Module stat_selectors 

Statistical-test-based feature selectors.

Three selectors that choose features based on p-values obtained from a statistical test (e.g., ANOVA F-test, chi-squared test):

• SelectFpr — False Positive Rate: selects every feature whose p-value is below alpha.
• SelectFdr — False Discovery Rate: applies the Benjamini-Hochberg procedure to control the expected proportion of false positives.
• SelectFwe — Family-Wise Error: applies the Bonferroni correction (alpha / n_features) to control the probability of any false positive.

All three take a pre-computed vector of p-values (one per feature) at fit time, allowing integration with any upstream scoring function.

Structs

FittedSelectFdr

A fitted SelectFdr holding the selected indices.

FittedSelectFpr

A fitted SelectFpr holding the selected indices.

FittedSelectFwe

A fitted SelectFwe holding the selected indices.

SelectFdr

Select features controlling the false discovery rate via the Benjamini-Hochberg procedure.

SelectFpr

Select features with p-values below alpha.

SelectFwe

Select features controlling the family-wise error rate via the Bonferroni correction.