How do I count the NaN values in a column in pandas DataFrame?

In Pandas, NaN (Not a Number) values often indicate missing or undefined data. To count how many NaNs exist in a particular column, you can combine isna() (or isnull()) and sum():

import pandas as pd
import numpy as np

df = pd.DataFrame({
    "A": [1, 2, np.nan, 4],
    "B": [np.nan, 5, 6, np.nan]
})

# Count the NaNs in column 'A'
nan_count_A = df["A"].isna().sum()
print("Number of NaNs in column A:", nan_count_A)

# Count the NaNs in column 'B'
nan_count_B = df["B"].isna().sum()
print("Number of NaNs in column B:", nan_count_B)

1. df["A"].isna() creates a boolean Series that marks where values in column "A" are NaN (True) or not (False).
2. .sum() on this boolean Series adds up the True values (counting them as 1 each), giving the total number of NaNs.

Counting NaNs in All Columns

If you ever need to count NaNs across all columns at once:

df.isna().sum()

This returns a Series with each column name and its corresponding NaN count.

Level Up Your Python & Data Skills

To handle missing data effectively in larger projects, you’ll benefit from a strong foundation in Python. Grokking Python Fundamentals by DesignGurus.io can help reinforce essential Python concepts, coding best practices, and more advanced techniques, ensuring you’re well-equipped for data manipulation and analysis.

With these approaches, counting and handling NaN values in your DataFrame becomes a straightforward step in your data cleaning workflow!