Add suffix to dataframe column without changing the structure-CodePudding

I am trying to add a suffix to a few columns of the pandas dataframe without touching the existing columns. I tried with the below code but it removes all the other columns:

df_client.iloc[:,-3:].add_suffix('_P1')

I can break the dataframe in two, rename columns of the 2nd dataframe and merge them back, but I wanted to know if there is an easier option available in pandas.

Sample input: df:

Name DL PL KH DL PL KH 
A     1  1  2  3  4  1
B     1  1  2  3  4  1

Expected output: df:

Name DL PL KH DL_P1 PL_P1 KH_P1 
A     1  1  2    3    4     1
B     1  1  2    3    4     1

CodePudding user response：

When you use df.columns it outputs a collection of strings. These are the column names.

If you have a list of column names you want to change to, let's say it's called new_list, you can do df.columns = new_list. This changes the names of all the columns, without changing data. (Gotta make sure the lists are the same length)

So just prepare a list of new column names.

new_columns = list(df.columns[:-3])   [c   "_P1" for c in df.columns[-3:]]

Then simply df.columns = new_columns.

CodePudding user response：

No there isn't a builtin function, you can use a custom function:

def suffix():
    yield ''
    i = 0
    while True:
        i  = 1
        yield f'_P{i}'

from collections import defaultdict
d = defaultdict(suffix)

df.columns  = df.columns.map(lambda x: x next(d[x]))

output:

  Name  DL  PL  KH  DL_P1  PL_P1  KH_P1
0    A   1   1   2      3      4      1
1    B   1   1   2      3      4      1