calculate business days including custom holidays for two dates columns in python

Question

I am trying to get the business days between two date columns by adding a new date_difference column.

What have I tried?

bd = workday(ll.date_done,ll.order_date, cal)
bd = networkdays(ll.date_done,ll.order_date, cal)
km = np.busday_count(dd.date(), od.date(), holidays = cal)

An error I am getting.

'Series' object has no attribute 'days'

What I am trying to do.

# Table Name: ll
order_date  date_done
2017-04-09  2017-04-16
2017-04-09  2017-04-18
2017-04-10  2017-04-20

What I am expecting:

order_date  date_done      Date_Difference
2017-04-09  2017-04-16        4
2017-04-09  2017-04-18        7
2017-04-10  2017-04-20        6

What Calculation I am doing?

I have holidays list excluding which I am trying to get the business days difference from two dates. The above date_difference calculation is just dummy numbers.

Answer 1

You can use numpy.bus_daycount() if you pass it the correct types as:

Code:

df['date_diff'] = np.busday_count(
    df.order_date.values.astype('datetime64[D]'),
    df.date_done.values.astype('datetime64[D]'))

Test Code:

import pandas as pd
import numpy as np

from io import StringIO
df = pd.read_fwf(StringIO(u"""
    order_date  date_done
    2017-04-09  2017-04-16
    2017-04-09  2017-04-18
    2017-04-10  2017-04-20"""), header=1)
df.order_date = pd.to_datetime(df.order_date, infer_datetime_format=True)
df.date_done = pd.to_datetime(df.date_done, infer_datetime_format=True)

df['date_diff'] = np.busday_count(
    df.order_date.values.astype('datetime64[D]'),
    df.date_done.values.astype('datetime64[D]'))

print(df)

Results:

  order_date  date_done  date_diff
0 2017-04-09 2017-04-16          5
1 2017-04-09 2017-04-18          6
2 2017-04-10 2017-04-20          8