python - 无法将对象数据类型转换为日期时间格式

Question

我想将列转换为日期时间格式并将日期和时间分开。我已经得到了给定的日期时间转换的解决方案 from the link below:和代码:

df2['date']=pd.to_datetime(df2['date'].astype(str), format='%Y%m%d%H%M',errors='coerce')

但是，每当我尝试查看 df2['date'] 的数据类型时，该列仍然是对象数据类型。当我尝试使用代码将其分离时:

df2['date'], df2['Time'] = df2['date'].dt.normalize(), df['date'].dt.time

它给了我错误:

AttributeError: 'DataFrame' object has no attribute 'dt'

当我打印 df2.columns 时，输出为

MultiIndex(levels=[[u'ALT', u'AW', u'CLG', u'DEWP', u'DIR', u'GUS', u'H', u'L', u'M', u'MAX', u'MIN', u'MW', u'PCP01', u'PCP06', u'PCP24', u'PCPXX', u'SD', u'SKC', u'SLP', u'SPD', u'STP', u'TEMP', u'VSB', u'W', u'date'], [u'F', u'GMT', u'MPH', u'Mb', u'Miles', u'Unnamed: 10_level_1', u'Unnamed: 12_level_1', u'Unnamed: 13_level_1', u'Unnamed: 14_level_1', u'Unnamed: 15_level_1', u'Unnamed: 16_level_1', u'Unnamed: 17_level_1', u'Unnamed: 18_level_1', u'Unnamed: 19_level_1', u'Unnamed: 20_level_1', u'Unnamed: 3_level_1', u'Unnamed: 6_level_1', u'Unnamed: 7_level_1', u'Unnamed: 8_level_1', u'Unnamed: 9_level_1', u'inches']],
           labels=[[24, 4, 19, 5, 2, 17, 7, 8, 6, 22, 11, 11, 11, 11, 1, 1, 1, 1, 23, 21, 3, 18, 0, 20, 9, 10, 12, 13, 14, 15, 16], [1, 15, 2, 2, 16, 17, 18, 19, 5, 4, 6, 7, 8, 9, 10, 11, 12, 13, 14, 0, 0, 3, 20, 3, 0, 0, 20, 20, 20, 20, 20]])

df2.dtypes 的输出为:

date   GMT                     object
DIR    Unnamed: 3_level_1      object
SPD    MPH                    float64
GUS    MPH                      int64
CLG    Unnamed: 6_level_1      object
SKC    Unnamed: 7_level_1      object
L      Unnamed: 8_level_1      object
M      Unnamed: 9_level_1      object
H      Unnamed: 10_level_1     object
VSB    Miles                  float64
MW     Unnamed: 12_level_1     object
       Unnamed: 13_level_1     object
       Unnamed: 14_level_1     object
       Unnamed: 15_level_1     object
AW     Unnamed: 16_level_1     object
       Unnamed: 17_level_1     object
       Unnamed: 18_level_1     object
       Unnamed: 19_level_1     object
W      Unnamed: 20_level_1     object
TEMP   F                        int64
DEWP   F                       object
SLP    Mb                      object
ALT    inches                  object
STP    Mb                      object
MAX    F                       object
MIN    F                       object
PCP01  inches                  object
PCP06  inches                  object
PCP24  inches                  object

PCPXX  inches                  object
SD     inches                  object

Answer 1

列中有MultiIndex，因此对于选择列需要元组:

df2[('date', 'GMT')] = df2[('date', 'GMT')].dt.normalize()
df2[('Time', 'GMT')] = df2[('date', 'GMT')].dt.time