TypeError: descriptor 'strftime' for 'datetime.date' objects doesn't apply-CodePudding

I have a data frame (test_df) with the following columns and datatypes:

|     Column          |     Data Type    |
|---------------------|------------------|
|    flowId           |         object   |
|    flowName         |         object   |
|    executionId      |         object   |
|    startedAt        |         datetime64[ns, tzlocal()]       |
|    lastUpdatedAt    |         datetime64[ns, tzlocal()]       |
|    dataPullStartTime|         datetime64[ns, tzlocal()]       |
|    dataPullEndTime  |         datetime64[ns, tzlocal()]       |
|    bytesProcessed   |         float64       |
|    bytesWritten     |         float64       |
|    recordsProcessed |         float64       |

Before converting this data frame to a JSON type format, I want to change the datetime columns to string with strftime.

test_df[['startedAt','lastUpdatedAt', 'dataPullStartTime', 'dataPullEndTime']] = 
test_df[['startedAt','lastUpdatedAt', 
'dataPullStartTime','dataPullEndTime']].apply(datetime.strftime('%Y-%m-%d 
%H:%M:%S.%f'), axis=1)

However, I always receive the following error:

TypeError: descriptor 'strftime' for 'datetime.date' objects doesn't apply to a 'str' object

I don't get the problem because the data type of the respective columns is datetime not str. Can someone help here? I already tried out multiple solutions from stack, but I was unable to solve the issue.

Thanks in advance!

CodePudding user response：

you can do it one column by one

   for col in ['startedAt','lastUpdatedAt', 'dataPullStartTime','dataPullEndTime'] : 
        test_df[col] = test_df[col].apply(lambda x: datetime.strftime(x, '%Y-%m-%d %H:%M:%S.%f'))

CodePudding user response：

apply need a function, not an object as args.

df["dataPullEndTime"].apply(lambda x: datetime.strftime(x, '%Y-%m-%d %H:%M:%S.%f'))

Reproduce

datetime.strftime('%Y-%m-%d %H:%M:%S.%f')

TypeError: descriptor 'strftime' requires a 'datetime.date' object but received a 'str'

CodePudding user response：

You should rather use the vectorial dt.strfime for each column:

cols = ['startedAt','lastUpdatedAt', 'dataPullStartTime', 'dataPullEndTime']
df[cols] = df[cols].apply(lambda c: c.dt.strftime('%Y-%m-%d %H:%M:%S.%f'))

or, for in place modification:

cols = ['startedAt','lastUpdatedAt', 'dataPullStartTime', 'dataPullEndTime']
df.update(df[cols].apply(lambda c: c.dt.strftime('%Y-%m-%d %H:%M:%S.%f')))

example:

df = pd.DataFrame({'col1': ['2000-01-01'], 'col2': ['2022-01-01']})
df['col1'] = pd.to_datetime(df['col1'])
df['col2'] = pd.to_datetime(df['col2'])
print(df)

        col1       col2
0 2000-01-01 2022-01-01

cols = ['col1', 'col2']
df.update(df[cols].apply(lambda c: c.dt.strftime('%Y-%m-%d %H:%M:%S.%f')))
print(df)

                         col1                        col2
0  2000-01-01 00:00:00.000000  2022-01-01 00:00:00.000000