Package org.jblas

Examples of org.jblas.DoubleMatrix$ColumnsAsListView

86
87
88
89
90
91
92
     *
     * @return A DoubleMatrix[3] array of U, S, V such that A = U * diag(S) * V'
     */

public static DoubleMatrix []  jblas_fullSVD( double [][]A) {
    return org.jblas.Singular.fullSVD(new DoubleMatrix(A));
}
View Full Code Here
100
101
102
103
104
105
106
     * @param A
     * @return A DoubleMatrix[3] array of U, S, V such that A = U * diag(S) * V'
     */

public static DoubleMatrix []  jblas_sparseSVD( double [][]A) {
    return org.jblas.Singular.sparseSVD(new DoubleMatrix(A));
}
View Full Code Here
106
107
108
109
110
111
112
}


public static ComplexDoubleMatrix []  jblas_sparseSVD( double [][]Areal, double [][] Aimag) {
    return org.jblas.Singular.sparseSVD(
            new ComplexDoubleMatrix(new DoubleMatrix(Areal),  new DoubleMatrix(Aimag)));
}
View Full Code Here
118
119
120
121
122
123
124
     * @param A DoubleMatrix of dimension m * n
     * @return A min(m, n) vector of singular values.
     */

public static DoubleMatrix jblas_SPDValues(double [][]A) {
    return  org.jblas.Singular.SVDValues(new DoubleMatrix(A));
}
View Full Code Here
130
131
132
133
134
135
136
     * @return A real-valued (!) min(m, n) vector of singular values.
     */

public static DoubleMatrix jblas_SPDValues(double [][]Areal, double [][]Aimag) {
    return  org.jblas.Singular.SVDValues(
            new ComplexDoubleMatrix(new DoubleMatrix(Areal), new DoubleMatrix(Aimag)));
}
View Full Code Here
8
9
10
11
12
13
14
15
16
     * @param M
     * @return e^{\eta_j} / \sum_i e^{\eta_i}
     */
    @Override
    public DoubleMatrix valueAt(DoubleMatrix M) {
        DoubleMatrix exp = MatrixFunctions.exp(M);
        DoubleMatrix sums = exp.columnSums();
        return exp.diviRowVector(sums);
    }
View Full Code Here
19
20
21
22
23
24
25
26
     * @param X input double matrix
     * @return derivative of softmax, has the same formula as sigmoid :)
     */
    @Override
    public DoubleMatrix derivativeAt(DoubleMatrix X) {
        DoubleMatrix M = valueAt(X);
        return M.mul((M.mul(-1)).addi(1));
    }
View Full Code Here
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
    public static boolean check(DifferentiableFunction Func) {
        int size = Func.dimension();
        double p = size;
        int attempts = 10;
        while (attempts > 0) {
            DoubleMatrix xMat = DoubleMatrix.rand(size);
//      DoubleMatrix xMat = attempts != 10 ? DoubleMatrix.rand(size) : DoubleMatrix.ones(size).mul(0.1);
            double[] x = xMat.data;
            double ReturnedCost = Func.valueAt(x);
            double[] ReturnedGradient = Func.derivativeAt(x);
            double[] NumericalGradient = new double[size];
            double PartCosts;

            double Mean = 2e-6 * ((1 + xMat.norm2()) / p);
            for (int i = 0; i < size; i++) {
                double[] e = DoubleMatrix.zeros(size).data;
                e[i] = 1;
                DoubleArrays.scale(e, Mean);
                double[] y = DoubleArrays.add(x, e);
View Full Code Here
17
18
19
20
21
22
23
24
     * @param M input double matrix
     * @return tanh_prime = (1-M.^2);
     */
    @Override
    public DoubleMatrix derivativeAt(DoubleMatrix M) {
        DoubleMatrix Squared = M.mul(M);
        return (Squared.muli(-1)).addi(1);
    }
View Full Code Here
15
16
17
18
19
20
21
22
23
24
25
     *         diag(1-x.^2)./nrm - y*x'./nrm^3
     */
    @Override
    public DoubleMatrix derivativeAt(DoubleMatrix M) {
        double norm = M.norm2();
        DoubleMatrix Squared = M.mul(M);
        DoubleMatrix y = M.sub(Squared.mul(M));
        DoubleMatrix p1 = DoubleMatrix.diag((Squared.mul(-1)).add(1)).divi(norm);
        DoubleMatrix p2 = (y.mmul(M.transpose())).divi(Math.pow(norm, 3));
        return p1.subi(p2);
    }
View Full Code Here
TOP

Related Classes of org.jblas.DoubleMatrix$ColumnsAsListView

  • com.github.pmerienne.trident.ml.classification.AROWClassifier
  • com.numb3r3.common.opt.GradientChecker
  • com.numb3r3.common.opt.Norm1Tanh
  • com.numb3r3.common.opt.Sigmoid
  • com.numb3r3.common.opt.Softmax
  • com.numb3r3.common.opt.Tanh
  • org.apache.spark.mllib.recommendation.JavaALSSuite
  • org.apache.spark.mllib.regression.JavaRidgeRegressionSuite
  • org.gd.spark.opendl.downpourSGD.Backpropagation.AutoEncoder
  • org.gd.spark.opendl.downpourSGD.Backpropagation.AutoEncoder$AEOptimizer
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.