Convert string to date with multiple formats

I have dates in the following format in the same file:

"%m/%d/%Y" --> YEAR with 4 digits
"%m/%d/%y" --> YEAR with 2 digits

I want to be able to parse both formats. Here’s my attempt:

df[["field1", "field2"]] = pd.to_datetime(
    df[["field1", "field2"]], format="%m/%d/%Y", errors="coerce"
).fillna(pd.to_datetime(df[["field1", "field2"]], format="%m/%d/%y"))

But that is not working:

ValueError: to assemble mappings requires at least that [year, month, day] be specified: [day,month,year] is missing

If I try with only one serie:

df["field1"] = pd.to_datetime(
    df["field1"], format="%m/%d/%Y", errors="coerce"
).fillna(pd.to_datetime(df["field1"], format="%m/%d/%y"))

I get ValueError and doesn’t seem like both formats are being tested.

ValueError: unconverted data remains: 21

What is the most efficient way to use different date formats?

>Solution :

IIUC, Just remove the format parameter from pd.to_datetime and let the inherit formatting handle the change from two digit years to four digit years.

pd.to_datetime(pd.Series(['01/01/2023', '01/02/23', '12/24/2021', '11/23/21']))

Output:

0   2023-01-01
1   2023-01-02
2   2021-12-24
3   2021-11-23
dtype: datetime64[ns]