package modelkit

You can search for identifiers within the package.

in-package search v0.2.0

Portable classical machine learning workflows for OCaml

Install

dune-project

Dependency

github.com Readme Changelog Edit opam file Versions (4)

Authors

A

Asara

Maintainers

A

Asara developers <devs@asara.io>

Sources

modelkit-0.5.0.tbz

sha256=1fe8fa7c7f904dd098a21a2ca30fd69230750b8cf8aa2ae481b97a16531b47b4

sha512=c946cd1ac014726f4d21791e14d806a205680e6edfa2f80ed8d3680f24f48ca3c5a2f89d5afa68c11eac4053ead448b6381fb3d06dc8ff129097af6c69b262fb

doc/modelkit/Modelkit/Column_transformer/Supervised/index.html

Module `Column_transformer.Supervised`

Target-aware composition for Pipeline.Supervised pipelines. All active children receive the same training targets in row order; sample weights retain each child's opt-in policy. Adapt ordinary stages with Pipeline.Supervised.unsupervised. The unsupervised composition's naming, allocation, empty-input, and deterministic seed rules apply. Target and weight length errors fail before any pipeline stage fits. Inference uses fitted values without targets.

type 'kind t

type 'kind branch

val transformer : 
  columns:Column_selector.t ->
  'kind Pipeline.Supervised.stage ->
  'kind branch

val passthrough : 
  name:string ->
  columns:Column_selector.t ->
  ('kind branch, Error.t) result

val drop : 
  name:string ->
  columns:Column_selector.t ->
  ('kind branch, Error.t) result

val create : 
  ?remainder:remainder ->
  ?max_output_features:int ->
  'kind branch array ->
  ('kind t, Error.t) result

val stage : 
  name:string ->
  'kind t ->
  ('kind Pipeline.Supervised.stage, Error.t) result