CODEWITHSUNDEEP
pythonpython-3.xpandas
Chris90
Chris90

Reputation: 1998

Value Counts with Condition?

How can I use a filter for if a column is = for a value counts?

For example

If I have df column below as:

Column
Happy
Sad
Mad
Glad
Happy 
Mad
Glad 
Glad

I want to run a df['Column'].value_counts(normalize = True)

But I only want to see the % for Glad, this is a sample df, but the actual df as many volume values and instead of sifting through I just want to see the % for Glad for example. Thanks!

Upvotes: 1

Views: 1001

Answers (2)

David Erickson
David Erickson

Reputation: 16683

Just add .loc['Glad'] to the end of your code, because value_counts() creates a pandas series allowing you to filter for Glad using loc:

df['Column'].value_counts(normalize=True).loc['Glad']
0.375

OR you can use shape[0] to get total rows of filtered dataframe for 'Glad' values as a proportion of total rows of the full dataframe:

df[df['Column'] == 'Glad'].shape[0] / df.shape[0]
0.375

Upvotes: 3

RavinderSingh13
RavinderSingh13

Reputation: 133458

Could you please try following, written as per your shown samples, this will give all column's percentage here.

import pandas as pd
df['Column'].value_counts(normalize=True) * 100

OR taking inspiration from David answer to get string Glad's percentage only use .loc for above solution.

df['Column'].value_counts(normalize=True).loc['Glad'] * 100

Upvotes: 2

Related Questions