = scipy.stats.ks_2samp(x, y, alternative = alternative, mode = mode)
return {"ks-value": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def es_twosample(self, x, y, t = (0.4, 0.8)):
2020-04-13 19:58:04 +00:00
results = scipy.stats.epps_singleton_2samp(x, y, t = t)
return {"es-value": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def mw_rank(self, x, y, use_continuity = True, alternative = None):
2020-04-13 19:58:04 +00:00
results = scipy.stats.mannwhitneyu(x, y, use_continuity = use_continuity, alternative = alternative)
return {"u-value": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def mw_tiecorrection(self, rank_values):
2020-04-13 19:58:04 +00:00
results = scipy.stats.tiecorrect(rank_values)
return {"correction-factor": results}
2020-04-13 19:58:04 +00:00
def rankdata(self, a, method = 'average'):
2020-04-13 19:58:04 +00:00
results = scipy.stats.rankdata(a, method = method)
return results
2020-04-13 19:58:04 +00:00
def wilcoxon_ranksum(self, a, b): # this seems to be superceded by Mann Whitney Wilcoxon U Test
2020-04-13 19:58:04 +00:00
results = scipy.stats.ranksums(a, b)
return {"u-value": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def wilcoxon_signedrank(self, x, y = None, zero_method = 'wilcox', correction = False, alternative = 'two-sided'):
2020-04-13 19:58:04 +00:00
results = scipy.stats.wilcoxon(x, y = y, zero_method = zero_method, correction = correction, alternative = alternative)
return {"t-value": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def kw_htest(self, *args, nan_policy = 'propagate'):
2020-04-13 19:58:04 +00:00
results = scipy.stats.kruskal(*args, nan_policy = nan_policy)
return {"h-value": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def friedman_chisquare(self, *args):
2020-04-13 19:58:04 +00:00
results = scipy.stats.friedmanchisquare(*args)
return {"chisquared-value": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def bm_wtest(self, x, y, alternative = 'two-sided', distribution = 't', nan_policy = 'propagate'):
2020-04-13 19:58:04 +00:00
results = scipy.stats.brunnermunzel(x, y, alternative = alternative, distribution = distribution, nan_policy = nan_policy)
return {"w-value": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def combine_pvalues(self, pvalues, method = 'fisher', weights = None):
2020-04-13 19:58:04 +00:00
results = scipy.stats.combine_pvalues(pvalues, method = method, weights = weights)
return {"combined-statistic": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def jb_fitness(self, x):
2020-04-13 19:58:04 +00:00
results = scipy.stats.jarque_bera(x)
return {"jb-value": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def ab_equality(self, x, y):
2020-04-13 19:58:04 +00:00
results = scipy.stats.ansari(x, y)
return {"ab-value": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def bartlett_variance(self, *args):
2020-04-13 19:58:04 +00:00
results = scipy.stats.bartlett(*args)
return {"t-value": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def levene_variance(self, *args, center = 'median', proportiontocut = 0.05):
2020-04-13 19:58:04 +00:00
results = scipy.stats.levene(*args, center = center, proportiontocut = proportiontocut)
return {"w-value": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def sw_normality(self, x):
2020-04-13 19:58:04 +00:00
results = scipy.stats.shapiro(x)
return {"w-value": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def shapiro(self, x):
2020-04-13 19:58:04 +00:00
return "destroyed by facts and logic"
