python test_pandas.py

Posted 2021-05-09

#!/usr/bin/env python3
# -*- coding: utf-8 -*-

from datetime import datetime
import itertools
import pandas as pd
import pprint
import pytz
import unittest

pp = pprint.PrettyPrinter(indent=4)


class TestPandas(unittest.TestCase):

    def setUp(self):
        self.f = 'cpu.csv'
        self.t = '24'
        self.df = pd.read_csv(self.f, header=0)

    def tearDown(self):
        pass

    def test_len_df(self):
        expected = 7445
        self.assertEqual(expected, len(self.df))

    def test_add_date_column(self):
        expected = 9
        self.df['date'] = self.df['time'].str.replace(' ..:..:..$', '')
        rows = self.df['date'].value_counts()
        self.assertEqual(expected, rows['2016/07/25'])

    def test_less_than_24(self):
        expected_indexes = ['2016/12/13', '2016/07/25']
        expected_values = [20, 9]
        self.df['date'] = self.df['time'].str.replace(' ..:..:..$', '')
        rows = self.df['date'].value_counts()
        self.assertEqual(expected_indexes,
                         rows.where(rows < 24).dropna().index.tolist())
        self.assertEqual(expected_values,
                         rows.where(rows < 24).dropna().tolist())

    def test_unixtime(self):
        expected = 1469426400
        time_jst = '2016/7/25  15:00:00'
        jst_dt_naive = datetime.strptime(time_jst,
                                         '%Y/%m/%d %H:%M:%S')
        jst = pytz.timezone('Asia/Tokyo')
        jst_dt = jst.localize(jst_dt_naive)
        self.assertEqual(expected, int(jst_dt.timestamp()))

    def test_add_rows(self):
        expected = []
        column_names = ['unixtime', 'time', 'date']
        row1 = ('1469426400', '2016/12/13  23:59', '2016/12/13')
        row2 = ('1469426400', '2016/07/25  23:59', '2016/07/25')
        self.df['date'] = self.df['time'].str.replace(' ..:..:..$', '')
        df_append = pd.DataFrame([row1]*4 + [row2]*15, columns=column_names)
        self.df = self.df.append(df_append, ignore_index=True)
        rows = self.df['date'].value_counts()
        self.assertEqual(expected,
                         rows.where(rows < 24).dropna().index.tolist())

    def test_grouper(self):
        column1_names = [1, 2, 3, 4, 5, 6, 7, 8, 9]
        column2_names = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
        expected_1 = [(1, 2, 3), (4, 5, 6), (7, 8, 9)]
        expected_2 = [(1, 2, 3), (4, 5, 6), (7, 8, 9), (10, None, None)]
        num = 3
        self.assertEqual(expected_1, list(self.grouper(column1_names, num)))
        self.assertEqual(expected_2, list(self.grouper(column2_names, num)))

    def grouper(self, seq, n, fillvalue=None):
        args = [iter(seq)] * n
        return itertools.zip_longest(*args, fillvalue=fillvalue)


if __name__ == '__main__':
    unittest.main()