|
1
|
|
|
import numpy as np |
|
2
|
|
|
import pandas as pd |
|
3
|
|
|
import unittest |
|
4
|
|
|
from klib.describe import _missing_vals, corr_mat |
|
5
|
|
|
|
|
6
|
|
|
if __name__ == '__main__': |
|
7
|
|
|
unittest.main() |
|
8
|
|
|
|
|
9
|
|
|
|
|
10
|
|
|
class Test__missing_vals(unittest.TestCase): |
|
11
|
|
|
|
|
12
|
|
|
@classmethod |
|
13
|
|
|
def setUpClass(cls): |
|
14
|
|
|
cls.data_mv_df = pd.DataFrame([[1, np.nan, 3, 4], |
|
15
|
|
|
[None, 4, 5, None], |
|
16
|
|
|
['a', 'b', pd.NA, 'd'], |
|
17
|
|
|
[True, False, 7, pd.NaT]], |
|
18
|
|
|
columns=['Col1', 'Col2', 'Col3', 'Col4']) |
|
19
|
|
|
|
|
20
|
|
|
cls.data_mv_array = np.array([[1, np.nan, 3, 4], |
|
21
|
|
|
[None, 4, 5, None], |
|
22
|
|
|
['a', 'b', pd.NA, 'd'], |
|
23
|
|
|
[True, False, 7, pd.NaT]]) |
|
24
|
|
|
|
|
25
|
|
|
cls.data_mv_list = [[1, np.nan, 3, 4], |
|
26
|
|
|
[None, 4, 5, None], |
|
27
|
|
|
['a', 'b', pd.NA, 'd'], |
|
28
|
|
|
[True, False, 7, pd.NaT]] |
|
29
|
|
|
|
|
30
|
|
|
def test_mv_total(self): |
|
31
|
|
|
# Test total missing values |
|
32
|
|
|
self.assertAlmostEqual(_missing_vals(self.data_mv_df)['mv_total'], 5) |
|
33
|
|
|
self.assertAlmostEqual(_missing_vals(self.data_mv_array)['mv_total'], 5) |
|
34
|
|
|
self.assertAlmostEqual(_missing_vals(self.data_mv_list)['mv_total'], 5) |
|
35
|
|
|
|
|
36
|
|
|
def test_mv_rows(self): |
|
37
|
|
|
# Test missing values for each row |
|
38
|
|
|
self.assertAlmostEqual(_missing_vals(self.data_mv_df)['mv_rows'][0], 1) |
|
39
|
|
|
self.assertAlmostEqual(_missing_vals(self.data_mv_df)['mv_rows'][1], 2) |
|
40
|
|
|
self.assertAlmostEqual(_missing_vals(self.data_mv_df)['mv_rows'][2], 1) |
|
41
|
|
|
self.assertAlmostEqual(_missing_vals(self.data_mv_df)['mv_rows'][3], 1) |
|
42
|
|
|
|
|
43
|
|
|
def test_mv_cols(self): |
|
44
|
|
|
# Test missing values for each column |
|
45
|
|
|
self.assertAlmostEqual(_missing_vals(self.data_mv_df)['mv_cols'][0], 1) |
|
46
|
|
|
self.assertAlmostEqual(_missing_vals(self.data_mv_df)['mv_cols'][1], 1) |
|
47
|
|
|
self.assertAlmostEqual(_missing_vals(self.data_mv_df)['mv_cols'][2], 1) |
|
48
|
|
|
self.assertAlmostEqual(_missing_vals(self.data_mv_df)['mv_cols'][3], 2) |
|
49
|
|
|
|
|
50
|
|
View Code Duplication |
def test_mv_rows_ratio(self): |
|
|
|
|
|
|
51
|
|
|
# Test missing values ratio for each row |
|
52
|
|
|
self.assertAlmostEqual(_missing_vals(self.data_mv_df)['mv_rows_ratio'][0], 0.25) |
|
53
|
|
|
self.assertAlmostEqual(_missing_vals(self.data_mv_df)['mv_rows_ratio'][1], 0.5) |
|
54
|
|
|
self.assertAlmostEqual(_missing_vals(self.data_mv_df)['mv_rows_ratio'][2], 0.25) |
