Skipping non-date time values when converting to datetime in Pandas

Question

I have a dataframe where one of the columns has data that I want to turn into datetime. But there are other, miscellaneous values in that column. Do I need to clean out those values before I convert? I keep getting an error about these rouge values gumming up my pd.to_datetime function.

Is there a way I can 'skip' over non-date values and convert only date values to datetime?

Answer 1

You can use the errors='coerce' option, e.g.:

df = pd.DataFrame({'date': '2021-01-01', 'foo', '2021-03-31'})

#          date
# 0  2021-01-01
# 1         foo
# 2  2021-03-31

df.date = pd.to_datetime(df.date, errors='coerce')

#          date
# 0  2021-01-01
# 1         NaT
# 2  2021-03-31