[FIXED] How to create multiple diff columns in pandas? At once

Issue

Check the following dataset:

d = {'col1':[2,4,5,6],'col2':[8,8,6,1],'col3':[1,2,3,4],'col4':[4,4,4,4]}

df = pd.DataFrame(d,index=['Item1','Item2','Item1','Item2'])

Lets say i want to interact through the columns, and create a diff column. That gives me difference beetween col1, col2 and them another diff column that gives me col3,col4. And so on like in a window by window style. Difference columns.

d = {'col1':[2,4,5,6],'col2':[8,8,6,1],'diff_1':[-6,-4,1,5],'col3':[1,2,3,4],'col4':[4,4,4,4],'diff_2':[-3,-2,-1,0]}
    
df = pd.DataFrame(d,index=['Item1','Item2','Item1','Item2'])

Solution

You can check groupby then concat back

out = pd.concat([y.assign(**{'diff_{0}'.format(x+1): y.iloc[:,0] - y.iloc[:,1]})for x , y in df.groupby(np.arange(df.shape[1])//2,axis=1)],axis=1)
Out[166]: 
       col1  col2  diff_1  col3  col4  diff_2
Item1     2     8      -6     1     4      -3
Item2     4     8      -4     2     4      -2
Item1     5     6      -1     3     4      -1
Item2     6     1       5     4     4       0

Answered By - BENY