Missing-value imputation#
scikitplot.impute combines standard scikit-learn-style imputation surfaces with an experimental approximate-nearest-neighbor imputer for larger or index-oriented workflows.
Important
ANNImputer is experimental and is not imported into the public namespace until you explicitly enable it. Approximate-neighbor imputation changes the computational trade-off; it does not guarantee the same neighbors or values as exact KNN imputation.
Use stable sklearn-compatible imputers or explicitly enable ANNImputer.
Understand the re-exported MissingIndicator and SimpleImputer surface.
Approximate-neighbor imputation with Annoy or Voyager backends.
Choose private/public/external index ownership and understand data exposure.
Stable imports, experimental activation, fit/transform behavior, and feature names.
Approximation, fallback behavior, optional backends, and current review findings.
Which page should I use?#
Start with Getting started. Then choose the page that owns the decision you are making: architecture and lifecycle, configuration, persistence, performance, public API, or limits. The guide intentionally separates those concerns so examples do not accidentally become policy.
Examples#
See Impute for executable examples that accompany this guide.
See also#
Further reading#
Voyager is optimized for modern nearest-neighbor search.
HNSW (Hierarchical Navigable Small World) provides better accuracy and speed,
outperforming Annoy in most use cases, especially when precision is important.
See also