将数据框中的对象（时间）类型列转换为日期时间

Question

I am struggling to convert a object type to datetime. It's a column from a dataframe which has time values in format hh:mm:ss.

dataframe: df_train column: time values in format: hh:mm:ss

Below are the options which I already tried without any luck

Opt1:

df_train['time'] = pd.to_datetime(df_train['time'])
Error:
TypeError: <class 'datetime.time'> is not convertible to datetime

Opt2:

df_train['time'] = pd.to_datetime(df_train['time'], format='%H:%M:%S').dt.time
Outcome:
The code is working but the type remain as object.

Opt3:

df_train['time'] = pd.to_datetime(df_train['time'].str.strip(), format='%H:%M:%S')
Outcome:
value changed to NaT

Opt4:

df_train['time'] = pd.to_datetime(df_train['time']).dt.time
Error:
TypeError: <class 'datetime.time'> is not convertible to datetime

Appreciate your suggestions to convert this column to type datetime.

Answer 1

Just try this :

dat['column']  = pd.to_datetime(dat['column'])

OR pd.read_csv(file, parse_dates=['column']) when reading data from file

Answer 2

我想很近，您需要的是to_timedelta for timedeltas，并将python对象时间转换为字符串：

df_train['time'] = pd.to_timedelta(df_train['time'].astype(str))