Implementation:Scikit learn Scikit learn VarianceThreshold

Knowledge Sources	Scikit_learn Scikit-learn Docs
Domains	Machine Learning, Feature Selection
Last Updated	2026-02-08 15:00 GMT

Overview

Concrete tool for removing low-variance features from datasets, provided by scikit-learn.

Description

The VarianceThreshold class is a feature selector that removes all features whose variance does not meet a specified threshold. It operates only on features (X) and not on targets (y), making it suitable for unsupervised feature selection. It supports both dense and sparse input matrices and allows NaN values.

Usage

Use this selector as a simple baseline feature selection step to remove constant or near-constant features before applying more sophisticated feature selection or modeling techniques.

Code Reference

Source Location

• Repository: scikit-learn
• File: sklearn/feature_selection/_variance_threshold.py

Signature

class VarianceThreshold(SelectorMixin, BaseEstimator):
    def __init__(self, threshold=0.0):

Import

from sklearn.feature_selection import VarianceThreshold

I/O Contract

Inputs

Name	Type	Required	Description
threshold	float	No	Variance threshold; features below this are removed (default 0.0)
X	array-like of shape (n_samples, n_features)	Yes	Training data for fitting

Outputs

Name	Type	Description
variances_	ndarray of shape (n_features,)	Variances of individual features
n_features_in_	int	Number of features seen during fit
X_transformed	ndarray	Reduced feature matrix with low-variance features removed

Usage Examples

Basic Usage

from sklearn.feature_selection import VarianceThreshold

X = [[0, 0, 1], [0, 1, 0], [1, 0, 0], [0, 1, 1], [0, 1, 0], [0, 1, 1]]
sel = VarianceThreshold(threshold=0.16)
X_selected = sel.fit_transform(X)
print(X_selected.shape)  # Features with variance > 0.16 are kept

Related Pages

• Principle:Scikit_learn_Scikit_learn_Feature_Selection