lixueclaire / parallel-svd

parallel SVD algorithm (Jacobis Rotations) using OpenMP

parallel SVD Using Jacobis Rotations

Parallel SVD using jacobis rotations, implemented in OpenMP.

generate M*N matrix

	python randomMatrix.py M N

M = # of columns
N = # of Rows
Matrix must be squared (M=N)

serial algorithm

	g++ SVD.cpp -o svd
	./svd M N -t -d

-t = print out Timing and # of Iterations
-p = print out Results (U, S, V)
-d = Generate the Octave files for debug and verify correctness

parallel algorithm

	mpic++ -fopenmp OMP_SVD.cpp -o omp_svd
	mpiexec -n 2 ./omp_svd M N -t -d

validate if the result is right

	g++ Validation.cpp -o validation
	./validation

parallel-svd's Issues

General matrix

Hi. This implements svd for a square matrix, how to compute a similar thing for M!=N?