2020-04-13 19:58:04 +00:00
def ad_onesample(self, x, dist = 'norm'):
2020-04-13 19:58:04 +00:00
results = scipy.stats.anderson(x, dist = dist)
return {"d-value": results[0], "critical-values": results[1], "significance-value": results[2]}
def ad_ksample(self, samples, midrank = True):
2020-04-13 19:58:04 +00:00
results = scipy.stats.anderson_ksamp(samples, midrank = midrank)
return {"d-value": results[0], "critical-values": results[1], "significance-value": results[2]}
2020-04-13 19:58:04 +00:00
def binomial(self, x, n = None, p = 0.5, alternative = 'two-sided'):
2020-04-13 19:58:04 +00:00
results = scipy.stats.binom_test(x, n = n, p = p, alternative = alternative)
return {"p-value": results}
2020-04-13 19:58:04 +00:00
def fk_variance(self, *args, center = 'median', proportiontocut = 0.05):
2020-04-13 19:58:04 +00:00
results = scipy.stats.fligner(*args, center = center, proportiontocut = proportiontocut)
return {"h-value": results[0], "p-value": results[1]} # unknown if the statistic is an h value
2020-04-13 19:58:04 +00:00
def mood_mediantest(self, *args, ties = 'below', correction = True, lambda_ = 1, nan_policy = 'propagate'):
2020-04-13 19:58:04 +00:00
results = scipy.stats.median_test(*args, ties = ties, correction = correction, lambda_ = lambda_, nan_policy = nan_policy)
return {"chisquared-value": results[0], "p-value": results[1], "m-value": results[2], "table": results[3]}
2020-04-13 19:58:04 +00:00
def mood_equalscale(self, x, y, axis = 0):
2020-04-13 19:58:04 +00:00
results = scipy.stats.mood(x, y, axis = axis)
return {"z-score": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def skewtest(self, a, axis = 0, nan_policy = 'propogate'):
2020-04-13 19:58:04 +00:00
results = scipy.stats.skewtest(a, axis = axis, nan_policy = nan_policy)
return {"z-score": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def kurtosistest(self, a, axis = 0, nan_policy = 'propogate'):
2020-04-13 19:58:04 +00:00
results = scipy.stats.kurtosistest(a, axis = axis, nan_policy = nan_policy)
return {"z-score": results[0], "p-value": results[1]}
2020-04-13 19:58:04 +00:00
def normaltest(self, a, axis = 0, nan_policy = 'propogate'):
2020-04-13 19:58:04 +00:00
results = scipy.stats.normaltest(a, axis = axis, nan_policy = nan_policy)
2020-05-04 19:50:36 +00:00
return {"z-score": results[0], "p-value": results[1]}
class Array(): # tests on nd arrays independent of basic_stats
2020-05-04 19:50:36 +00:00
def elementwise_mean(self, *args): # expects arrays that are size normalized
return np.mean([*args], axis = 0)
def elementwise_median(self, *args):
return np.median([*args], axis = 0)
def elementwise_stdev(self, *args):
return np.std([*args], axis = 0)
def elementwise_variance(self, *args):
return np.var([*args], axis = 0)
def elementwise_npmin(self, *args):
return np.amin([*args], axis = 0)
def elementwise_npmax(self, *args):
return np.amax([*args], axis = 0)
def elementwise_stats(self, *args):
_mean = self.elementwise_mean(*args)
_median = self.elementwise_median(*args)
_stdev = self.elementwise_stdev(*args)
_variance = self.elementwise_variance(*args)
_min = self.elementwise_npmin(*args)
_max = self.elementwise_npmax(*args)
2020-05-13 01:19:58 +00:00
return _mean, _median, _stdev, _variance, _min, _max
def normalize(self, array):
a = np.atleast_1d(np.linalg.norm(array))
a[a==0] = 1
2020-05-13 01:39:23 +00:00
return array / np.expand_dims(a, -1)
2020-05-13 01:19:58 +00:00
def add(self, *args):
temp = np.array([])
2020-05-13 01:21:11 +00:00
for a in args:
