org.grouplens.lenskit.transform.quantize

Class QuantizedRatingPredictor

java.lang.Object

org.grouplens.lenskit.basic.AbstractRatingPredictor

org.grouplens.lenskit.transform.quantize.QuantizedRatingPredictor

All Implemented Interfaces:

RatingPredictor

public class QuantizedRatingPredictor
extends AbstractRatingPredictor
implements RatingPredictor

A rating predictor wrapper that quantizes predictions.

Constructor Summary

Constructors

Constructor and Description
QuantizedRatingPredictor(ItemScorer scorer, ItemScorer baseline, Quantizer q) Construct a new quantized predictor.

Method Summary

Modifier and Type	Method and Description
void	predict(long user, MutableSparseVector scores) Predict for items in a vector.

Methods inherited from class org.grouplens.lenskit.basic.AbstractRatingPredictor

predict, predict

Methods inherited from class java.lang.Object

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

Methods inherited from interface org.grouplens.lenskit.RatingPredictor

predict, predict

Constructor Detail

QuantizedRatingPredictor

@Inject
public QuantizedRatingPredictor(@PrimaryScorer
                                        ItemScorer scorer,
                                        @Nullable @BaselineScorer
                                        ItemScorer baseline,
                                        Quantizer q)

Construct a new quantized predictor.

Parameters:

scorer - The item scorer to use.

baseline - A baseline scorer to fall back to.

q - The quantizer.

Method Detail

predict

public void predict(long user,
                    @Nonnull
                    MutableSparseVector scores)

Description copied from interface: RatingPredictor

Predict for items in a vector. The key domain of the provided vector is the items whose predictions are requested, and the predict method sets the values for each item to its predict (or unsets it, if no prediction can be provided). The previous values are discarded.

Specified by:

predict in interface RatingPredictor

Parameters:

user - The user ID.

scores - The prediction output vector. Its key domain is the items to score.