Search in sources :

Example 6 with Array2DRowRealMatrix

use of org.apache.commons.math3.linear.Array2DRowRealMatrix in project gatk-protected by broadinstitute.

the class SparkConverter method convertSparkMatrixToRealMatrix.

/**
     * Convert a local (not distributed) Spark Matrix to an Apache Commons matrix.
     *
     * @param r Never {@code null}
     * @return Not {@code null}
     */
public static RealMatrix convertSparkMatrixToRealMatrix(final Matrix r) {
    final RealMatrix result = new Array2DRowRealMatrix(r.numRows(), r.numCols());
    final double[] columnMajorMat = r.toArray();
    for (int i = 0; i < r.numRows(); i++) {
        result.setRow(i, Arrays.copyOfRange(columnMajorMat, i * r.numCols(), i * r.numCols() + r.numCols()));
    }
    return result;
}
Also used : Array2DRowRealMatrix(org.apache.commons.math3.linear.Array2DRowRealMatrix) RealMatrix(org.apache.commons.math3.linear.RealMatrix) Array2DRowRealMatrix(org.apache.commons.math3.linear.Array2DRowRealMatrix)

Example 7 with Array2DRowRealMatrix

use of org.apache.commons.math3.linear.Array2DRowRealMatrix in project gatk-protected by broadinstitute.

the class HDF5PCACoveragePoNCreationUtilsUnitTest method simpleEigensampleData.

@DataProvider(name = "singleEigensample")
public Object[][] simpleEigensampleData() {
    final List<Object[]> result = new ArrayList<>();
    final int NUM_TARGETS = 10;
    final int NUM_SAMPLES = 5;
    final List<Target> targets = IntStream.range(0, NUM_TARGETS).boxed().map(i -> new Target("target_" + i, new SimpleInterval("1", 100 * i + 1, 100 * i + 5))).collect(Collectors.toList());
    final List<String> columnNames = IntStream.range(0, NUM_SAMPLES).boxed().map(i -> "sample_" + i).collect(Collectors.toList());
    double[][] countsArray = new double[NUM_TARGETS][NUM_SAMPLES];
    final RealMatrix counts = new Array2DRowRealMatrix(countsArray);
    // All row data is the same (0,1,2,3,4...)
    final double[] rowData = IntStream.range(0, NUM_SAMPLES).boxed().mapToDouble(i -> i).toArray();
    for (int i = 0; i < NUM_TARGETS; i++) {
        counts.setRow(i, rowData);
    }
    new ReadCountCollection(targets, columnNames, counts);
    result.add(new Object[] { new ReadCountCollection(targets, columnNames, counts) });
    return result.toArray(new Object[result.size()][]);
}
Also used : IntStream(java.util.stream.IntStream) SVD(org.broadinstitute.hellbender.utils.svd.SVD) DataProvider(org.testng.annotations.DataProvider) BaseTest(org.broadinstitute.hellbender.utils.test.BaseTest) JavaSparkContext(org.apache.spark.api.java.JavaSparkContext) Level(org.apache.logging.log4j.Level) MatrixSummaryUtils(org.broadinstitute.hellbender.utils.MatrixSummaryUtils) Test(org.testng.annotations.Test) Random(java.util.Random) OptionalInt(java.util.OptionalInt) ParamUtils(org.broadinstitute.hellbender.utils.param.ParamUtils) ArrayList(java.util.ArrayList) Mean(org.apache.commons.math3.stat.descriptive.moment.Mean) Pair(org.apache.commons.lang3.tuple.Pair) Message(org.apache.logging.log4j.message.Message) Assert(org.testng.Assert) Median(org.apache.commons.math3.stat.descriptive.rank.Median) HDF5File(org.broadinstitute.hdf5.HDF5File) Marker(org.apache.logging.log4j.Marker) AbstractLogger(org.apache.logging.log4j.spi.AbstractLogger) Array2DRowRealMatrix(org.apache.commons.math3.linear.Array2DRowRealMatrix) IOUtils(org.broadinstitute.hellbender.utils.io.IOUtils) SimpleInterval(org.broadinstitute.hellbender.utils.SimpleInterval) Collectors(java.util.stream.Collectors) File(java.io.File) DoubleStream(java.util.stream.DoubleStream) List(java.util.List) Percentile(org.apache.commons.math3.stat.descriptive.rank.Percentile) Logger(org.apache.logging.log4j.Logger) ReadCountCollection(org.broadinstitute.hellbender.tools.exome.ReadCountCollection) Stream(java.util.stream.Stream) Target(org.broadinstitute.hellbender.tools.exome.Target) SVDFactory(org.broadinstitute.hellbender.utils.svd.SVDFactory) RealMatrix(org.apache.commons.math3.linear.RealMatrix) SparkContextFactory(org.broadinstitute.hellbender.engine.spark.SparkContextFactory) PoNTestUtils(org.broadinstitute.hellbender.tools.pon.PoNTestUtils) ReadCountCollection(org.broadinstitute.hellbender.tools.exome.ReadCountCollection) ArrayList(java.util.ArrayList) Target(org.broadinstitute.hellbender.tools.exome.Target) Array2DRowRealMatrix(org.apache.commons.math3.linear.Array2DRowRealMatrix) RealMatrix(org.apache.commons.math3.linear.RealMatrix) Array2DRowRealMatrix(org.apache.commons.math3.linear.Array2DRowRealMatrix) SimpleInterval(org.broadinstitute.hellbender.utils.SimpleInterval) DataProvider(org.testng.annotations.DataProvider)

