spparms('umfpack', 0);
% spparms('umfpack', 1);
% setup example matrix:
J = ones(3,4);
J = blkdiag(J,J);
J = [J; 2*ones(1,8)];
J = [J 3*ones(7,1)];
J = blkdiag(J,J);
J = [J; zeros(4,18); 4*ones(1,18)];
J = [J 5*ones(19,1)];
J = sparse(J);
[rows cols vals] = find(J);
subplot(2,2,1);
spy(J)
rblocks = [1 3; 4 6; 8 10; 11 13; 15 18];
cblocks = [1 4; 5 8; 10 13; 14 17];
rlist = [7 14 19];
clist = [9 18 19];
N = length(rlist);
new_nrows = N*size(rblocks,1)-1;
for i = 1:size(rblocks,1)-1;
end
idx = zeros(size(rows));
for r=rlist
idx = idx | rows==r;
end
for c=clist
idx = idx | cols==c;
end
% load('jac_po', 'J', 'b', 'x');
% tic; J\b; toc
% tic; [L,U,P,Q,R] = lu(J); toc
% tic; [L,U,P,Q] = lu(J); toc