I have a dataframe like this:

And I would like an output of this:

Thank you!

CodePudding user response：

Try this:

def split_date_range(row):
    start, end = row[['start', 'end']]
    dates = []

    s = start
    while True:
        e = s   pd.DateOffset(years=1) - pd.Timedelta(days=1)
        if e < end:
            dates.append((s, e))
            s = e   pd.Timedelta(days=1)
        else:
            dates.append((s, end))
            break
    return dates

tmp = df.set_index('id').apply(split_date_range, axis=1).explode()
result = pd.DataFrame([ 
    [id_, *timerange] for id_, timerange in tmp.iteritems()
], columns=['id', 'start', 'end'])

The code uses id_ to avoid collision with the id() built-in function.