On Nov 4, 7:27 pm, "analys...@hotmail.com" <analys...@hotmail.com> wrote: > > Yes, please post your solution. > > Right now I am thinking of following the suggestion of implementing > primal simplex with bounded variables with only two rows.
Sorry for the delay; the INFORMS conference has kept me hopping. My algorithm is a bit long to explain here, so I posted it on my blog (http://orinanobworld.blogspot.com/2010/11/continuous-knapsack-with- count.html). I have proof-of-concept code (in Java), although I didn't bother to code the bisection search for the first phase (I just brute-force it). If you want to see the code, you can reply directly to me (or to the list, or drop a comment on my blog, your choice) and send me your e-mail address.