Package de.labathome.trustconstr.sparse

Class SparseLinearOperator

java.lang.Object

de.labathome.trustconstr.sparse.SparseLinearOperator

public final class SparseLinearOperator
extends Object

Adapter from a CSRMatrix or CSCMatrix to the existing LinearOperator interface used throughout the algorithm. Mirrors the role of scipy.sparse.linalg.LinearOperator produced from a sparse matrix via aslinearoperator(A).

The LinearOperator contract is typed in terms of Matrix, so this adapter does the double[] ↔ Matrix conversion at the boundary on each apply.

Method Summary

Modifier and Type	Method	Description
static LinearOperator	forMatvec(CSCMatrix a)
static LinearOperator	forMatvec(CSRMatrix a)
static LinearOperator	forRmatvec(CSCMatrix a)
static LinearOperator	forRmatvec(CSRMatrix a)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Method Details

forMatvec

public static LinearOperator forMatvec(CSRMatrix a)

Parameters:

a - CSR matrix

Returns:

a LinearOperator that computes y = A x

forMatvec

public static LinearOperator forMatvec(CSCMatrix a)

Parameters:

a - CSC matrix

Returns:

a LinearOperator that computes y = A x

forRmatvec

public static LinearOperator forRmatvec(CSRMatrix a)

Parameters:

a - CSR matrix

Returns:

a LinearOperator that computes y = A^T x

forRmatvec

public static LinearOperator forRmatvec(CSCMatrix a)

Parameters:

a - CSC matrix

Returns:

a LinearOperator that computes y = A^T x