Low Variance¶

skfeature.function.statistical_based.low_variance

Description¶

Low Variance removes features whose variance falls below a threshold. It is a simple unsupervised filter useful for dropping constant or near-constant features before applying other methods.

Usage¶

import numpy as np
from sklearn.datasets import load_iris

from skfeature.function.statistical_based import low_variance

X, y = load_iris(return_X_y=True)

# drop features with variance below the threshold
X_selected = low_variance.low_variance_feature_selection(X, threshold=0.5)

Parameters¶

X: numpy array, shape (n_samples, n_features) — input data
threshold: float, default 0.0 — minimum variance a feature must have to be kept
mode: {"rank", "index"}, default "rank"

Returns¶

X_selected: transformed data with low-variance features removed

References¶

Based on sklearn.feature_selection.VarianceThreshold.