0

I would like to maximize $P = 2pr + 3pq + 2rq$ given that $p+q+3r=1 \wedge p,q,r \geq 0 \wedge p,q,r \leq 1.$ I was considering to develop $P$ as a function of three variables and using standard Lagrange Optimization, but I was wondering if it would be better to reduce $P$ to a function of two variables and then solve for these said variables. I sometimes find working in convex optimization to be slightly tedious.

Erica
  • 1
  • It's tedious no matter which way you go. When I substitute for $p$, I get the only critical point is outside the region, so your maximum is on the boundardy. And testing the $p$ boundaries in particular are significant work. I haven't tried both ways, but I suspect Lagrange would actually be less tedious. – Paul Sinclair Dec 09 '15 at 03:27
  • This isn't convex, actually. – Michael Grant Dec 09 '15 at 14:31

0 Answers0