Count business days between several columns-CodePudding

I have a df that contains four columns:

       col1           col2         col3           col4
0  2021-10-18    2021-10-27    2021-11-08      2021-11-21   
1  2021-01-30    2021-02-01    2021-02-28      2021-03-18   
2  2021-08-17    2021-08-20    2021-09-02      2021-09-07

I would like to count the number of businessdays between the columns. So far I did this:

col1 = [d.date() for d in df['col1']]
col2 = [d.date() for d in df['col2 ']]
col3 = [d.date() for d in df['col3']]
col4 = [d.date() for d in df['col4 ']]

df['bday1'] = np.busday_count(col1, col2)
df['bday2'] = np.busday_count(col2, col3)
df['bday3'] = np.busday_count(col3, col4)

Is there a way to do this more efficiently? For example, for a case when there would be more columns? I'd be grateful for your help!

CodePudding user response：

Try this:

cols = ['col1', 'col2', 'col3', 'col4']

for i, (col1, col2) in enumerate(zip(cols[:-1], cols[1:])):
    df[f'bday{i 1}'] = np.busday_count(
        df[col1].dt.date.to_list(),
        df[col2].dt.date.to_list(),
    )

You just need to specify the columns you want to use.