CODEWITHSUNDEEP
pythonpandasdataframe
noobee
noobee

Reputation: 57

Printing Python dictionary as Pandas value_counts for multiple columns side by side

I am new to Data Science with Python, Numpy, and Pandas. So, excuse me if my question is very trivial.

Is there a way to print the value_counts of any two or more columns from a Python dictionary?

For example, right now I am just printing the value_counts of the columns I want like below:

import os
import numpy as np
import pandas as pd
import matplotlib as plot

dict_h = {'a':['APPLE', 'DONUT', 'APPLE', 'APPLE', 'APPLE', 'DONUT', 'PEAR'],
          'b':['PEAR', 'DONUT', 'DONUT', 'DONUT', 'APPLE', 'PEAR', 'DONUT'],
          'c':['APPLE', 'APPLE', 'APPLE','DONUT','DONUT','DONUT','PEAR'],
          'd':['PEAR', 'PEAR', 'PEAR','DONUT','DONUT','DONUT','PEAR']}

print('\n')
print('Orignal Dict:')
print(dict_h)
y = pd.DataFrame.from_dict(dict_h, orient='index')
print('\n')
print(y.describe())
print('\n')
print(y[0].value_counts())
print('\n')
print(y[1].value_counts())

That prints:

...
PEAR     2
APPLE    2
Name: 0, dtype: int64


DONUT    2
PEAR     1
APPLE    1
Name: 1, dtype: int64

but instead, I was hoping to use print like this:

         0    1   
   -------------  
   APPLE 2    1 
   PEAR  2    1      
   DONUT NaN  2
   -------------

Thank you. Drew

Upvotes: 2

Views: 181

Answers (1)

Henry Yik
Henry Yik

Reputation: 22503

To get your desired output, use transpose and melt with crosstab:

s = df.T.melt()

print (pd.crosstab(s["value"], s["variable"]))

variable  0  1  2  3  4  5  6
value                        
APPLE     2  1  2  1  2  0  0
DONUT     0  2  1  3  2  3  1
PEAR      2  1  1  0  0  1  3

Upvotes: 2

Related Questions