2020-05-13 01:19:58 +00:00
temp += a
return temp
def mul(self, *args):
temp = np.array([])
2020-05-13 01:21:11 +00:00
for a in args:
2020-05-13 01:19:58 +00:00
temp *= a
return temp
def neg(self, array):
return -array
def inv(self, array):
return 1/array
def dot(self, a, b):
return np.dot(a, b)
def cross(self, a, b):
return np.cross(a, b)
def sort(self, array): # depreciated
warnings.warn("Array.sort has been depreciated in favor of Sort")
array_length = len(array)
if array_length <= 1:
return array
middle_index = int(array_length / 2)
left = array[0:middle_index]
right = array[middle_index:]
left = self.sort(left)
right = self.sort(right)
return self.__merge(left, right)
def __merge(self, left, right):
sorted_list = []
left = left[:]
right = right[:]
while len(left) > 0 or len(right) > 0:
if len(left) > 0 and len(right) > 0:
if left[0] <= right[0]:
sorted_list.append(left.pop(0))
else:
sorted_list.append(right.pop(0))
elif len(left) > 0:
sorted_list.append(left.pop(0))
elif len(right) > 0:
sorted_list.append(right.pop(0))
return sorted_list
def search(self, arr, x):
return self.__search(arr, 0, len(arr) - 1, x)
def __search(self, arr, low, high, x):
if high >= low:
mid = (high + low) // 2
if arr[mid] == x:
return mid
elif arr[mid] > x:
return binary_search(arr, low, mid - 1, x)
else:
return binary_search(arr, mid + 1, high, x)
else:
return -1
class Sort: # if you haven't used a sort, then you've never lived
def quicksort(self, a):
def sort(array):
less = []
equal = []
greater = []
if len(array) > 1:
pivot = array[0]
for x in array:
if x < pivot:
less.append(x)
elif x == pivot:
equal.append(x)
elif x > pivot:
greater.append(x)
return sort(less)+equal+sort(greater)
else:
return array
return np.array(sort(a))
def mergesort(self, a):
def sort(array):
array = array
if len(array) >1:
middle = len(array) // 2
L = array[:middle]
R = array[middle:]
sort(L)
sort(R)
i = j = k = 0
while i < len(L) and j < len(R):
if L[i] < R[j]:
array[k] = L[i]
i+= 1
else:
array[k] = R[j]
j+= 1
k+= 1
while i < len(L):
array[k] = L[i]
i+= 1
k+= 1
while j < len(R):
array[k] = R[j]
j+= 1
k+= 1
return array
return sort(a)
def introsort(self, a):
def sort(array, start, end, maxdepth):
array = array
if end - start <= 1:
return
elif maxdepth == 0:
heapsort(array, start, end)
else:
p = partition(array, start, end)
sort(array, start, p + 1, maxdepth - 1)
sort(array, p + 1, end, maxdepth - 1)
return array
def partition(array, start, end):
pivot = array[start]
i = start - 1
j = end
while True:
i = i + 1
while array[i] < pivot:
i = i + 1
j = j - 1
while array[j] > pivot:
j = j - 1
if i >= j:
return j
swap(array, i, j)
def swap(array, i, j):
array[i], array[j] = array[j], array[i]
def heapsort(array, start, end):
build_max_heap(array, start, end)
for i in range(end - 1, start, -1):
swap(array, start, i)
max_heapify(array, index=0, start=start, end=i)
def build_max_heap(array, start, end):
def parent(i):
return (i - 1)//2
length = end - start
index = parent(length - 1)
while index >= 0:
max_heapify(array, index, start, end)
index = index - 1
def max_heapify(array, index, start, end):
def left(i):
return 2*i + 1
def right(i):
return 2*i + 2
size = end - start
l = left(index)
r = right(index)
if (l < size and array[start + l] > array[start + index]):
largest = l
else:
largest = index
if (r < size and array[start + r] > array[start + largest]):
largest = r
if largest != index:
swap(array, start + largest, start + index)
max_heapify(array, largest, start, end)
maxdepth = (len(a).bit_length() - 1)*2
