How to Use var in pandas for Variance Calculation

In pandas, you use the var() method on a DataFrame or Series to calculate the variance of numeric data. This method computes the sample variance by default, which measures how spread out the data is around the mean.

Syntax

The var() method can be called on a pandas DataFrame or Series. It has optional parameters to control the calculation.

• axis: Choose whether to calculate variance across rows (axis=1) or columns (axis=0, default).
• ddof: Delta degrees of freedom. Default is 1 for sample variance.
• numeric_only: If True, only include numeric data.

python

DataFrame.var(axis=0, ddof=1, numeric_only=None, **kwargs)

Series.var(ddof=1, **kwargs)

Example

This example shows how to calculate variance for each numeric column in a DataFrame using var(). It also shows variance for a single Series.

python

import pandas as pd

data = {
    'math': [90, 80, 70, 60, 85],
    'english': [88, 92, 85, 87, 90],
    'history': [75, 78, 80, 72, 70]
}
df = pd.DataFrame(data)

# Variance of each column
variance_df = df.var()

# Variance of math scores only
variance_math = df['math'].var()

print('Variance of each subject:')
print(variance_df)
print('\nVariance of math scores:')
print(variance_math)

Output

Variance of each subject:
math       112.5
english     7.7
history    13.5
dtype: float64

Variance of math scores:
112.5

Common Pitfalls

Common mistakes when using var() include:

• Forgetting that var() calculates sample variance by default (ddof=1), not population variance.
• Applying var() on non-numeric columns without setting numeric_only=True, which can cause errors.
• Misunderstanding the axis parameter, which changes the direction of calculation.

python

import pandas as pd

data = {'A': [1, 2, 3], 'B': ['x', 'y', 'z']}
df = pd.DataFrame(data)

# This will raise an error because column B is non-numeric
# df.var()

# Correct way: ignore non-numeric columns
variance = df.var(numeric_only=True)
print(variance)

Output

A    1.0
dtype: float64

Quick Reference

Summary tips for using var() in pandas:

• Use df.var() to get variance of each numeric column.
• Use series.var() for variance of a single column.
• Set ddof=0 for population variance.
• Use numeric_only=True to avoid errors with non-numeric data.
• Use axis=1 to calculate variance across rows instead of columns.

Key Takeaways

Use pandas var() method to calculate variance of numeric data in DataFrames or Series.

By default, var() calculates sample variance with ddof=1.

Set numeric_only=True to avoid errors from non-numeric columns.

Use axis parameter to control calculation direction (columns or rows).

Set ddof=0 to calculate population variance if needed.