org.apache.ignite.ml.pipeline

Class Pipeline<K,V,R>

java.lang.Object

org.apache.ignite.ml.pipeline.Pipeline<K,V,R>

Type Parameters:

K - Type of a key in upstream data.

V - Type of a value in upstream data.

R - Type of a result in upstream feature extractor.

public class Pipeline<K,V,R>
extends Object

A simple pipeline, which acts as a global trainer which produce a Pipeline Model. A Pipeline consists of a sequence of stages, each of which is either a Preprocessing Stage or a Trainer. When fit() method is called, the stages are executed in order.

Constructor Summary

Constructors

Constructor and Description
Pipeline()

Method Summary

Modifier and Type	Method and Description
Pipeline<K,V,R>	addFeatureExtractor(IgniteBiFunction<K,V,R> featureExtractor) Adds feature extractor as a zero stage.
Pipeline<K,V,R>	addLabelExtractor(IgniteBiFunction<K,V,Double> lbExtractor) Adds a label extractor for the produced model.
Pipeline<K,V,R>	addPreprocessor(PreprocessingTrainer preprocessor) Adds a preprocessor.
Pipeline<K,V,R>	addTrainer(DatasetTrainer trainer) Adds a trainer.
PipelineMdl<K,V>	fit(Ignite ignite, IgniteCache<K,V> cache) Fits the pipeline to the input cache.
PipelineMdl<K,V>	fit(Map<K,V> data, int parts) Fits the pipeline to the input mock data.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

Pipeline

public Pipeline()

Method Detail

addFeatureExtractor

public Pipeline<K,V,R> addFeatureExtractor(IgniteBiFunction<K,V,R> featureExtractor)

Adds feature extractor as a zero stage.

Parameters:

featureExtractor - The parameter value.

Returns:

The updated Pipeline.

addLabelExtractor

public Pipeline<K,V,R> addLabelExtractor(IgniteBiFunction<K,V,Double> lbExtractor)

Adds a label extractor for the produced model.

Parameters:

lbExtractor - The parameter value.

Returns:

The updated Pipeline.

addPreprocessor

public Pipeline<K,V,R> addPreprocessor(PreprocessingTrainer preprocessor)

Adds a preprocessor.

Parameters:

preprocessor - The parameter value.

Returns:

The updated Pipeline.

addTrainer

public Pipeline<K,V,R> addTrainer(DatasetTrainer trainer)

Adds a trainer.

Parameters:

trainer - The parameter value.

Returns:

The updated Pipeline.

fit

public PipelineMdl<K,V> fit(Ignite ignite,
                            IgniteCache<K,V> cache)

Fits the pipeline to the input cache.

Parameters:

ignite - Ignite instance.

cache - Ignite cache with upstream data.

Returns:

The fitted model based on chain of preprocessors and final trainer.

fit

public PipelineMdl<K,V> fit(Map<K,V> data,
                            int parts)

Fits the pipeline to the input mock data.

Parameters:

data - Data.

parts - Number of partitions.

Returns:

The fitted model based on chain of preprocessors and final trainer.