Python Pandas整数YYYYMMDD到日期时间

to_datetime

In [92]:t = 20070530pd.to_datetime(str(t), format='%Y%m%d')Out[92]:Timestamp('2007-05-30 00:00:00')

例：

In [94]:t = 20070530df = pd.Dataframe({'date':[t]*10})dfOut[94]:       date0  200705301  200705302  200705303  200705304  200705305  200705306  200705307  200705308  200705309  20070530In [98]:df['DateTime'] = df['date'].apply(lambda x: pd.to_datetime(str(x), format='%Y%m%d'))dfOut[98]:       date   DateTime0  20070530 2007-05-301  20070530 2007-05-302  20070530 2007-05-303  20070530 2007-05-304  20070530 2007-05-305  20070530 2007-05-306  20070530 2007-05-307  20070530 2007-05-308  20070530 2007-05-309  20070530 2007-05-30In [99]:df.dtypesOut[99]:date      int64DateTime    datetime64[ns]dtype: object

编辑

实际上，将类型转换为字符串然后将整个系列转换为日期时间要快得多，而不是对每个值调用apply：

In [102]:df['DateTime'] = pd.to_datetime(df['date'].astype(str), format='%Y%m%d')dfOut[102]:       date   DateTime0  20070530 2007-05-301  20070530 2007-05-302  20070530 2007-05-303  20070530 2007-05-304  20070530 2007-05-305  20070530 2007-05-306  20070530 2007-05-307  20070530 2007-05-308  20070530 2007-05-309  20070530 2007-05-30

时机

In [104]:%timeit df['date'].apply(lambda x: pd.to_datetime(str(x), format='%Y%m%d'))100 loops, best of 3: 2.55 ms per loopIn [105]:%timeit pd.to_datetime(df['date'].astype(str), format='%Y%m%d')1000 loops, best of 3: 396 µs per loop