sci.op-research

humberto.bortolossi@gmail.com wrote:
> I've found a symbolic package for the REDUCE system: SYMOPT. There is
> technical report about it:
>
> http://www.fmi.uni-passau.de/forschung/mip-berichte/MIP-0203.ps
>

Humberto,

This is interesting, but I suspect it is useful only for relatively
small problems. The number of potential vertices (and typically the
number of actual vertices) grows exponentially with the number of
variables (including slacks). Then you have to deal with combinations
of inequalities to decide which of those points exist, and which are
feasible, and then which of them might be optimal. Thus when Frank
Kampas ran your example, which has 4C2 = 6 possible vertices, he
encountered upwards of 10 distinct cases.

If you are only interested in small problems, though, it seems
Mathematica is an option.

Cheers,
Paul