change date format with timezone to yyyy-mm-dd hh:mm:ss

Question

I am familiar with strftime however they require date to be in datetime format. In my case I get data that looks like 01/Oct/2019:07:34:34 +0900 . I would like to convert it into 2019-10-01 07:34:34 .

NOTE: my date is in object datatype.

I've read date format with timezone and https://www.guru99.com/date-time-and-datetime-classes-in-python.html however they both assume I have datetime format that looks like datetime(2019, 10, 1, 07, 34, 34)

thanks to month name to month number and vice versa in python .

I've created a function that changes the format

import calendar
month_dict = dict((v,k) for k,v in enumerate(calendar.month_abbr))

def date_formatter(date):
    day, month, year_and_time = date.split()[0].split('/')
    year, hour, minute, second = year_and_time.split(":")

    month = month_dict[month]

    return f'{year}-{month}-{day} {hour}:{minute}:{second}'

then since I am working with dataframe, use apply + lambda to apply to all rows.

df['new_date'] = df['date'].apply(lambda row: date_formatter(row))

This works perfectly and people might be telling me to just use it, I am just curious if there is any alternative way.

Answer 1

If the date you provided is a string, then use

from datetime import datetime

myDate = '01/Oct/2019:07:34:34 +0900'

dt = datetime.strptime(myDate, '%d/%b/%Y:%H:%M:%S %z')
dt_to_string = dt.strftime('%Y-%m-%d %H:%M:%S')
print(dt_to_string)

2019-10-01 07:34:34

Theres this for future use: https://strftime.org/

EDIT: Saw you were in a dataframe, use

df['Datetime'] = pd.to_datetime(df['DateTimeColumn'], format='%d/%b/%Y:%H:%M:%S %z')

Answer 2

from datetime import datetime

old_date = "01/Oct/2019:07:34:34 +0900"
new_date = datetime.strptime(date, "%d/%b/%Y:%X %z").strftime("%Y-%m-%d %X")

print(new_date)

2019-10-01 07:34:34

Check out this cool page: https://strftime.org/