熊猫优雅地处理具有不同日期时间格式的列

Question

I have a column with a birthdate. Some are NA, some 01.01.2016 but some contain 01.01.2016 01:01:01 Filtering the NA values works fine. But handling the different date formats seems clumsy. Is it possible to have pandas handle these gracefully and eg for a birthdate only interpret the date and not fail?

Answer 1

pd.to_datetime() will handle multiple formats

>>> ser = pd.Series(['NaT', '01.01.2016', '01.01.2016 01:01:01'])
>>> pd.to_datetime(ser)
0                   NaT
1   2016-01-01 00:00:00
2   2016-01-01 01:01:01
dtype: datetime64[ns]