[pandas] General pd.datetime tricks

!pip install jpholiday
from jpholiday import year_holidays
import pandas as pd
import numpy as np

def random_datetimes_or_dates(start, end, out_format='datetime', n=10):
    start_u = start.value//10**9
    end_u = end.value//10**9

    return pd.DatetimeIndex((10**9*np.random.randint(start_u, end_u, n, dtype=np.int64)).view('M8[ns]'))

holidays = pd.to_datetime(np.array(year_holidays(2017))[:,0])

start = pd.to_datetime('2017-05-01')
end = pd.to_datetime('2018-01-01')

df = pd.DataFrame({'date': random_datetimes_or_dates(start, end, out_format= 'datetime', n=1000).sort_values()})
df['date'] = pd.to_datetime(df['date'], format='%Y-%m-%d', errors='ignore')


# Convert to 4 seasons
# print('seasons: ', [(month%12 + 3)//3 for month in range(1, 13)])
# pandas built-in functions
df['year'] = df['date'].dt.year
df['month'] = df['date'].dt.month
df['week'] = df['date'].dt.week
df['day'] = df['date'].dt.day
df['hour'] = df['date'].dt.hour
df['minute'] = df['date'].dt.minute
df['second'] = df['date'].dt.second
df['dayofweek'] = df['date'].dt.dayofweek
df['dayofweek_str'] = df['date'].dt.day_name()

# numerically convert date to season
df['season'] = (df['date'].dt.month%12 + 3)//3

# japanese public holidays
df['is_holidays'] = df['date'].isin(holidays).astype(int)

# Late night: 0 ~ 7, Morning: 7 ~ 12, Lunch: 12 ~ 13, Afternoon: 13 ~ 18 Evening: 18 ~ 24
b = [0,6,11,12,17,24]
l = [0, 1, 2, 3, 4]
df['session'] = pd.cut(df['date'].dt.hour, bins=b, labels=l, include_lowest=True)