Example 8 with Array2DRowRealMatrix

use of org.apache.commons.math3.linear.Array2DRowRealMatrix in project gatk by broadinstitute.

the class HDF5PCACoveragePoNCreationUtilsUnitTest method readCountAndPercentileData.

// this is duplicated from ReadCountCollectionUtilsUnitTest
@DataProvider(name = "readCountAndPercentileData")
public Object[][] readCountAndPercentileData() {
    final double[] percentiles = new double[] { 1.0, 2.5, 5.0, 10.0, 25.0 };
    final List<Object[]> result = new ArrayList<>();
    final Random rdn = new Random(13);
    final int columnCount = 100;
    final int targetCount = 100;
    final List<String> columnNames = IntStream.range(0, columnCount).mapToObj(i -> "sample_" + (i + 1)).collect(Collectors.toList());
    final List<Target> targets = IntStream.range(0, targetCount).mapToObj(i -> new Target("target_" + (i + 1))).collect(Collectors.toList());
    for (final double percentile : percentiles) {
        final double[][] counts = new double[columnCount][targetCount];
        for (int i = 0; i < counts.length; i++) {
            for (int j = 0; j < counts[0].length; j++) {
                counts[i][j] = rdn.nextDouble();
            }
        }
        final ReadCountCollection readCounts = new ReadCountCollection(targets, columnNames, new Array2DRowRealMatrix(counts, false));
        result.add(new Object[] { readCounts, percentile });
    }
    return result.toArray(new Object[result.size()][]);
}
Also used : IntStream(java.util.stream.IntStream) SVD(org.broadinstitute.hellbender.utils.svd.SVD) DataProvider(org.testng.annotations.DataProvider) BaseTest(org.broadinstitute.hellbender.utils.test.BaseTest) JavaSparkContext(org.apache.spark.api.java.JavaSparkContext) Level(org.apache.logging.log4j.Level) MatrixSummaryUtils(org.broadinstitute.hellbender.utils.MatrixSummaryUtils) Test(org.testng.annotations.Test) Random(java.util.Random) OptionalInt(java.util.OptionalInt) ParamUtils(org.broadinstitute.hellbender.utils.param.ParamUtils) ArrayList(java.util.ArrayList) Mean(org.apache.commons.math3.stat.descriptive.moment.Mean) Pair(org.apache.commons.lang3.tuple.Pair) Message(org.apache.logging.log4j.message.Message) Assert(org.testng.Assert) Median(org.apache.commons.math3.stat.descriptive.rank.Median) HDF5File(org.broadinstitute.hdf5.HDF5File) Marker(org.apache.logging.log4j.Marker) AbstractLogger(org.apache.logging.log4j.spi.AbstractLogger) Array2DRowRealMatrix(org.apache.commons.math3.linear.Array2DRowRealMatrix) IOUtils(org.broadinstitute.hellbender.utils.io.IOUtils) SimpleInterval(org.broadinstitute.hellbender.utils.SimpleInterval) Collectors(java.util.stream.Collectors) File(java.io.File) DoubleStream(java.util.stream.DoubleStream) List(java.util.List) Percentile(org.apache.commons.math3.stat.descriptive.rank.Percentile) Logger(org.apache.logging.log4j.Logger) ReadCountCollection(org.broadinstitute.hellbender.tools.exome.ReadCountCollection) Stream(java.util.stream.Stream) Target(org.broadinstitute.hellbender.tools.exome.Target) SVDFactory(org.broadinstitute.hellbender.utils.svd.SVDFactory) RealMatrix(org.apache.commons.math3.linear.RealMatrix) SparkContextFactory(org.broadinstitute.hellbender.engine.spark.SparkContextFactory) PoNTestUtils(org.broadinstitute.hellbender.tools.pon.PoNTestUtils) ReadCountCollection(org.broadinstitute.hellbender.tools.exome.ReadCountCollection) ArrayList(java.util.ArrayList) Target(org.broadinstitute.hellbender.tools.exome.Target) Random(java.util.Random) Array2DRowRealMatrix(org.apache.commons.math3.linear.Array2DRowRealMatrix) DataProvider(org.testng.annotations.DataProvider)

