R help - Sep 2010 - non-linear integer optimization?

Are there any packages that do non-linear integer otimization?  Looked at
lpSolve but i'm pretty sure it only works with linear programming and not
non-linear, tried "L-BFGS-B" optim after flooring all my params, works
somewhat but seems really inefficient.  Anything else I should look at?
-- 
View this message in context:
http://r.789695.n4.nabble.com/non-linear-integer-optimization-tp2551409p2551409.html
Sent from the R help mailing list archive at Nabble.com.

darckeen <darckeen <at> hotmail.com> writes:
> 
> Are there any packages that do non-linear integer otimization?  Looked at
> lpSolve but i'm pretty sure it only works with linear programming and
not
> non-linear, tried "L-BFGS-B" optim after flooring all my params,
works
> somewhat but seems really inefficient.  Anything else I should look at?
The most general answer is:  There is no R package for MINLP problems !

But: Would have been nice if you had done the following things:

  - Have a look into the 'optimization' task view
  - Provide an example of a function you would like to optimize
    (can it be reduced to a quadratic or convex problem?).
  - Tell whether you need to do it repeatedly or only a few times
    (e.g., utilizing an interface to the NEOS solvers).
  - How many integer variables, binary or really integer, etc.
    (could you do your own branch-and-bound, e.g.)?

Providing as little information as above makes communicating difficult.

Hans Werner

P. S.:  The R-Forge project "Integer and Nonlinear Optimization in R
(RINO)"
        will (hopefully) make some of those COIN-OR solvers available.