Example usage for org.apache.mahout.classifier.df.node Node classify

Introduction

In this page you can find the example usage for org.apache.mahout.classifier.df.node Node classify.

Prototype

public abstract double classify(Instance instance);

Source Link

Document

predicts the label for the instance

Usage

From source file:com.wsc.myexample.decisionForest.MyDecisionForest.java

License:Apache License

/**
 * Classifies the data and calls callback for each classification
 *//*  w  w w  . java  2 s.  c  om*/
public void classify(Data data, double[] predictions) {
    Preconditions.checkArgument(data.size() == predictions.length,
            "predictions.length must be equal to data.size()");

    if (data.isEmpty()) {
        return; // nothing to classify
    }

    for (Node tree : trees) {
        for (int index = 0; index < data.size(); index++) {
            predictions[index] = tree.classify(data.get(index));
        }
    }
}

---

From source file:com.wsc.myexample.decisionForest.MyDecisionForest.java

License:Apache License

/**
 * predicts the label for the instance//from w w  w  .jav  a2s  .  c o m
 * 
 * @param rng
 *          Random number generator, used to break ties randomly
 * @return -1 if the label cannot be predicted
 */
public double classify(Dataset dataset, Random rng, Instance instance) {
    if (dataset.isNumerical(dataset.getLabelId())) {
        double sum = 0;
        int cnt = 0;
        for (Node tree : trees) {
            double prediction = tree.classify(instance);
            if (prediction != -1) {
                sum += prediction;
                cnt++;
            }
        }
        return sum / cnt;
    } else {
        int[] predictions = new int[dataset.nblabels()];
        for (Node tree : trees) {
            double prediction = tree.classify(instance);
            if (prediction != -1) {
                predictions[(int) prediction]++;
            }
        }

        if (DataUtils.sum(predictions) == 0) {
            return -1; // no prediction available
        }

        return DataUtils.maxindex(rng, predictions);
    }
}

---