Module iterative_imputer 

Iterative imputer: fill missing values by modeling each feature as a function of all other features.

IterativeImputer performs round-robin imputation: for each feature with missing values, it fits a simple Ridge regression on the non-missing rows using the other features as predictors, then predicts the missing values. This process is repeated for max_iter iterations or until convergence.

Initial Imputation

Before the iterative process begins, missing values are filled using a simple strategy (mean by default). This initial imputation provides a starting point for the regression models.

Structs

FittedIterativeImputer

A fitted iterative imputer that stores per-feature Ridge regression coefficients learned during fitting.

IterativeImputer

An unfitted iterative imputer.

Enums

InitialStrategy

Strategy for the initial imputation before iterative refinement.