bernoulliestimatortest.java

一个自然语言处理的Java开源工具包。LingPipe目前已有很丰富的功能

package com.aliasi.test.unit.stats;

import com.aliasi.stats.BernoulliDistribution;
import com.aliasi.stats.BernoulliEstimator;

import com.aliasi.test.unit.BaseTestCase;

import com.aliasi.util.AbstractExternalizable;

import java.io.*;

public class BernoulliEstimatorTest extends BaseTestCase {

    public void testOne() throws IOException, ClassNotFoundException {
        BernoulliEstimator coin = new BernoulliEstimator();
        coin.train(true);
        coin.train(true);
        coin.train(false);
        coin.train(true,1);
        coin.train(true,5);
        coin.train(false,7);
        assertFairCoin(coin);


        assertTrue(AbstractExternalizable.compile(coin)
                   instanceof BernoulliDistribution);

        coin.train(true,8);
        assertEquals(2.0/3.0,coin.successProbability(),0.0001);
        assertEquals(1.0/3.0,coin.probability(0l),0.0001);
    }


    static void assertFairCoin(BernoulliDistribution coin) {
        assertEquals(0l,coin.minOutcome());
        assertEquals(1l,coin.maxOutcome());
        assertEquals(2,coin.numDimensions());
        assertEquals("failure",coin.label(0l));
        assertEquals("success",coin.label(1l));
        assertEquals(0.50,coin.successProbability(),0.0001);
        assertEquals(0.50,coin.probability(0l),0.0001);
        assertEquals(0.50,coin.probability(1l),0.0001);

        assertEquals(0.00,coin.cumulativeProbabilityLess(-1l),0.0001);
        assertEquals(0.50,coin.cumulativeProbabilityLess(0l),0.0001);
        assertEquals(1.00,coin.cumulativeProbabilityLess(1l),0.0001);
        assertEquals(1.00,coin.cumulativeProbabilityLess(2l),0.0001);
        assertEquals(1.00,coin.cumulativeProbabilityGreater(-1l),0.0001);
        assertEquals(1.00,coin.cumulativeProbabilityGreater(0l),0.0001);
        assertEquals(0.50,coin.cumulativeProbabilityGreater(1l),0.0001);
        assertEquals(0.00,coin.cumulativeProbabilityGreater(2l),0.0001);
        assertEquals(0.00,coin.cumulativeProbability(-2l,-1l),0.0001);
        assertEquals(0.00,coin.cumulativeProbability(3l,120l),0.0001);
        assertEquals(0.50,coin.cumulativeProbability(-1l,0l),0.0001);
        assertEquals(0.50,coin.cumulativeProbability(1l,1l),0.0001);
        assertEquals(0.50,coin.cumulativeProbability(1l,15l),0.0001);
        assertEquals(1.00,coin.cumulativeProbability(0l,1l),0.0001);
        assertEquals(1.00,coin.cumulativeProbability(0l,14l),0.0001);
        assertEquals(1.00,coin.cumulativeProbability(-12l,14l),0.0001);

        assertEquals(1.0,coin.entropy(),0.0001);
        assertEquals(0.5,coin.mean(),0.0001);
        assertEquals((0.5 * 0.5 + 0.5*0.5)/2.0,coin.variance(),0.0001);
    }
}