Package com.datarobot.prediction

Interface IClassificationPredictor

All Superinterfaces:

IEventProvider<ClassificationScoreEventInfo>, IPredictorInfo

public interface IClassificationPredictor
extends IPredictorInfo, IEventProvider<ClassificationScoreEventInfo>

Interface for multi class and binary classification models.

Method Summary

Modifier and Type	Method	Description
String[]	getClassLabels()	Model specific class labels that will be predicted.
Map<String,Double>	score(Iterable<?> row)	Scores a single row.
Map<String,Double>	score(Map<String,?> row)	Scores a single row.
Score<Map<String,Double>>	scoreWithExplanations(Iterable<?> row)	Scoring with the default prediction explanations parameters.
Score<Map<String,Double>>	scoreWithExplanations(Iterable<?> row, ExplanationParams predictionExplanationParams)	Scoring with some overridden prediction explanation parameters.
Score<Map<String,Double>>	scoreWithExplanations(Map<String,?> row)	Scoring with the default prediction explanations parameters.
Score<Map<String,Double>>	scoreWithExplanations(Map<String,?> row, ExplanationParams predictionExplanationParams)	Scoring with some overridden prediction explanation parameters.

Methods inherited from interface com.datarobot.prediction.IEventProvider

getScoreEvent

Methods inherited from interface com.datarobot.prediction.IPredictorInfo

getDefaultPredictionExplanationParams, getFeatures, getModelId, getModelInfo, getPredictorClass

Method Detail

getClassLabels

String[] getClassLabels()

Model specific class labels that will be predicted.

NOTE:

These labels will be used as a key set for outputs of score(Iterable) and score(Map).

Returns:

array of class labels.

score

Map<String,Double> score(Map<String,?> row)

Scores a single row.

Parameters:

row - row with feature values.

Returns:

class label to its probability.

score

Map<String,Double> score(Iterable<?> row)

Scores a single row.

NOTE:

Here Iterable<?> row is the feature values in the same order as they goes in IPredictorInfo.getFeatures().

Parameters:

row - row with feature values in right order.

Returns:

class label to its probability.

scoreWithExplanations

Score<Map<String,Double>> scoreWithExplanations(Map<String,?> row)

Scoring with the default prediction explanations parameters.

Returns:

scores with explanations.

scoreWithExplanations

Score<Map<String,Double>> scoreWithExplanations(Iterable<?> row)

Scoring with the default prediction explanations parameters.

Returns:

scores with explanations.

scoreWithExplanations

Score<Map<String,Double>> scoreWithExplanations(Map<String,?> row,
                                                      ExplanationParams predictionExplanationParams)

Scoring with some overridden prediction explanation parameters.

Parameters:

predictionExplanationParams - prediction explanations parameters to override.

Returns:

scores with explanations.

scoreWithExplanations

Score<Map<String,Double>> scoreWithExplanations(Iterable<?> row,
                                                      ExplanationParams predictionExplanationParams)

Scoring with some overridden prediction explanation parameters.

Parameters:

predictionExplanationParams - prediction explanations parameters to override.

Returns:

scores with explanations.