|
55
|
|
|
self.assertAlmostEqual(_missing_vals(self.data_mv_df)['mv_rows_ratio'][3], 0.25) |
|
56
|
|
|
|
|
57
|
|
|
# Test if missing value ratio is between 0 and 1 |
|
58
|
|
|
self.assertTrue(0 <= _missing_vals(self.data_mv_df)['mv_rows_ratio'][0] <= 1) |
|
59
|
|
|
self.assertTrue(0 <= _missing_vals(self.data_mv_df)['mv_rows_ratio'][1] <= 1) |
|
60
|
|
|
self.assertTrue(0 <= _missing_vals(self.data_mv_df)['mv_rows_ratio'][2] <= 1) |
|
61
|
|
|
self.assertTrue(0 <= _missing_vals(self.data_mv_df)['mv_rows_ratio'][3] <= 1) |
|
62
|
|
|
|
|
63
|
|
View Code Duplication |
def test_mv_cols_ratio(self): |
|
|
|
|
|
|
64
|
|
|
# Test missing values ratio for each row |
|
65
|
|
|
self.assertAlmostEqual(_missing_vals(self.data_mv_df)['mv_cols_ratio'][0], 0.25) |
|
66
|
|
|
self.assertAlmostEqual(_missing_vals(self.data_mv_df)['mv_cols_ratio'][1], 0.25) |
|
67
|
|
|
self.assertAlmostEqual(_missing_vals(self.data_mv_df)['mv_cols_ratio'][2], 0.25) |
|
68
|
|
|
self.assertAlmostEqual(_missing_vals(self.data_mv_df)['mv_cols_ratio'][3], 0.5) |
|
69
|
|
|
|
|
70
|
|
|
# Test if missing value ratio is between 0 and 1 |
|
71
|
|
|
self.assertTrue(0 <= _missing_vals(self.data_mv_df)['mv_cols_ratio'][0] <= 1) |
|
72
|
|
|
self.assertTrue(0 <= _missing_vals(self.data_mv_df)['mv_cols_ratio'][1] <= 1) |
|
73
|
|
|
self.assertTrue(0 <= _missing_vals(self.data_mv_df)['mv_cols_ratio'][2] <= 1) |
|
74
|
|
|
self.assertTrue(0 <= _missing_vals(self.data_mv_df)['mv_cols_ratio'][3] <= 1) |
|
75
|
|
|
|
|
76
|
|
|
|
|
77
|
|
|
class Test_corr_mat(unittest.TestCase): |
|
78
|
|
|
|
|
79
|
|
|
@classmethod |
|
80
|
|
|
def setUpClass(cls): |
|
81
|
|
|
cls.data_corr = pd.DataFrame([[1, 0, 3j, 4], |
|
82
|
|
|
[3, 4, 5, 6], |
|
83
|
|
|
['a', 'b', pd.NA, 'd'], |
|
84
|
|
|
[5, False, np.nan, pd.NaT]], |
|
85
|
|
|
columns=['Col1', 'Col2', 'Col3', 'Col4']) |
|
86
|
|
|
|
|
87
|
|
|
cls.data_corr_list = [1, 2, -3, 4j, 5, 0] |
|
88
|
|
|
|
|
89
|
|
|
def test_output_type(self): |
|
90
|
|
|
# Test conversion from pd.io.formats.style.Styler to pd.core.frame.DataFrame |
|
91
|
|
|
self.assertTrue(type(corr_mat(self.data_corr)), type(pd.DataFrame)) |
|
92
|
|
|
self.assertTrue(type(corr_mat(self.data_corr_list)), type(pd.DataFrame)) |
|
93
|
|
|
|
|
94
|
|
|
def test_output_shape(self): |
|
95
|
|
|
# Test for output of equal dimensions |
|
96
|
|
|
self.assertEqual(corr_mat(self.data_corr).data.shape[0], corr_mat(self.data_corr).data.shape[1]) |
|
97
|
|
|
self.assertEqual(corr_mat(self.data_corr_list).data.shape[0], corr_mat(self.data_corr_list).data.shape[1]) |
|
98
|
|
|
|
akanz1 /
klib
Loading history...