tests_as_linear_models.ipynb

Kernel: Python 3

Show that standard tests can be implemented as inference on linear models.

Code is from https://www.georgeho.org/tests-as-linear/

!git clone https://github.com/eigenfoo/tests-as-linear.git

Cloning into 'tests-as-linear'...
remote: Enumerating objects: 227, done.
remote: Counting objects: 100% (6/6), done.
remote: Compressing objects: 100% (6/6), done.
remote: Total 227 (delta 1), reused 0 (delta 0), pack-reused 221
Receiving objects: 100% (227/227), 5.05 MiB | 20.05 MiB/s, done.
Resolving deltas: 100% (129/129), done.

%cd tests-as-linear

/content/tests-as-linear

%ls

cheatsheets/        index.html   README.md             scripts/
CODE_OF_CONDUCT.md  LICENSE.txt  requirements-dev.txt  tests_as_linear/
CONTRIBUTING.md     Makefile     requirements.txt      tests-as-linear.ipynb

import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import patsy
import scipy
import statsmodels.api as sm
import statsmodels.formula.api as smf
from tests_as_linear import plots, utils

np.random.seed(1618)

plots.pearson_spearman_plot()
plt.show()