pairwise-test-python/pairwise.py at main · istevenep/pairwise-test-python · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
"""
This module provides functions to perform pairwise tests on a pandas DataFrame.
"""

import itertools
import pandas as pd
from scipy.stats import chi2_contingency, fisher_exact


def _bonferroni_correct(p_value, n_tests):
    """Return the Bonferroni-corrected p-value capped at 1.0."""
    return min(p_value * n_tests, 1.0)


def chisquare(input, col1, col2):
    """
    Given a a pandas dataframe (input) and the name of 2 columns (col1, col2), computes chi2 test for each pairwise combination of col1 and returns p value
    col1 must be the fixed variable (es. Group)

    Arguments:
        input: a pandas DataFrame
        col1: name of the column with the fixed variable (es. Group)
        col2: name of the column with the dependent variable (es. Sex)

    Prints:
        chi2 square p value for each pairwise combinations of col1
    Returns:
        pandas DataFrame with columns 'Variable1', 'Variable2', 'p-value',
        and 'bonferroni p-value'
    """
    series1 = input[str(col1)]
    series2 = input[str(col2)]

    contingency = pd.crosstab(series2, series1)

    x = list(itertools.combinations(
        range(0, series1.nunique()), 2))
    print("Pairwise chi2 test comparison for combinations of: ",
          str(col1), " and variable: ", str(col2))

    results = []
    n_tests = len(x)

    for i in x:
        a, b = i
        c, p, dof, expected = chi2_contingency(contingency.iloc[:, [a, b]])
        bonferroni_p = _bonferroni_correct(p, n_tests)
        print("chi2 test between: ", contingency.columns[a],
              " and ", contingency.columns[b])
        print("pvalue: ", p)
        print("Bonferroni corrected pvalue: ", bonferroni_p)
        results.append(
            [contingency.columns[a], contingency.columns[b], p, bonferroni_p]
        )
        print("\n")

    return pd.DataFrame(
        results,
        columns=['Variable1', 'Variable2', 'p-value', 'bonferroni p-value']
    )


def fisher(input, col1, col2):
    """
    Given a a pandas dataframe (input) and the name of 2 columns (col1, col2), computes Fisher's exact test for each pairwise combination of col1 and returns p value
    col1 must be the fixed variable (es. Group)

    Arguments:
        input: a pandas DataFrame
        col1: name of the column with the fixed variable (es. Group)
        col2: name of the column with the dependent variable (es. Sex)

    Prints:
        Fisher's exact test p value for each pairwise combinations of col1
    Returns:
        pandas DataFrame with columns 'Variable1', 'Variable2', 'p-value',
        and 'bonferroni p-value'
    """
    series1 = input[str(col1)]
    series2 = input[str(col2)]

    contingency = pd.crosstab(series2, series1)

    x = list(itertools.combinations(
        range(0, series1.nunique()), 2))
    print("Pairwise Fisher's exact test comparison for combinations of: ",
          str(col1), " and variable: ", str(col2))

    results = []
    n_tests = len(x)

    for i in x:
        a, b = i
        table = contingency.iloc[:, [a, b]].values
        _, p = fisher_exact(table)
        bonferroni_p = _bonferroni_correct(p, n_tests)
        print("Fisher's exact test between: ", contingency.columns[a],
              " and ", contingency.columns[b])
        print("pvalue: ", p)
        print("Bonferroni corrected pvalue: ", bonferroni_p)
        results.append(
            [contingency.columns[a], contingency.columns[b], p, bonferroni_p]
        )
        print("\n")

    return pd.DataFrame(
        results,
        columns=['Variable1', 'Variable2', 'p-value', 'bonferroni p-value']
    )