Examples with SingularMatrixException org.apache.ignite.ml.math.exceptions.math.SingularMatrixException used on opensource projects

Example 1 with SingularMatrixException

use of org.apache.ignite.ml.math.exceptions.math.SingularMatrixException in project ignite by apache.

the class LUDecomposition method solve.

/**
 * @param b Vector to solve using this decomposition.
 * @return Solution vector.
 */
public Vector solve(Vector b) {
    final int m = pivot.size();
    if (b.size() != m)
        throw new CardinalityException(b.size(), m);
    if (singular)
        throw new SingularMatrixException();
    final double[] bp = new double[m];
    // Apply permutations to b
    for (int row = 0; row < m; row++) bp[row] = b.get((int) pivot.get(row));
    // Solve LY = b
    for (int col = 0; col < m; col++) {
        final double bpCol = bp[col];
        for (int i = col + 1; i < m; i++) bp[i] -= bpCol * lu.get(i, col);
    }
    // Solve UX = Y
    for (int col = m - 1; col >= 0; col--) {
        bp[col] /= lu.get(col, col);
        final double bpCol = bp[col];
        for (int i = 0; i < col; i++) bp[i] -= bpCol * lu.get(i, col);
    }
    return b.like(m).assign(bp);
}

Example 2 with SingularMatrixException

use of org.apache.ignite.ml.math.exceptions.math.SingularMatrixException in project ignite by apache.

the class GmmTrainer method updateModel.

/**
 * Gets older model and returns updated model on given data.
 *
 * @param dataset Dataset.
 * @param model Model.
 * @return Updated model.
 */
@NotNull
private UpdateResult updateModel(Dataset<EmptyContext, GmmPartitionData> dataset, GmmModel model) {
    boolean isConverged = false;
    int countOfIterations = 0;
    double maxProbInDataset = Double.NEGATIVE_INFINITY;
    while (!isConverged) {
        MeanWithClusterProbAggregator.AggregatedStats stats = MeanWithClusterProbAggregator.aggreateStats(dataset, countOfComponents);
        Vector clusterProbs = stats.clusterProbabilities();
        Vector[] newMeans = stats.means().toArray(new Vector[countOfComponents]);
        A.ensure(newMeans.length == model.countOfComponents(), "newMeans.size() == count of components");
        A.ensure(newMeans[0].size() == initialMeans[0].size(), "newMeans[0].size() == initialMeans[0].size()");
        List<Matrix> newCovs = CovarianceMatricesAggregator.computeCovariances(dataset, clusterProbs, newMeans);
        try {
            List<MultivariateGaussianDistribution> components = buildComponents(newMeans, newCovs);
            GmmModel newModel = new GmmModel(clusterProbs, components);
            countOfIterations += 1;
            isConverged = isConverged(model, newModel) || countOfIterations > maxCountOfIterations;
            model = newModel;
            maxProbInDataset = GmmPartitionData.updatePcxiAndComputeLikelihood(dataset, clusterProbs, components);
        } catch (SingularMatrixException | IllegalArgumentException e) {
            String msg = "Cannot construct non-singular covariance matrix by data. " + "Try to select other initial means or other model trainer. Iterations will stop.";
            environment.logger().log(MLLogger.VerboseLevel.HIGH, msg);
            isConverged = true;
        }
    }
    return new UpdateResult(model, maxProbInDataset);
}

Example 3 with SingularMatrixException

use of org.apache.ignite.ml.math.exceptions.math.SingularMatrixException in project ignite by apache.

the class LUDecomposition method solve.

/**
 * @param b Matrix to solve using this decomposition.
 * @return Solution matrix.
 */
public Matrix solve(Matrix b) {
    final int m = pivot.size();
    if (b.rowSize() != m)
        throw new CardinalityException(b.rowSize(), m);
    if (singular)
        throw new SingularMatrixException();
    final int nColB = b.columnSize();
    // Apply permutations to b
    final double[][] bp = new double[m][nColB];
    for (int row = 0; row < m; row++) {
        final double[] bpRow = bp[row];
        final int pRow = (int) pivot.get(row);
        for (int col = 0; col < nColB; col++) bpRow[col] = b.get(pRow, col);
    }
    // Solve LY = b
    for (int col = 0; col < m; col++) {
        final double[] bpCol = bp[col];
        for (int i = col + 1; i < m; i++) {
            final double[] bpI = bp[i];
            final double luICol = lu.get(i, col);
            for (int j = 0; j < nColB; j++) bpI[j] -= bpCol[j] * luICol;
        }
    }
    // Solve UX = Y
    for (int col = m - 1; col >= 0; col--) {
        final double[] bpCol = bp[col];
        final double luDiag = lu.getX(col, col);
        for (int j = 0; j < nColB; j++) bpCol[j] /= luDiag;
        for (int i = 0; i < col; i++) {
            final double[] bpI = bp[i];
            final double luICol = lu.get(i, col);
            for (int j = 0; j < nColB; j++) bpI[j] -= bpCol[j] * luICol;
        }
    }
    return b.like(b.rowSize(), b.columnSize()).assign(bp);
}