David Joyner wrote:
The paper http://www.lehigh.edu/~tkr2/research/papers/MILP04.pdf provides a survey of "non-commercial" LP solvers. It appears that COIN-OR's Symphany package solves a wider class of problems by that (in one comparison at least) lp_solve is relatively fast. Do you agree? Can you offer other comparisons with Symphany?Sage currently ships only with cvxopt but, AFAIK, cvxopt does not handle MILPs.
See also http://wiki.sagemath.org/optimization for plans to include solvers in Sage. I've just read a benchmark paper comparing lp_solve, glpk and Coin-Or's clp, the last always winning by 1 or 2 orders of magnitude in computation time, but it's from 2006, so maybe a bit dated.
-- Robert Schwarz <m...@rschwarz.net> Get my public key at http://rschwarz.net/key.asc
signature.asc
Description: OpenPGP digital signature
