CODEWITHSUNDEEP
pythonpandas
emilaz
emilaz

Reputation: 2082

Pandas compare each row to list

Given a pandas dataframe like the following

    Q1  Q2  Q3
0   'A' 'B' 'A'
1   'B' 'B' 'C'

and a list containing values I want to compare against

['B','B','A']

How do I obtain for each row how many column entries coincided with the respective list entry? So the output should be some form of

[2/3, 2/3]

Little code snippet for the above:

import pandas as pd
df = pd.DataFrame({'Q1':['A','B'],'Q2':['B','B'],'Q3':['A','C']})
true_answers = ['B','B','A']

Upvotes: 1

Views: 229

Answers (1)

jezrael
jezrael

Reputation: 863611

Compare values by list (it has to be same length like number of columns) and get mean using DataFrame.eq method:

s = df.eq(true_answers).mean(axis=1)
print (s)
0    0.666667
1    0.666667
dtype: float64

If want list use:

print (list(s))
[0.6666666666666666, 0.6666666666666666]

Detail:

print (df.eq(true_answers))
      Q1    Q2     Q3
0  False  True   True
1   True  True  False

Upvotes: 4

Related Questions