Filter dataframe with multicolumn index in first & second levels-CodePudding

I have a data frame with two columns of an index like this.

             A                   B         
             one     two      three   four
ID
A       0.895717 -1.206412  1.431256 -1.170299
B       0.410835  0.132003 -0.076467  1.130127
C      -1.413681  1.024180  0.875906  0.974466

I want to filter it by first & second levels as mentioned below:

I want to mention only the first level.

df[A]

             A                
             one     two    
ID
A       0.895717 -1.206412  
B       0.410835  0.132003 
C      -1.413681  1.024180

I want to mention columns of second-level like this:

df[A[two],B[three,four]]

             A             B         
ID          two         three   four
A       -1.206412     1.431256 -1.170299
B       0.132003     -0.076467  1.130127
C       1.024180      0.875906  0.974466

Kindly help!

CodePudding user response：

First selection:

df[['A']]
           A          
         one       two
ID                    
A   0.895717 -1.206412
B   0.410835  0.132003
C  -1.413681  1.024180

Second one:

df[[('A', 'one'), ('B', 'three'), ('B', 'four')]]

           A         B          
         one     three      four
ID                              
A   0.895717  1.431256 -1.170299
B   0.410835 -0.076467  1.130127
C  -1.413681  0.875906  0.974466

If really you have many combinations and want to compress the selector you could do something like:

l = [['A', 'one'], ['B', ['three', 'four']]]
idx = [(a,b) for A,B in l
       for a in (A if isinstance(A, list) else [A])
       for b in (B if isinstance(B, list) else [B])
       ]
df[idx]