Renaming columns in pandas dataframe using regular expressions

Question

   Y2010  Y2011  Y2012  Y2013  test
0  86574  77806  93476  99626  2
1  60954  67873  65135  64418  4
2    156    575    280    330  6
3   1435   1360   1406   1956  7
4   3818   7700   6900   5500  8

Is there a way to rename the columns of this dataframe from Y2010... to 2010.. i.e. removing the initial 'Y'. I want to use regular expressions since I have quite a few such columns. I tried this:

df.rename(df.filter(regex='^Y\d{4}').columns.values, range(2010, 2013 + 1, 1))

--EDIT: The dataframe doees include columns which do not start with a 'Y'

Community · Accepted Answer · 2017-05-23 11:53:34Z

13

I'd use map:

In [11]: df.columns.map(lambda x: int(x[1:]))
Out[11]: array([2010, 2011, 2012, 2013])

In [12]: df.columns = df.columns.map(lambda x: int(x[1:]))

In [13]: df
Out[13]:
    2010   2011   2012   2013
0  86574  77806  93476  99626
1  60954  67873  65135  64418
2    156    575    280    330
3   1435   1360   1406   1956
4   3818   7700   6900   5500

Edit: I forgot the most-popular pandas question:

In [21]: df.rename(columns=lambda x: int(x[1:]))
Out[21]:
    2010   2011   2012   2013
0  86574  77806  93476  99626
1  60954  67873  65135  64418
2    156    575    280    330
3   1435   1360   1406   1956
4   3818   7700   6900   5500

If you have additional columns, I would probably write a proper function (rather than a lambda):

def maybe_rename(col_name):
    if re.match(r"^Y\d{4}", col_name):
        return int(col_name[1:])
    else:
        return col_name

In [31]: df.rename(columns=maybe_rename)
Out[31]:
    2010   2011   2012   2013  test
0  86574  77806  93476  99626     2
1  60954  67873  65135  64418     4
2    156    575    280    330     6
3   1435   1360   1406   1956     7
4   3818   7700   6900   5500     8

edited May 23, 2017 at 11:53

CommunityBot

11 silver badge

answered Sep 24, 2015 at 0:29

Andy Hayden

378k110 gold badges640 silver badges546 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

user308827 Over a year ago

my bad, the dataframe does include columns which do not start with a 'Y', which is why I was using filter. Will update question to reflect that

Evan Over a year ago

FWIW, maybe_rename worked as written on a MultiIndex, although the only regex matches were in the top level of the Multi-index. I'm not sure it would have behaved otherwise. Cheers!

Collectives™ on Stack Overflow

Renaming columns in pandas dataframe using regular expressions

1 Answer 1

2 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

2 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related