4

I have a one CSV files in which I want rename some of the columns with same name. my initial code looks like this

df = pd.read_csv('New.csv')

I extracted selected columns from dataframe with this code

df.columns[1::3]

this slicing gets every third columns. Now I want to rename those every third columns with same names but trying to rename my columns like this gives me an error

df.columns[1::3]= ['SomeName']
raise TypeError("Index does not support mutable operations")

Is there any way I can rename multiple columns with same name in pandas ?

Any other suggestions than doing this manually ?

Improve this question
1
  • If need dynamic rename, what is pattern for new column names? Commented May 17, 2018 at 8:19

3 Answers 3

Reset to default
3

I think best here is use rename with unique new columns names like:

df = pd.DataFrame({'A':list('abcdef'),
                   'B':[4,5,4,5,5,4],
                   'C':[7,8,9,4,2,3],
                   'D':[1,3,5,7,1,0],
                   'E':[5,3,6,9,2,4],
                   'F':list('aaabbb')})

print (df)
   A  B  C  D  E  F
0  a  4  7  1  5  a
1  b  5  8  3  3  a
2  c  4  9  5  6  a
3  d  5  4  7  9  b
4  e  5  2  1  2  b
5  f  4  3  0  4  b

d = dict(zip(df.columns[1::3], range(len(df.columns[1::3]))))
print (d)
{'B': 0, 'E': 1}


df = df.rename(columns=d)
print (df)
   A  0  C  D  1  F
0  a  4  7  1  5  a
1  b  5  8  3  3  a
2  c  4  9  5  6  a
3  d  5  4  7  9  b
4  e  5  2  1  2  b
5  f  4  3  0  4  b

Or:

d = dict(zip(df.columns[1::3], 
             ['name{}'.format(x) for x in range(len(df.columns[1::3]))]))
print (d)
{'B': 'name0', 'E': 'name1'}

df = df.rename(columns=d)
print (df)
   A  name0  C  D  name1  F
0  a      4  7  1      5  a
1  b      5  8  3      3  a
2  c      4  9  5      6  a
3  d      5  4  7      9  b
4  e      5  2  1      2  b
5  f      4  3  0      4  b

Not recommended solution is rename for same column names:

d = dict.fromkeys(df.columns[1::3], 'Name')
print (d)
{'B': 'Name', 'E': 'Name'}

df = df.rename(columns=d)
print (df)
   A  Name  C  D  Name  F
0  a     4  7  1     5  a
1  b     5  8  3     3  a
2  c     4  9  5     6  a
3  d     5  4  7     9  b
4  e     5  2  1     2  b
5  f     4  3  0     4  b

because if want seelct column Name it return all columns in DataFrame:

print (df['Name'])
   Name  Name
0     4     5
1     5     3
2     4     6
3     5     9
4     5     2
5     4     4
Improve this answer
Sign up to request clarification or add additional context in comments.

3 Comments

The reason I want to use same column name here is that, I want to group those column names.
like d = dict.fromkeys(df.columns[1::3], 'Name') and then df['name'].sum(axis=1) ? Then it is possible use
No, the target is to use that grouping columns as an index not to sum.
1

df.columns is of type pandas.indexes.base.Index which is why you are getting the TypeError. If you convert it to a list, then you can update (rename) using the slice, and set df.columns to that updated list.

This works for me:

lst = list(df.columns)
lst[1::3] = ['someName']*len(lst[1::3])
df.columns = lst

or to have unique new column names (as @jezrael pointed out it might not be recommended to use the same name):

lst = list(df.columns)
lst[1::3] = ['someName{}'.format(x) for x in range(len(lst[1::3]))] 
df.columns = lst
Improve this answer

Comments

1

Underlying pandas index objects are numpy arrays.

You can take advantage of this fact to slice and assign using numpy conventions.

Data from @jezrael. The need to extract values explicitly is necessitated by this known issue.

df = pd.DataFrame({'A':list('abcdef'),
                   'B':[4,5,4,5,5,4],
                   'C':[7,8,9,4,2,3],
                   'D':[1,3,5,7,1,0],
                   'E':[5,3,6,9,2,4],
                   'F':list('aaabbb')})

arr = df.columns.values
arr[1::3] = range(2)
df.columns = arr

print(df)

   A  0  C  D  1  F
0  a  4  7  1  5  a
1  b  5  8  3  3  a
2  c  4  9  5  6  a
3  d  5  4  7  9  b
4  e  5  2  1  2  b
5  f  4  3  0  4  b
Improve this answer

5 Comments

No, rather not. very bad answer. stackoverflow.com/questions/43291781/…
In my opinion to dangerous do it, so downvoted. But if change answer, I remove downvote, no problem.
Just test it in pandas 0.23.0 and still problem, so dont use it.
In my opinion the best dont use it, because too weird behaviour.
@jezrael, But I'm not using it. I'm taking a copy of the values, changing them and reassigning back.. this isn't the same as your linked question at all.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.