-1

I need help in a price model optimization. I am trying to maximize Sale based on several conditions.I have already done optimization in Excel using solver(GRG Nonlinear) but want to do in R since solver has limitations(Microsoft Excel Solver has a limit of 200 decision variables, for both linear and nonlinear problems).

user144700
  • 60
  • 1
  • 11
  • What have you tried? As it is this question is very broad and not about a specific programming problem you have. – John Paul Dec 15 '15 at 20:51
  • There's no question here. Certainly there is not a programming question. There should be a question mark, and a list of stuff you've already tried, and what you expected to see, and what you saw... – BadZen Dec 15 '15 at 20:51
  • @BadZen I am new to r programming and want to know how can I perform optimization for n number of sku. What are the packages which can be helpful for non linear optimization and also a simple example of optimization might help me. – user144700 Dec 15 '15 at 20:58
  • @user144700 - Please read http://stackoverflow.com/help/how-to-ask – BadZen Dec 15 '15 at 21:00
  • 1
    In particular, this is not your first stop for that sort of question. Research first, then try, then SO if there is a problem/misunderstanding... no one here is going to do your basic open-ended research, that's not what this site is for. =) – BadZen Dec 15 '15 at 21:02
  • @BadZen All right sir. – user144700 Dec 15 '15 at 21:07
  • This is not a well written SO question but for posterity in addition to the CRAN Task View on optimization you should check out John Nash's book on optimization in R (son of the famous John Nash). – Hack-R Apr 05 '18 at 15:50

1 Answers1

0

Excel's NLP solver is based on Lasdon's GRG2 solver. I don't think this is available under R. We don't know the exact form of your model and its size (details like whether the constraints are linear or not, whether the objective is linear, quadratic or otherwise nonlinear etc), so it is difficult to recommend a particular solver. Here is a list of solvers available under R. Opposed to good LP solvers that basically can solve whatever you throw at them, NLP solvers are a little bit more fragile and may require a little bit more hand-holding (things like scaling, initial point, bounds come to mind).

Erwin Kalvelagen
  • 15,677
  • 2
  • 14
  • 39