0

How to change dd-mm-yyyy date format to yyyy-dd-mm in pandas. I have a datefield which is already in dd-mm-yyyy format but when I try 

df[('date')] = pd.to_datetime(df[('date')]).dt.strftime('%Y-%m-%d')

it gives output a yyyy-dd-mm

jpp
  • 159,742
  • 34
  • 281
  • 339
Rahul
  • 23
  • 1
  • 4

3 Answers3

1

I believe this is what you needed.

import pandas as pd
df = pd.read_csv("dates.csv")

df

   id        date
0   1  25/06/2018
1   2  14-11-2005
2   3  03/10/2010
3   4  13-08-2008
4   5  05-05-2005

Here no need to specify the format as you have tried.

df['date'] =pd.to_datetime(df['date'])
df

 id       date
0   1 2018-06-25
1   2 2005-11-14
2   3 2010-03-10
3   4 2008-08-13
4   5 2005-05-05
msr_003
  • 1,205
  • 2
  • 10
  • 25
  • Hi,even if I don't specify it gives me result in yyyy-mm-dd... Does it matter if the earlier date format is dd/mm/yyyy delimiter is (/) rather than (-) ?? – Rahul Aug 30 '18 at 06:03
  • Can you please specify in which format you wanted it to be? Ex:- `yyyy-mm-dd` (or) `yyyy/mm/dd` (or) `dd-mm-yyyy` (or) `dd/mm/yyyy` – msr_003 Aug 30 '18 at 06:20
  • My date format is dd/mm/yyyy but i want it in yyyy-mm-dd format.. But when i do pd.datetime it gives me result as yyyy-dd-mm – Rahul Aug 30 '18 at 06:32
  • I've tested with `dd/mm/yyyy` format and the same function is giving me in `yyyy-mm-dd` format, not in `yyyy-dd-mm` format. – msr_003 Aug 30 '18 at 06:37
0

Pandas datetime series data do not have an inherent string format.

datetime values are stored internally as integers. For more details, see this answer. String representations are just that, representations. For example, when you use the print command, a specific string representation is used so that data is displayed in a human-readable way.

For most purposes, you should not worry about the representation. If you need a format different to the default representation, i.e. "YYYY-MM-DD", you can use pd.Series.dt.strftime and specify a string format. For this Python's strftime directives is a useful resource.

jpp
  • 159,742
  • 34
  • 281
  • 339
0

Use this:

import pandas as pd

df['date'] = pd.to_datetime(df['date'],format='%d-%m-%Y').dt.strftime('%Y-%m-%d')#specify input format '%d-%m-%Y' and output format '%Y-%m-%d' or change output as desired i.e. %d/%m/%Y to give dd/mm/yyyy
Ines
  • 141
  • 1
  • 6