return sort(a, 0, len(a), maxdepth)
def heapsort(self, a):
def sort(array):
array = array
n = len(array)
for i in range(n//2 - 1, -1, -1):
heapify(array, n, i)
for i in range(n-1, 0, -1):
array[i], array[0] = array[0], array[i]
heapify(array, i, 0)
return array
def heapify(array, n, i):
array = array
largest = i
l = 2 * i + 1
r = 2 * i + 2
if l < n and array[i] < array[l]:
largest = l
if r < n and array[largest] < array[r]:
largest = r
if largest != i:
array[i],array[largest] = array[largest],array[i]
heapify(array, n, largest)
return array
return sort(a)
def insertionsort(self, a):
def sort(array):
array = array
for i in range(1, len(array)):
key = array[i]
j = i-1
while j >=0 and key < array[j] :
array[j+1] = array[j]
j -= 1
array[j+1] = key
return array
return sort(a)
def timsort(self, a, block = 32):
BLOCK = block
def sort(array, n):
array = array
for i in range(0, n, BLOCK):
insertionsort(array, i, min((i+31), (n-1)))
size = BLOCK
while size < n:
for left in range(0, n, 2*size):
mid = left + size - 1
right = min((left + 2*size - 1), (n-1))
merge(array, left, mid, right)
size = 2*size
return array
def insertionsort(array, left, right):
array = array
for i in range(left + 1, right+1):
temp = array[i]
j = i - 1
while j >= left and array[j] > temp :
array[j+1] = array[j]
j -= 1
array[j+1] = temp
return array
def merge(array, l, m, r):
len1, len2 = m - l + 1, r - m
left, right = [], []
for i in range(0, len1):
left.append(array[l + i])
for i in range(0, len2):
right.append(array[m + 1 + i])
i, j, k = 0, 0, l
while i < len1 and j < len2:
if left[i] <= right[j]:
array[k] = left[i]
i += 1
else:
array[k] = right[j]
j += 1
k += 1
while i < len1:
array[k] = left[i]
k += 1
i += 1
while j < len2:
array[k] = right[j]
k += 1
j += 1
return sort(a, len(a))
def selectionsort(self, a):
array = a
for i in range(len(array)):
min_idx = i
for j in range(i+1, len(array)):
if array[min_idx] > array[j]:
min_idx = j
array[i], array[min_idx] = array[min_idx], array[i]
return array
def shellsort(self, a):
array = a
n = len(array)
gap = n//2
while gap > 0:
for i in range(gap,n):
temp = array[i]
j = i
while j >= gap and array[j-gap] >temp:
array[j] = array[j-gap]
j -= gap
array[j] = temp
gap //= 2
return array
def bubblesort(self, a):
def sort(array):
for i, num in enumerate(array):
try:
if array[i+1] < num:
array[i] = array[i+1]
array[i+1] = num
sort(array)
except IndexError:
pass
return array
return sort(a)
def cyclesort(self, a):
def sort(array):
array = array
writes = 0
for cycleStart in range(0, len(array) - 1):
item = array[cycleStart]
pos = cycleStart
for i in range(cycleStart + 1, len(array)):
if array[i] < item:
pos += 1
if pos == cycleStart:
continue
while item == array[pos]:
pos += 1
array[pos], item = item, array[pos]
writes += 1
while pos != cycleStart:
pos = cycleStart
for i in range(cycleStart + 1, len(array)):
if array[i] < item:
pos += 1
while item == array[pos]:
pos += 1
array[pos], item = item, array[pos]
writes += 1
return array
return sort(a)
def cocktailsort(self, a):
def sort(array):
array = array
n = len(array)
swapped = True
start = 0
end = n-1
while (swapped == True):
swapped = False
for i in range (start, end):
if (array[i] > array[i + 1]) :
array[i], array[i + 1]= array[i + 1], array[i]
swapped = True
if (swapped == False):
break
swapped = False
end = end-1
for i in range(end-1, start-1, -1):
if (array[i] > array[i + 1]):
array[i], array[i + 1] = array[i + 1], array[i]
swapped = True
start = start + 1
return array
return sort(a)