cvxopt.coneqp runtime on large scale problems (is there any limitation?)

[Mason]

For a QP model, is there any limitation on the number of variables (n) and constraints(m) cvxopt is able to handle efficiently? 

In my case, for a QP model with inequality constraint, m = 57046 and n = 57041 it took around 500 seconds to solve it (Gurobi is able to solve in 7 seconds). I appreciate if anyone can comment on runtime? 

Thanks,

[Joachim Dahl]

There is no hard limit, but certainly commercial solvers are much efficient for sparse problems.  One of the purposes of CVXOPT was to be able to write specialized KKT solvers for various conic problems, without having to rewrite the entire surrounding functionality.  Writing highly tuned general-purposes solvers large-scale sparse problems was not an ambition.

--
You received this message because you are subscribed to the Google Groups "CVXOPT" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cvxopt+unsubscribe@googlegroups.com.
To post to this group, send email to cvx...@googlegroups.com.
Visit this group at https://groups.google.com/group/cvxopt.
For more options, visit https://groups.google.com/d/optout.