net.maizegenetics.stats.PCA

Class PrinComp

net.maizegenetics.stats.PCA.PrinComp

public class PrinComp

Constructor Detail

PrinComp

public PrinComp(DoubleMatrix data,
                net.maizegenetics.stats.PCA.PrinComp.PC_TYPE type)

The class uses singular value decomposition to find the eigenvalues, eigenvectors and principal components of either the covariance or correlation matrix of the data. If covariance, then the result is the equivalent of finding the eighvalue decomposition of XX'/(n-1) where X is the data matrix with the column means subtracted from the columns. If correlation, the column values are also scaled. That is, after the mean is subtracted, the values are divided by the standard deviation.

Parameters:

data - a matrix of data

type - should the analysis use the covariance (cov) or the correlation (corr) matrix of the data

Method Detail

getEigenValues

public kotlin.Array[] getEigenValues()

Returns:

org.jetbrains.dokka.NodeRenderContent@f70aa56: a double[] of eigenvalues from the decomposition of either the covariance or correlation matrix of the data

getEigenValuesAsColumnVector

public DoubleMatrix getEigenValuesAsColumnVector()

Returns:

org.jetbrains.dokka.NodeRenderContent@373838e6: a column vector of eigenvalues from the decomposition of either the covariance or correlation matrix of the data

getEigenvalueMatrix

public DoubleMatrix getEigenvalueMatrix()

Returns:

org.jetbrains.dokka.NodeRenderContent@c66b7dd: a square matrix with diagonal equal to the eigenvalues of the covariance or correlation matrix of the data

getEigenVectors

public DoubleMatrix getEigenVectors()

Returns:

org.jetbrains.dokka.NodeRenderContent@1d847c42: the eigenvectors from the decomposition of either the covariance or correlation matrix of the data

getPrincipalComponents

public DoubleMatrix getPrincipalComponents()

calculated as data * eigenvectors

Returns:

org.jetbrains.dokka.NodeRenderContent@172c47be: all of the principal components