Hello,

I'm interested in finding out the sparsity fraction of a MatrixXd without having to write helper function loops etc i.e. count the zeros and then get:
sparsity fraction = zeroCount/M*N

Is there some provision in Eigen for this use-case somewhere in whatever form? I need it to automatically switch between dense and sparse solvers.

TIA,
Best regards,
Giovanni

That's easy: double(mat.nonZeros())/double(mat.size())

hm wait, the above works for a sparse matrix, so for a MatrixXd: double((mat.array()!=0).count()) / double(mat.size())