Java DoubleMatrix.mul方法代码示例

import org.jblas.DoubleMatrix; //导入方法依赖的package包/类
/**
 * Element-wise derivated sigmoid
 * @param x DoubleMatrix 
 * @return DoubleMatrix sigmoid'()
 */
protected static DoubleMatrix sigmoidDerivate(DoubleMatrix x) {
    if (x != null) {
        DoubleMatrix ones = DoubleMatrix.ones(x.rows, x.columns);
        return x.mul(ones.sub(x));
    }
    return null;
}
 

import org.jblas.DoubleMatrix; //导入方法依赖的package包/类
public double trainOn(MultitaggerTrainingSentence sentence, int i, double learningRate, double cumProbCutoff, int maxK) {
	MultitaggerTrainingItem item = sentence.getItems().get(i);
	int goldRank = item.getGoldIndex();
	if(goldRank < 0) {
		return 0.0;
		// Can't train on missing categories? (or maybe should return single top category)
	}
	double firstProb = first.predict(sentence, i);
	double continueProb = 1.0-firstProb;
	ArrayList<Double> probs = new ArrayList<Double>();
	probs.add(firstProb);
	ArrayList<Double> cumProbs = new ArrayList<Double>();
	cumProbs.add(1.0 - continueProb);
	int k = 1;
	DoubleMatrix category = this.categoryEmbeddings.getVec(item.getCat(0));
	this.history = category.mul(firstProb);
	while(continueProb >= cumProbCutoff && k < maxK) {
		category = this.categoryEmbeddings.getVec(item.getCat(k));
		
		double laterProb = later.predict(sentence, i, k, history);
		probs.add(laterProb);
		continueProb *= (1.0 - laterProb);
		cumProbs.add(1.0 - continueProb);
		this.history.addi(category.mul(cumProbs.get(k)));
		k++;
		
		
	}
	
	return 0.0;
}
 

import org.jblas.DoubleMatrix; //导入方法依赖的package包/类
private void coalesceDown(int currTreeInterval) {
		List<Node> parentLines = mascotInput.get().treeIntervalsInput.get().getLineagesAdded(currTreeInterval);
		if (parentLines.size()!=1){
			System.err.println("to many lineages, while coalescening down");
			System.exit(0);
		}
		Node parentNode = parentLines.get(0);
		
		if (!parentNode.isRoot()){
			DoubleMatrix start = stateProbabilities[parentNode.getNr() - nrSamples];
			DoubleMatrix end = stateProbabilitiesDown[parentNode.getParent().getNr() - nrSamples];
			DoubleMatrix flow = TransitionProbabilities[parentNode.getNr()];
			DoubleMatrix otherSideInfo = end.div(start.transpose().mmul(flow));
			// get rid of NaN from division by 0
			for (int i = 0; i < otherSideInfo.length; i++)
				if (Double.isNaN(otherSideInfo.get(i)))
					otherSideInfo.put(i, 0.0);
				
				
				
			DoubleMatrix conditional = flow.mmul(otherSideInfo);
			conditional = conditional.mul(start);
			stateProbabilitiesDown[parentNode.getNr() - nrSamples] = conditional.div(conditional.sum());
//			if (!(conditional.get(0) >= 0.0 && conditional.get(0)<=1.0))
//				conditional.print();
		}else{
//			DoubleMatrix d1 = stateProbabilities[parentNode.getLeft().getNr() - nrSamples];
//			DoubleMatrix d2 = stateProbabilities[parentNode.getRight().getNr() - nrSamples];
//			DoubleMatrix f1 = TransitionProbabilities[parentNode.getLeft().getNr()];
//			DoubleMatrix f2 = TransitionProbabilities[parentNode.getLeft().getNr()];
//			d1.transpose().mmul(f1).print();
//			d2.transpose().mmul(f2).print();
//			d1.print();
//			d2.print();
//			System.out.println();
//			f1.print();
//			f2.print();
//			System.out.println();
//			stateProbabilities[parentNode.getNr() - nrSamples].print();
//			
//			System.out.println();
//			System.exit(0);
			stateProbabilitiesDown[parentNode.getNr() - nrSamples] = stateProbabilities[parentNode.getNr() - nrSamples];
    	}
	}
 

import org.jblas.DoubleMatrix; //导入方法依赖的package包/类
protected double processCoalEvent(int t, int currTreeInterval) {
  	List<DoubleMatrix> coalVectors = stateProbabilities.getCoalescentVectors(intervals, currTreeInterval);
  	DoubleMatrix pvec1, pvec2;	
  	pvec1 = coalVectors.get(0);
  	pvec2 = coalVectors.get(1);
  	
  	
List<Node> parentLines = intervals.getLineagesAdded(currTreeInterval);
if (parentLines.size() > 1) throw new RuntimeException("Unsupported coalescent at non-binary node");			
//Add parent to activeLineage and initialise parent's state prob vector
Node parentNode = parentLines.get(0);


//Compute parent lineage state probabilities in p				
DoubleMatrix F,Y;
  	Y = ts.getYs()[t];
  	F = ts.getFs()[t];		
  	
  	 if (forgiveYInput.get()) {
		Y.maxi(1.0);
	} else { // however, Y_i > 1e-12 by default
		Y.maxi(1e-12); 
	}
 
  	 double lambda;
   /* Compute Lambda Sum */
DoubleMatrix pa;
   DoubleMatrix pi_Y = pvec1.div(Y);
   pi_Y.reshape(numStates, 1);	    	
   DoubleMatrix pj_Y = pvec2.div(Y);
   pj_Y.reshape(numStates, 1);	
   pa = pi_Y.mul(F.mmul(pj_Y));	    	
   pa.addi(pj_Y.mul(F.mmul(pi_Y)));
   pa.reshape(1,numStates);
   lambda = pa.sum();
   pa.divi(lambda);
			
stateProbabilities.addLineage(parentNode.getNr(),pa);
	
sp[parentNode.getNr() - nrSamples] = (pa);
 
//Remove child lineages
List<Node> coalLines = intervals.getLineagesRemoved(currTreeInterval);
stateProbabilities.removeLineageNr(coalLines.get(0).getNr() );
stateProbabilities.removeLineageNr(coalLines.get(1).getNr()); 
	
if (fsCorrectionsInput.get()) {
	doFiniteSizeCorrections(parentNode,pa);
}

return lambda;
  }