how to change to datetime format in python-CodePudding

     Date                      

    2021/8/1
    8-5-2021
    8-6-2021:08:00:00 PM

I would like all the values in this column to be in the format yyyy-m-dd.

This is what I have been trying but it gives me an error saying unknown string format:

df['Date']= pd.to_datetime(df['Date'].str.split(':', n=1).str[0])

CodePudding user response：

You could apply a custom function on your Date column to parse the date values.

import pandas as pd
import io
from datetime import datetime

temp_data=u"""Date
8/1/2021
8-5-2021
8-6-2021:08:00:00 PM
"""

data = pd.read_csv(io.StringIO(temp_data), sep=";", parse_dates=False)

def to_date(string_date):
    formats=["%d/%m/%Y","%d-%m-%Y","%d-%m-%Y:%I:%M:%S %p"]
    parsed_date=None
    for format in formats:
        try:
            parsed_date=datetime.strptime('8-6-2021:08:00:00 PM', "%d-%m-%Y:%I:%M:%S %p").date()
            return parsed_date
        except ValueError:
            pass
    raise RuntimeError(f"Unable to parse date {string_date} with available formats={formats}")

You can add new formats in the to_date function to parse any new date format.

data['Date']=data['Date'].apply(lambda row: to_date(row))

>>> data
        Date
0  2021-06-08
1  2021-06-08
2  2021-06-08

CodePudding user response：

Make use of the format option. For more info, check the to_datetime documentation.

df['Date'] = pd.to_datetime(df['Date'], format='%Y%m%d')