Routine Name: parallerMatrixVectorProduct

Author: Christopher Johnson

Language: C++11. Tested with g++ compiler.

Declared in “parallelOperations.h”

Description/Purpose: Multiplies Matrix by Vector. Parallel implementation.

Input: Matrix a and Vector v, where Matrix is a typedef of std::vector<std::vector<double>> and Vector is a typedef of std::vector<double>.

Output: Returns a Vector, typedef of std::vector<double>. The size is equal to rows(a).

Usage/Example:

A Matrix must first be defined, and then the method may be called.

Matrix b(5, Vector(5,1)); \\5*5 matrix of all ones
b = matrixAdd(b,identityMatrix(5));
Vector v{1,2,3,4,5};
std::cout << "B * v is\n"
printVector(parallelMatrixVectorProduct(b,v));

Output from the lines above:

B * v is
[16, 17, 18, 19, 20]

Implementation/Code: The following is the code for parallelMatrixVectorProduct(a,v)

Vector parallelMatrixVectorProduct(const Matrix &a, const Vector &v)
{
	int m = a.size();
	int n = a[0].size();
	int i,j;
	Vector sol(m,0);

	#pragma omp parallel for private(j)
	for (i=0;i<m;++i)
		for (j=0;j<n;++j)
			sol[i]+=a[i][j]*v[j];
	return sol;
}

Last Modified: October/2017