I have a table where there are four columns consisting of dates in simple numbers (without any separator). There are some rows in 4th column having 0 instead of date, which I want to skip(not drop) and change the format of other dates to DD/MM/YYYY

The output should be:

Logic: The dates in the first table should change to a format of DD/MM/YYYY and for the 4th column it should ignore the 0.

Thanks.

CodePudding user response：

You can convert date4 column to datetime by pd.to_datetime with errors='coerce' which treats mismatched time as NaT then convert it to string with .dt.strftime and at last fill the NaT with original mismatched time.

df['date4'] = (pd.to_datetime(df['date4'], format='%Y%m%d', errors='coerce')
               .dt.strftime('%d/%m/%Y')
               .fillna(df['date4'])) # or fillna(0)

print(df)

     Date1.     date2     date3       date4
0  20200122  20201012  20210424  10/07/2018
1  20190315  20210120  20180710           0
2  20220514  20200122  20200122           0
3  20180317  20200122  20201003  22/01/2020
4  20201003  20200122  20200122  03/12/2017
5  20171203  20200122  20200122           0
6  20210423  20200122  20200122  15/03/2019