Skip to content

Matrix Decompositions

Thomas Shields edited this page Aug 24, 2016 · 1 revision

Available decompositions:

Name Description
CholeskyDecompositor This class represents Cholesky decomposition of matrices.
EigenDecompositor This class represents Eigen decomposition of matrices.
LUDecompositor This class represents LU decomposition of matrices.
QRDecompositor This class represents QR decomposition of matrices.
RawLUDecompositor
RawQRDecompositor
SingularValueDecompositor This class represents singular value decomposition of matrices.

An example of how to decompose a matrix with SVD is below - the other decompositors can be used the same way.

SingularValueDecompositor decomp = new SingularValueDecompositor(matrix);

Matrix[] result = decomp.decompose();

Matrix U = result[0];
Matrix D = result[1];
Matrix V = result[2];
Clone this wiki locally