SimplePredictiveModelOptimizerBuilder.java

package quickml.supervised.predictiveModelOptimizer;

import com.google.common.base.Preconditions;
import quickml.data.instances.Instance;
import quickml.supervised.PredictiveModel;
import quickml.supervised.PredictiveModelBuilder;
import quickml.supervised.crossValidation.SimpleCrossValidator;
import quickml.supervised.crossValidation.LossChecker;
import quickml.supervised.crossValidation.data.TrainingDataCycler;

import java.util.Map;

public class SimplePredictiveModelOptimizerBuilder<PM extends PredictiveModel, T extends Instance> {

    private Map<String, ? extends FieldValueRecommender> valuesToTest;
    private PredictiveModelBuilder modelBuilder;
    private TrainingDataCycler<T> dataCycler;
    private LossChecker<PM, T> lossChecker;
    private int iterations = 5;


    public SimplePredictiveModelOptimizerBuilder<PM, T> valuesToTest(Map<String, ? extends FieldValueRecommender> valuesToTest) {
        this.valuesToTest = valuesToTest;
        return this;
    }

    public SimplePredictiveModelOptimizerBuilder<PM, T> lossChecker(LossChecker<PM, T> lossChecker) {
        this.lossChecker = lossChecker;
        return this;
    }

    public SimplePredictiveModelOptimizerBuilder<PM, T> dataCycler(TrainingDataCycler<T> dataCycler) {
        this.dataCycler = dataCycler;
        return this;
    }

    public SimplePredictiveModelOptimizerBuilder<PM, T> modelBuilder(PredictiveModelBuilder modelBuilder) {
        this.modelBuilder = modelBuilder;
        return this;
    }

    public SimplePredictiveModelOptimizerBuilder<PM, T> iterations(int iterations) {
        this.iterations = iterations;
        return this;
    }

    public PredictiveModelOptimizer build() {
        Preconditions.checkArgument(modelBuilder != null, "You must supply a model builder");
        Preconditions.checkArgument(dataCycler != null, "You must supply a data cycler");
        Preconditions.checkArgument(lossChecker != null, "You must supply a loss checker");
        Preconditions.checkArgument(valuesToTest != null, "You must supply a map of configurations to check");

        return new PredictiveModelOptimizer(valuesToTest, new SimpleCrossValidator<>(modelBuilder, lossChecker, dataCycler), iterations);
    }

}