Simplify Boolean Indexing Conditions in Pandas

Question

I have two columns which either contain a "Yes" or "No" value:

• Partner
• Dependents

I originally wrote this code to generate another column based off the value in those two columns (see below):

conditions = [
    # either dependents, partner, or both 
    ((cleaned_df["Partner"] == "Yes") & (cleaned_df["Dependents"] == "Yes")) | ((cleaned_df["Partner"] == "No") & (cleaned_df["Dependents"] == "Yes")) | ((cleaned_df["Partner"] == "Yes") & (cleaned_df["Dependents"] == "No")),
    # neither partner nor dependents
    (cleaned_df["Partner"] == "No") & (cleaned_df["Dependents"] == "No")
]

However, the first condition is a bit wordy and I was wondering if there was a more eloquent way to rewrite this code. Thanks in advance!

Answer 1

I will take a minimum example:

import pandas as pd
data = {'Partner' : ['Yes','No','Yes','No'] , 'Dependents': ['No','No','Yes','Yes']}
df = pd.DataFrame(data)

print(df)
Partner Dependents
0   Yes No
1   No  No
2   Yes Yes
3   No  Yes

You can simply do:

(df["Partner"] == "Yes") | (df["Dependents"] == "Yes")

0     True
1    False
2     True
3     True
dtype: bool

Explanation: If any one is True then it will be True, if both of them are False then it will be False.