CODEWITHSUNDEEP
pythonpython-3.xpandas
Piyush S. Wanare
Piyush S. Wanare

Reputation: 4943

Pandas Change column date format

I have column in pandas dataframe as follows:

0    2018-04-06
1    2018-04-06
2    2018-04-09
3    2018-04-19
4    2018-04-19
5    2018-04-17

I want to convert this column into yyyy/mm/dd for which I have coaded as follows:

def change_date_format(x):
    if x != 'nan' and x != '' and x != ' ' and x != 0:
        x = parse(x, dayfirst=True).strftime("%Y-%m-%d")
        return x
    else:
        return ''

read4['Column Name'] = read4['Column Name'].apply(lambda x : change_date_format(x)  )

But it's convert as follows:

2018-06-04
2018-06-04
2018-09-04
2018-04-19
2018-04-19
2018-04-17

which ideally should be :

2018-04-06
2018-04-06
2018-04-09
2018-04-19
2018-04-19
2018-04-17

How do I force it do work as above. Basically it should consider the input also and depends on that it should work.

Upvotes: 2

Views: 357

Answers (2)

jezrael
jezrael

Reputation: 863741

I think need to_datetime with parameter errors='coerce' for convert not parseable values to NaT, then strftime and last replace:

read4['Column Name'] = (pd.to_datetime(read4['Column Name'], errors='coerce')
                          .dt.strftime("%Y-%m-%d")
                          .replace('NaT', ''))
  Column Name
0  2018-04-06
1  2018-04-06
2  2018-04-09
3  2018-04-19
4  2018-04-19
5  2018-04-17

Upvotes: 2

wookiekim
wookiekim

Reputation: 1176

Maybe you need to set your dayfirst flag to False?

x = parse(x, dayfirst=False).strftime("%Y-%m-%d")

Upvotes: 1

Related Questions