String replacement of column names in python

Question

I have a data frame df that contains 15 columns (See below)

A_phase_1,A_phase_2,A_phase_3,A_phase_4,A_phase_5, B_phase_1,B_phase_2,B_phase_3,B_phase_4,B_phase_5, C_phase_1,C_phase_2,C_phase_3,C_phase_4,C_phase_5.

While retaining columns with phase_1 for example=A_phase_1, B_phase_1 and C_phase_1,

I want to programmatically do the following:

1. delete the columns containing phase_2 that is = A_phase_2,B_phase_2,C_phase_2

2. replace the remainder column names phase_n with phase_n-1. for example,

all phase_3 become phase_2;

phase_4 --> phase_3;

phase_5 --> phase_4

So desired output should look like this :

A_phase_1,A_phase_2,A_phase_3,A_phase_4,

B_phase_1,B_phase_2,B_phase_3,B_phase_4,

C_phase_1,C_phase_2,C_phase_3,C_phase_4.

Thanks in advance!

@seaBean I got this error

Answer 1

You can use .filter() to filter the columns to delete. Then, use df.columns.map() with lambda function to rename the columns, as follows:

df = df.drop(df.filter(like='phase_2').columns, axis=1)

df.columns = df.columns.map(lambda x: (x[:-1] + str(int(x[-1]) -1)) if (x[-1].isdigit() and (int(x[-1]) > 1)) else x)

Input:

print(df)

  A_phase_1 A_phase_2 A_phase_3 A_phase_4 A_phase_5 B_phase_1 B_phase_2 B_phase_3 B_phase_4 B_phase_5 C_phase_1 C_phase_2 C_phase_3 C_phase_4 C_phase_5 X_phase_t
0        A1        A2        A3        A4        A5        B1        B2        B3        B4        B5        C1        C2        C3        C4        C5        Xt

Output:

print(df)

  A_phase_1 A_phase_2 A_phase_3 A_phase_4 B_phase_1 B_phase_2 B_phase_3 B_phase_4 C_phase_1 C_phase_2 C_phase_3 C_phase_4 X_phase_t
0        A1        A3        A4        A5        B1        B3        B4        B5        C1        C3        C4        C5        Xt

Answer 2

# 1.) delete phase 2 cols 
#    more precisely, keep only the other columns
df = df[[col for col in df.columns if col[-1]!='2']]
# This next could be done inline but keeping it separate here
rename_dict = {col:col[:-1]+str(int(col[-1])-1) for col in df.columns if col[-1] in ['3','4','5']}
df.rename(columns=rename_dict,inplace=True)

Answer 3

from pandas import DataFrame
from itertools import product

def ren(col):
    prefix, sufix = col.rsplit('_', maxsplit=1)
    sufix = int(sufix)
    return f'{prefix}_{max(sufix-1, 1)}'

# some dummy df
df = DataFrame({f'{letter}_phase_{n}':[i, i, i] 
                for i, (letter, n) in 
                enumerate(product('ABC', range(1, 6)), start=1)})

df = df.drop(df.filter(like='phase_2').columns, axis=1)
# df = df.drop([f'{letter}_phase_2' for letter in 'ABC'], axis=1) # alternative
df = df.rename(mapper=ren, axis=1)
print(df)