Routine Name: parallelMatrixMatrixProduct

Author: Christopher Johnson

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

Declared in “parallelOperations.h”

Description/Purpose: Performs Matrix-Matrix multiplication. Parallel implementation.

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

Output: Returns a Matrix, typedef of std::vector<std::vector<double>>. The size is equal rows(a) x cols(b).

Usage/Example:

Two Matrices must first be defined, and then the method may be called.

Matrix c = { {1,2,3},{4,5,6} };
Matrix d = { {7,8},{9,10},{11,12} };
std::cout << "C * D is\n"
printVector(parallelMatrixMatrixProduct(c,d));

Output from the lines above:

C * D is
[[58,64],
[139,154]]

Implementation/Code: The following is the code for parallelMatrixMatrixProduct(a,b)

Matrix parallelMatrixMatrixProduct(const Matrix &a, const Matrix &b)
{
    int n = a.size();
    int m = b.size();
    int r = b[0].size();
    int i,j,k;

    Matrix sol(n, Vector(r,0));

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

Last Modified: October/2017