Parallel Assembly Of A Sparse Matrix In Python

I'm trying to use mpi4py to assemble a very large sparse matrix in parallel. Each rank produces a sparse sub matrix (in scipy's dok format) that needs to be put in place in the ver

Solution 1:

For what is worth, using three element lists work pretty well as suggested by hpaulj. Here's a working example:

from mpi4py import MPI
from numpy import *
import scipy.sparse as ss
from timeit import default_timer as timer

comm = MPI.COMM_WORLD
rank = comm.Get_rank()

if rank == 0:
    tic = timer()      

# each rank generates a sparse matrix with N entries on the diagonal
N = 100000
block = ss.eye(N, format = 'coo')

# extract indices and values
out = [ block.data, block.row , block.col]
out[1] = out[1] + rank * N
out[2] = out[2] + rank * N

# gather all the arrays representing the submatrices
full_list = comm.gather(out,root=0)

if rank == 0:
    dat = concatenate([x[0] for x in full_list])
    row = concatenate([x[1] for x in full_list])
    col = concatenate([x[2] for x in full_list])
    final_result = ss.csr_matrix( ( dat, (row, col) ) )
    toc = timer()
    print 'Matrix assembled in', toc-tic, 'seconds'

The assembly is definitely much faster using coo matrices rather than dok.