Example 9 with Array2DRowRealMatrix

use of org.apache.commons.math3.linear.Array2DRowRealMatrix in project gatk by broadinstitute.

the class PoNTestUtils method readTsvIntoMatrix.

/**
     * Reads a very basic tsv (numbers separated by tabs) into a RealMatrix.
     * <p>Very little error checking happens in this method</p>
     *
     * @param inputFile readable file.  Not {@code null}
     * @return never {@code null}
     */
public static RealMatrix readTsvIntoMatrix(final File inputFile) {
    IOUtils.canReadFile(inputFile);
    final List<double[]> allData = new ArrayList<>();
    int ctr = 0;
    try {
        final CSVReader reader = new CSVReader(new FileReader(inputFile), '\t', CSVWriter.NO_QUOTE_CHARACTER);
        String[] nextLine;
        while ((nextLine = reader.readNext()) != null) {
            ctr++;
            allData.add(Arrays.stream(nextLine).filter(s -> StringUtils.trim(s).length() > 0).map(s -> Double.parseDouble(StringUtils.trim(s))).mapToDouble(d -> d).toArray());
        }
    } catch (final IOException ioe) {
        Assert.fail("Could not open test file: " + inputFile, ioe);
    }
    final RealMatrix result = new Array2DRowRealMatrix(allData.size(), allData.get(0).length);
    for (int i = 0; i < result.getRowDimension(); i++) {
        result.setRow(i, allData.get(i));
    }
    return result;
}
Also used : Array2DRowRealMatrix(org.apache.commons.math3.linear.Array2DRowRealMatrix) Arrays(java.util.Arrays) IOUtils(org.broadinstitute.hellbender.utils.io.IOUtils) CSVWriter(au.com.bytecode.opencsv.CSVWriter) CSVReader(com.opencsv.CSVReader) HDF5PCACoveragePoNCreationUtils(org.broadinstitute.hellbender.tools.pon.coverage.pca.HDF5PCACoveragePoNCreationUtils) IOException(java.io.IOException) TargetArgumentCollection(org.broadinstitute.hellbender.tools.exome.TargetArgumentCollection) StringUtils(org.apache.commons.lang3.StringUtils) CreatePanelOfNormals(org.broadinstitute.hellbender.tools.exome.CreatePanelOfNormals) OptionalInt(java.util.OptionalInt) ParamUtils(org.broadinstitute.hellbender.utils.param.ParamUtils) File(java.io.File) ArrayList(java.util.ArrayList) HDF5PCACoveragePoN(org.broadinstitute.hellbender.tools.pon.coverage.pca.HDF5PCACoveragePoN) List(java.util.List) PCACoveragePoN(org.broadinstitute.hellbender.tools.pon.coverage.pca.PCACoveragePoN) Assert(org.testng.Assert) Target(org.broadinstitute.hellbender.tools.exome.Target) Utils(org.broadinstitute.hellbender.utils.Utils) RealMatrix(org.apache.commons.math3.linear.RealMatrix) HDF5File(org.broadinstitute.hdf5.HDF5File) FileReader(java.io.FileReader) TargetCollection(org.broadinstitute.hellbender.tools.exome.TargetCollection) Array2DRowRealMatrix(org.apache.commons.math3.linear.Array2DRowRealMatrix) RealMatrix(org.apache.commons.math3.linear.RealMatrix) Array2DRowRealMatrix(org.apache.commons.math3.linear.Array2DRowRealMatrix) CSVReader(com.opencsv.CSVReader) ArrayList(java.util.ArrayList) FileReader(java.io.FileReader) IOException(java.io.IOException)

Example 10 with Array2DRowRealMatrix

use of org.apache.commons.math3.linear.Array2DRowRealMatrix in project gatk by broadinstitute.

the class HDF5PCACoveragePoNCreationUtilsUnitTest method readCountOnlyData.

