将多个ISO时间列复制到\u datetime

Question

我有多个包含ISO数据字符串的列，如

"CreateDate": "2020-04-30T06:12:29.424Z", "ApprovedDate": "2020-04-30T06:21:30.504Z"

当我用

df['CreateDate'] = pd.to_datetime(df['CreateDate'], infer_datetime_format=True, utc=True)

它很好用。但是

df[['CreateDate', 'ApprovedDate']] = pd.to_datetime(df[['CreateDate', 'ApprovedDate']], infer_datetime_format=True, utc=True)

给我

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

有没有办法解决这个问题？尝试了所有的日期时间选项，但没有成功

Answer 1

因为您有很好的ISO格式字符串，所以不需要使用任何关键字调用to_datetime。因此，一个简单的apply应该是一个选项：

import pandas as pd
df = pd.DataFrame({"CreateDate": ["2020-04-30T06:12:29.424Z"], 
                   "ApprovedDate": ["2020-04-30T06:21:30.504Z"]})

df[['CreateDate', 'ApprovedDate']] = df[['CreateDate', 'ApprovedDate']].apply(pd.to_datetime)
# df
#                         CreateDate                     ApprovedDate
# 0 2020-04-30 06:12:29.424000+00:00 2020-04-30 06:21:30.504000+00:00