Search code examples
pythonpython-3.xpandaspython-3.5pweave

print(tabulate(...)) to pretty print multiIndex pandas ?


Say i have a pandas data frame:

import numpy as np
import pandas as pd
from tabulate import tabulate

A = pd.DataFrame(np.random.randint(0,10,(3,6)), index= ['uno', 'dos', 'tres'])
A.columns = ['A','B','C','D','E','F']
A.index.names = ['type']
A.columns.names= ['group']

h = [A.index.names[0] +'/'+ A.columns.names[0]] + list(A.columns)
print(tabulate(A, headers= h, tablefmt= 'grid'))

which gives:

+--------------+-----+-----+-----+-----+-----+-----+
| type/group   |   A |   B |   C |   D |   E |   F |
+==============+=====+=====+=====+=====+=====+=====+
| uno          |   3 |   1 |   6 |   0 |   7 |   0 |
+--------------+-----+-----+-----+-----+-----+-----+
| dos          |   9 |   5 |   3 |   0 |   6 |   6 |
+--------------+-----+-----+-----+-----+-----+-----+
| tres         |   6 |   7 |   4 |   6 |   8 |   4 |
+--------------+-----+-----+-----+-----+-----+-----+

Now adding a layer:

iterable = [['A', 'B'], ['AA', 'BB', 'CC']]
A.columns = pd.MultiIndex.from_product(iterable,
                                       names= ['group', 'subgroup'])
A.index.names = ['type']

will give using a print statement:

group     A        B      
subgroup AA BB CC AA BB CC
type                      
uno       3  1  6  0  7  0
dos       9  5  3  0  6  6
tres      6  7  4  6  8  4

Typically, this does not present well on a document such as pweave.

How can i use print(tabulate(...)) such as i have each group and sub group to shown on separate line ?

Thanks


Solution

  • With tabulate version 0.8.1 or newer,

    import numpy as np
    import pandas as pd
    from tabulate import tabulate
    
    A = pd.DataFrame(np.random.randint(0,10,(3,6)), index= ['uno', 'dos', 'tres'])
    A.columns = ['A','B','C','D','E','F']
    A.index.names = ['type']
    A.columns.names= ['group']
    
    iterable = [['A', 'B'], ['AA', 'BB', 'CC']]
    A.columns = pd.MultiIndex.from_product(iterable, names= ['group', 'subgroup'])
    
    h = [A.index.names[0] +'/'+ A.columns.names[0]] + list(map('\n'.join, A.columns.tolist()))
    print(tabulate(A, headers= h, tablefmt= 'grid'))
    

    yields

    +--------------+------+------+------+------+------+------+
    | type/group   |    A |    A |    A |    B |    B |    B |
    |              |   AA |   BB |   CC |   AA |   BB |   CC |
    +==============+======+======+======+======+======+======+
    | uno          |    8 |    4 |    3 |    3 |    4 |    2 |
    +--------------+------+------+------+------+------+------+
    | dos          |    1 |    1 |    1 |    9 |    8 |    9 |
    +--------------+------+------+------+------+------+------+
    | tres         |    0 |    4 |    3 |    6 |    1 |    7 |
    +--------------+------+------+------+------+------+------+