How To Apply Hierarchy Or Multi-index To Pandas Columns

I have seen lots of examples on how to arrange dataframe row indexes hierarchically, but I am trying to do the same for columns and am not understanding the syntax: Given: df = pd.

Solution 1:

You can define MultiIndices using the from_arrays or from_tuples or from_product classmethods. Here is an example using from_arrays:

arrays = [[1, 2]*3, ['A', 'B', 'C']*2]
columns = pd.MultiIndex.from_arrays(arrays, names=['foo', 'bar'])

df = pd.DataFrame(np.random.randn(2,6),
                  columns=columns,
                  index= pd.date_range('20000103',periods=2))

yields

In [81]: df
Out[81]: 
foo                1         2         1         2         1         2
bar                A         B         C         A         B         C
2000-01-03  1.277234 -0.899547  0.040337 -0.878752 -0.524336  0.922440
2000-01-04 -1.706797  0.450379  1.510868 -2.539827 -1.909996 -0.003851

Defining a MultiIndex for the index is done in exactly the same way as for columns.