Package org.flag4j.operations.dense_sparse.csr.real

Class RealCsrDenseMatrixMultiplication

java.lang.Object

org.flag4j.operations.dense_sparse.csr.real.RealCsrDenseMatrixMultiplication

public class RealCsrDenseMatrixMultiplication
extends Object

This class contains low-level implementations of real sparse-dense matrix multiplication where the sparse matrix is in CSR format.

Constructor Summary

Constructors

Modifier	Constructor	Description
private	RealCsrDenseMatrixMultiplication()

Method Summary

Modifier and Type	Method	Description
static Matrix	standard(Matrix src1, CsrMatrix src2)	Computes the matrix multiplication between a real dense matrix and a real sparse CSR matrix.
static Matrix	standard(CsrMatrix src1, Matrix src2)	Computes the matrix multiplication between a real sparse CSR matrix and a real dense matrix.
static Matrix	standardTranspose(CsrMatrix src1, Matrix src2)	Computes the matrix multiplication between a real sparse CSR matrix and the transpose of a real dense matrix.
static Vector	standardVector(CsrMatrix src1, Vector src2)	Computes the matrix-vector multiplication between a real sparse CSR matrix and a real dense vector.
static Vector	standardVector(CsrMatrix src1, CooVector src2)	Computes the matrix-vector multiplication between a real sparse CSR matrix and a real sparse COO vector.

Methods inherited from class java.lang.Object

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

Constructor Details

RealCsrDenseMatrixMultiplication

private RealCsrDenseMatrixMultiplication()

Method Details

standard

public static Matrix standard(CsrMatrix src1,
 Matrix src2)

Computes the matrix multiplication between a real sparse CSR matrix and a real dense matrix. WARNING: If the first matrix is very large but not very sparse, this method may be slower than converting the first matrix to a dense matrix and calling Matrix.mult(Matrix).

Parameters:

src1 - First matrix in the matrix multiplication.

src2 - Second matrix in the matrix multiplication.

Returns:

The result of the matrix multiplication between src1 and src2.

Throws:

IllegalArgumentException - If src1 does not have the same number of columns as src2 has rows.

standard

public static Matrix standard(Matrix src1,
 CsrMatrix src2)

Computes the matrix multiplication between a real dense matrix and a real sparse CSR matrix. WARNING: If the second matrix is very large but not very sparse, this method may be slower than converting the second matrix to a dense matrix and calling Matrix.mult(Matrix).

Parameters:

src1 - First matrix in the matrix multiplication (dense matrix).

src2 - Second matrix in the matrix multiplication (sparse CSR matrix).

Returns:

The result of the matrix multiplication between src1 and src2.

Throws:

IllegalArgumentException - If src1 does not have the same number of columns as src2 has rows.

standardTranspose

public static Matrix standardTranspose(CsrMatrix src1,
 Matrix src2)

Computes the matrix multiplication between a real sparse CSR matrix and the transpose of a real dense matrix. WARNING: This method is likely slower than standard(src1, src2.T()) unless src1 has many more columns than rows and is very sparse.

Parameters:

src1 - First matrix in the matrix multiplication.

src2 - Second matrix in the matrix multiplication. Will be implicitly transposed.

Returns:

The result of the matrix multiplication between src1 and src2.

Throws:

IllegalArgumentException - If src1 and src2 do not have the same number of rows.

standardVector

public static Vector standardVector(CsrMatrix src1,
 Vector src2)

Computes the matrix-vector multiplication between a real sparse CSR matrix and a real dense vector.

Parameters:

src1 - The matrix in the multiplication.

src2 - Vector in multiplication. Treated as a column vector.

Returns:

The result of the matrix-vector multiplication.

Throws:

IllegalArgumentException - If the number of columns in src1 does not equal the length of src2.

standardVector

public static Vector standardVector(CsrMatrix src1,
 CooVector src2)

Computes the matrix-vector multiplication between a real sparse CSR matrix and a real sparse COO vector.

Parameters:

src1 - The matrix in the multiplication.

src2 - Vector in multiplication. Treated as a column vector in COO format.

Returns:

The result of the matrix-vector multiplication.

Throws:

IllegalArgumentException - If the number of columns in src1 does not equal the number of columns in src2.