1

I have a code with lpsolve package which quite standard way calculates minimal price for given constrains. That works ok.

My next task is to see if i exclude or change one constraint, how does prices change? It could be ok to have little bit higher price, but one constraint is excluded for example. Is there any way to calculate this sensitivity towards constraints? I understand that brutal way is to solve problem again with one constraint excluded, but it's too computationally expensive. Is there any algorithm or so?

I have seen something that one can calculate sensitivity towards coefficients in constraint. But maybe there is simple elegant way for this. Thanks!

Anna K
  • 81
  • 1
  • 2
  • 5
  • Commercial solvers have support for this kind of analysis (although i never used it). I don't know if lpsolve does support this (some [mentioning in the docs](http://web.mit.edu/lpsolve/doc/)), maybe you could also use [glpk](https://en.wikibooks.org/wiki/GLPK/Solution_information#Sensitivity_analysis_report) (also open-source; but we don't know why you decided lpsolve). But one remark: if lpsolve supports model-modification with efficient warm-starting (simplex!) don't be too scared about the computational factor. – sascha Nov 16 '17 at 03:40

0 Answers0