pandas_date_range.py

import numpy as np
import pandas as pd


np.random.seed(1234)

start_date = "2017-07-01"
end_date = "2017-07-02"
_date_range = pd.date_range(start_date, end_date, freq="H")
_size = _date_range.shape[0]
sample_log = pd.DataFrame({
    "created_at" : pd.date_range(start_date, end_date, freq="H"),
    "click" : np.random.binomial(10, 0.1, size=_size)
}).query("click >0")
sample_log.index = np.arange(1, sample_log.shape[0]+1, 1)

print("[before]\n",sample_log.head(10))

start_at, end_at = sample_log.created_at.min(), sample_log.created_at.max()
sample_log.index = sample_log.created_at
res = sample_log.reindex(pd.date_range(start_at, end_at, freq="H"))

print("[after]\n",res.head(10))

res.created_at = res.index
res.click = res.click.fillna(0)

print("[after2]\n",res.head(10))