Missing data in pandas.crosstab

Question

I\'m making some crosstabs with pandas:

a = np.array([\'foo\', \'foo\', \'foo\', \'bar\', \'bar\', \'foo\', \'foo\'], dtype=object)
b = np.array([\'one\', \'one\         


        

Answer 1

The crosstab function has a parameter called dropna which is set to True by default. This parameter defines whether empty columns (such as the one-shiny column) should be displayed or not.

I tried calling the funcion like this:

pd.crosstab(a, [b, c], rownames=['a'], colnames=['b', 'c'], dropna = False)

and this is what I got:

b     one          two       
c    dull  shiny  dull  shiny
a                            
bar     1      0     1      0
foo     2      0     1      2

Hope that was still helpful.

Answer 2

I don't think there is a way to do this, and crosstab calls pivot_table in the source, which doesn't seem to offer this either. I raised it as an issue here.

A hacky workaround (which may or may not be the same as you were already using...):

from itertools import product
ct = pd.crosstab(a, [b, c], rownames=['a'], colnames=['b', 'c'])
a_x_b = list(product(np.unique(b), np.unique(c)))
a_x_b = pd.MultiIndex.from_tuples(a_x_b)

In [15]: ct.reindex_axis(a_x_b, axis=1).fillna(0)
Out[15]:
      one          two
     dull  shiny  dull  shiny
a
bar     1      0     1      0
foo     2      0     1      2

If product is too slow, here is a numpy implementation of it.