org.apache.storm.pmml.runner

Interface PmmlModelRunner<I,R,P,S>

Type Parameters:

I - type of the input source. For Storm it is typically a Tuple

R - the type of extracted raw input

P - the type of preprocessed input

S - the type of predicted scores

All Superinterfaces:

ModelRunner

All Known Implementing Classes:

JPmmlModelRunner

public interface PmmlModelRunner<I,R,P,S>
extends ModelRunner

Runner for models defined using PMML.

Method Summary

Modifier and Type	Method and Description
R	extractRawInputs(I input) Extracts from the tuple the raw inputs that are to be scored according to the predictive model.
S	predictScores(P preProcInputs) Compute the predicted scores from the pre-processed inputs in the step above.
P	preProcessInputs(R rawInputs) Pre process inputs, i.e., remove missing fields, outliers, etc

Methods inherited from interface org.apache.storm.pmml.runner.ModelRunner

scoredTuplePerStream

Method Detail

extractRawInputs

R extractRawInputs(I input)

Extracts from the tuple the raw inputs that are to be scored according to the predictive model.

Parameters:

input - source from which to extract raw inputs

Returns:

raw inputs

preProcessInputs

P preProcessInputs(R rawInputs)

Pre process inputs, i.e., remove missing fields, outliers, etc

Parameters:

rawInputs - that are to be preprocessed

Returns:

preprocessed output

predictScores

S predictScores(P preProcInputs)

Compute the predicted scores from the pre-processed inputs in the step above.

Parameters:

preProcInputs - that are to be preprocessed

Returns:

predicted scores