@DataProvider(name = "readCountOnlyData")
public Object[][] readCountOnlyData() {
    final int repeats = 4;
    final List<Object[]> result = new ArrayList<>(repeats);
    final Random rdn = new Random(13);
    final int columnCount = 100;
    final int targetCount = 100;
    final List<String> columnNames = IntStream.range(0, columnCount).mapToObj(i -> "sample_" + (i + 1)).collect(Collectors.toList());
    final List<Target> targets = IntStream.range(0, targetCount).mapToObj(i -> new Target("target_" + (i + 1))).collect(Collectors.toList());
    for (int k = 0; k < repeats; k++) {
        final double[][] counts = new double[columnCount][targetCount];
        for (int i = 0; i < counts.length; i++) {
            for (int j = 0; j < counts[0].length; j++) {
                counts[i][j] = rdn.nextDouble();
            }
        }
        final ReadCountCollection readCounts = new ReadCountCollection(targets, columnNames, new Array2DRowRealMatrix(counts, false));
        result.add(new Object[] { readCounts });
    }
    return result.toArray(new Object[result.size()][]);
}
Also used : IntStream(java.util.stream.IntStream) SVD(org.broadinstitute.hellbender.utils.svd.SVD) DataProvider(org.testng.annotations.DataProvider) BaseTest(org.broadinstitute.hellbender.utils.test.BaseTest) JavaSparkContext(org.apache.spark.api.java.JavaSparkContext) Level(org.apache.logging.log4j.Level) MatrixSummaryUtils(org.broadinstitute.hellbender.utils.MatrixSummaryUtils) Test(org.testng.annotations.Test) Random(java.util.Random) OptionalInt(java.util.OptionalInt) ParamUtils(org.broadinstitute.hellbender.utils.param.ParamUtils) ArrayList(java.util.ArrayList) Mean(org.apache.commons.math3.stat.descriptive.moment.Mean) Pair(org.apache.commons.lang3.tuple.Pair) Message(org.apache.logging.log4j.message.Message) Assert(org.testng.Assert) Median(org.apache.commons.math3.stat.descriptive.rank.Median) HDF5File(org.broadinstitute.hdf5.HDF5File) Marker(org.apache.logging.log4j.Marker) AbstractLogger(org.apache.logging.log4j.spi.AbstractLogger) Array2DRowRealMatrix(org.apache.commons.math3.linear.Array2DRowRealMatrix) IOUtils(org.broadinstitute.hellbender.utils.io.IOUtils) SimpleInterval(org.broadinstitute.hellbender.utils.SimpleInterval) Collectors(java.util.stream.Collectors) File(java.io.File) DoubleStream(java.util.stream.DoubleStream) List(java.util.List) Percentile(org.apache.commons.math3.stat.descriptive.rank.Percentile) Logger(org.apache.logging.log4j.Logger) ReadCountCollection(org.broadinstitute.hellbender.tools.exome.ReadCountCollection) Stream(java.util.stream.Stream) Target(org.broadinstitute.hellbender.tools.exome.Target) SVDFactory(org.broadinstitute.hellbender.utils.svd.SVDFactory) RealMatrix(org.apache.commons.math3.linear.RealMatrix) SparkContextFactory(org.broadinstitute.hellbender.engine.spark.SparkContextFactory) PoNTestUtils(org.broadinstitute.hellbender.tools.pon.PoNTestUtils) Target(org.broadinstitute.hellbender.tools.exome.Target) Random(java.util.Random) Array2DRowRealMatrix(org.apache.commons.math3.linear.Array2DRowRealMatrix) ReadCountCollection(org.broadinstitute.hellbender.tools.exome.ReadCountCollection) ArrayList(java.util.ArrayList) DataProvider(org.testng.annotations.DataProvider)

Aggregations

Array2DRowRealMatrix (org.apache.commons.math3.linear.Array2DRowRealMatrix)141 RealMatrix (org.apache.commons.math3.linear.RealMatrix)101 Test (org.testng.annotations.Test)60 IntStream (java.util.stream.IntStream)31 BaseTest (org.broadinstitute.hellbender.utils.test.BaseTest)28 File (java.io.File)27 Collectors (java.util.stream.Collectors)25 ArrayList (java.util.ArrayList)24 Assert (org.testng.Assert)24 List (java.util.List)22 SimpleInterval (org.broadinstitute.hellbender.utils.SimpleInterval)22 Target (org.broadinstitute.hellbender.tools.exome.Target)18 java.util (java.util)15 Random (java.util.Random)14 ReadCountCollection (org.broadinstitute.hellbender.tools.exome.ReadCountCollection)14 ParamUtils (org.broadinstitute.hellbender.utils.param.ParamUtils)14 DataProvider (org.testng.annotations.DataProvider)14 Stream (java.util.stream.Stream)13 Arrays (java.util.Arrays)12 DoubleStream (java.util.stream.DoubleStream)12