1
|
|
|
import numpy as np |
2
|
|
|
import pandas as pd |
3
|
|
|
import unittest |
4
|
|
|
from klib.describe import _missing_vals |
5
|
|
|
|
6
|
|
|
if __name__ == '__main__': |
7
|
|
|
unittest.main() |
8
|
|
|
|
9
|
|
|
data_mv = pd.DataFrame([[1, np.nan, 3, 4], |
10
|
|
|
[None, 4, 5, ], |
11
|
|
|
['a', 'b', pd.NA, 'd'], |
12
|
|
|
[True, False, 7, pd.NaT]], |
13
|
|
|
columns=['Col1', 'Col2', 'Col3', 'Col4']) |
14
|
|
|
|
15
|
|
|
|
16
|
|
|
class Test__missing_vals(unittest.TestCase): |
17
|
|
|
|
18
|
|
|
def test_mv_total(self): |
19
|
|
|
# Test total missing values |
20
|
|
|
self.assertAlmostEqual(_missing_vals(data_mv)['mv_total'], 5) |
21
|
|
|
|
22
|
|
|
def test_mv_rows(self): |
23
|
|
|
# Test missing values for each row |
24
|
|
|
self.assertAlmostEqual(_missing_vals(data_mv)['mv_rows'][0], 1) |
25
|
|
|
self.assertAlmostEqual(_missing_vals(data_mv)['mv_rows'][1], 2) |
26
|
|
|
self.assertAlmostEqual(_missing_vals(data_mv)['mv_rows'][2], 1) |
27
|
|
|
self.assertAlmostEqual(_missing_vals(data_mv)['mv_rows'][2], 1) |
28
|
|
|
|
29
|
|
|
def test_mv_cols(self): |
30
|
|
|
# Test missing values for each column |
31
|
|
|
self.assertAlmostEqual(_missing_vals(data_mv)['mv_cols'][0], 1) |
32
|
|
|
self.assertAlmostEqual(_missing_vals(data_mv)['mv_cols'][1], 1) |
33
|
|
|
self.assertAlmostEqual(_missing_vals(data_mv)['mv_cols'][2], 1) |
34
|
|
|
self.assertAlmostEqual(_missing_vals(data_mv)['mv_cols'][3], 2) |
35
|
|
|
|
36
|
|
|
def test_mv_rows_ratio(self): |
37
|
|
|
# Test missing values ratio for each row |
38
|
|
|
self.assertAlmostEqual(_missing_vals(data_mv)['mv_rows_ratio'][0], 0.25) |
39
|
|
|
self.assertAlmostEqual(_missing_vals(data_mv)['mv_rows_ratio'][1], 0.5) |
40
|
|
|
self.assertAlmostEqual(_missing_vals(data_mv)['mv_rows_ratio'][2], 0.25) |
41
|
|
|
self.assertAlmostEqual(_missing_vals(data_mv)['mv_rows_ratio'][3], 0.25) |
42
|
|
|
|
43
|
|
|
def test_mv_cols_ratio(self): |
44
|
|
|
# Test missing values ratio for each row |
45
|
|
|
self.assertAlmostEqual(_missing_vals(data_mv)['mv_cols_ratio'][0], 0.25) |
46
|
|
|
self.assertAlmostEqual(_missing_vals(data_mv)['mv_cols_ratio'][1], 0.25) |
47
|
|
|
self.assertAlmostEqual(_missing_vals(data_mv)['mv_cols_ratio'][2], 0.25) |
48
|
|
|
self.assertAlmostEqual(_missing_vals(data_mv)['mv_cols_ratio'][3], 0.5) |
49
|
|
|
|