Indexing and Selecting Data  

  • Operators used in indexing and selecting data in a Data Frame: 

      [  ]   =   indexing operator
  .      =   attribute operator

  • Pandas supports 3 types of Multi-axes indexing:

    • dfObj.loc( .. ):     label based indexing

    • dfObj.iloc( .. ):   integer rank based indexing

  .loc[ ]: indexing using labels - selecting rows   

df = pd.DataFrame(np.random.randn(3, 4),
                  index = ['a','b','c'],
                  columns = ['A', 'B', 'C', 'D'])

print( df )

               A         B         C         D
      a -0.605034 -0.739834 -0.251246  0.585301
      b  1.522427 -0.322365 -0.168772  1.641832
      c -0.521770 -1.004734  0.757724 -0.698973



 Selecting rows:

   1 row                  slices of rows
   df.loc['b'] =          df.loc['b':'c'] =    ending index ≤ 'c' !! 
        A    1.522427                    A         B         C         D
	B   -0.322365          b  1.522427 -0.322365 -0.168772  1.641832
	C   -0.168772	       c -0.521770 -1.004734  0.757724 -0.698973
	D    1.641832

   Selected rows
   df.loc[ ['a','c'] ] =  (use a list)
             A         B         C         D
   a -0.605034 -0.739834 -0.251246  0.585301
   c -0.521770 -1.004734  0.757724 -0.698973

  .loc[ ]: indexing using labels - selecting columns   

df = pd.DataFrame(np.random.randn(3, 4),
                  index = ['a','b','c'],
                  columns = ['A', 'B', 'C', 'D'])

print( df )

               A         B         C         D
      a  0.126892 -0.207192 -0.024788  0.463379
      b -0.270875  0.393623  0.292516  0.192060
      c -0.241555 -0.105059  0.220437 -0.517377



 Selecting columns:

   1 column               slices of columns
   df.loc[:,'B'] =        df.loc[:,'B':'D'] = 
       a   -0.207192                     B         C         D
       b    0.393623           a -0.207192 -0.024788  0.463379
       c   -0.105059	       b  0.393623  0.292516  0.192060
                               c -0.105059  0.220437 -0.517377

   Selected columns
   df.loc[ :,['B','D'] ] =  (use a list)
                 B         D
       a -0.207192  0.463379
       b  0.393623  0.192060
       c -0.105059 -0.517377

  .iloc[ ]: indexing using integer ranks - selecting rows   

df = pd.DataFrame(np.random.randn(3, 4),
                  index = ['a','b','c'],
                  columns = ['A', 'B', 'C', 'D'])

print( df )

               A         B         C         D
      a -0.605034 -0.739834 -0.251246  0.585301
      b  1.522427 -0.322365 -0.168772  1.641832
      c -0.521770 -1.004734  0.757724 -0.698973



 Selecting rows:

   1 row                  slices of rows
   df.iloc[ 1 ] =          df.iloc[ 1 : 3 ] =   ending index < 3 !! 
        A    1.522427                    A         B         C         D
	B   -0.322365          b  1.522427 -0.322365 -0.168772  1.641832
	C   -0.168772	       c -0.521770 -1.004734  0.757724 -0.698973
	D    1.641832

   Selected rows
   df.iloc[ [ 0 , 2 ] ] =  (use a list)
             A         B         C         D
   a -0.605034 -0.739834 -0.251246  0.585301
   c -0.521770 -1.004734  0.757724 -0.698973

  .iloc[ ]: indexing using integer rank - selecting columns   

df = pd.DataFrame(np.random.randn(3, 4),
                  index = ['a','b','c'],
                  columns = ['A', 'B', 'C', 'D'])

print( df )

               A         B         C         D
      a  0.126892 -0.207192 -0.024788  0.463379
      b -0.270875  0.393623  0.292516  0.192060
      c -0.241555 -0.105059  0.220437 -0.517377



 Selecting columns:

   1 column               slices of columns
   df.iloc[:, 1 ] =        df.iloc[:, 1 : 4 ] =     ending index < 4 !! 
       a   -0.207192                     B         C         D
       b    0.393623           a -0.207192 -0.024788  0.463379
       c   -0.105059	       b  0.393623  0.292516  0.192060
                               c -0.105059  0.220437 -0.517377

   Selected columns
   df.iloc[ :,[ 1 , 3 ] ] =  (use a list)
                 B         D
       a -0.207192  0.463379
       b  0.393623  0.192060
       c -0.105059 -0.517377