CODEWITHSUNDEEP
pythonpandasdataframe
RRB
RRB

Reputation: 1

Pandas: Counts of rows with any of n True boolean columns values

df_con = df[['Traffic_Signal', 'Crossing', 'Junction', 'Bump', 'Railway']]

I have a data frame of boolean values. I want to create a count of total cases where any of the above columns are also True.

I started looping through the df with iterrows, but I'm getting errors I'm unfamiliar with...any guidance is appreciated!

This was my closest (I think) attempt:

df_con = df[['Traffic_Signal', 'Crossing', 'Junction', 'Bump', 'Railway']]

cols = ['Traffic_Signal', 'Crossing', 'Junction', 'Bump', 'Railway']

count = 0
for index, row in df_con.iterrows():
    for col in cols:
        if df_con.loc[col] == True:
            count += 1
            break
print(count)

Upvotes: 0

Views: 454

Answers (1)

kmiller96
kmiller96

Reputation: 329

I'd recommend that you use the .any() method to get a boolean for that row and then .sum() that result to get your final count.

Here is a very simple example using my own columns and data.

import pandas as pd

df = pd.DataFrame(
    {
        "excluded": [0, 1, 2, 3],
        "x": [True, True, True, False],
        "y": [True, False, False, False],
        "z": [True, True, False, False],
    }
)


dff = df[["x", "y", "z"]]
count = dff.any(axis="columns").sum()

print(count)  # Gives expected result of 3

Upvotes: 1

Related Questions