Model Infeasible troubleshooting (m.computeIIS() gets stuck)

回答済み

Dongjun Cai

2022年11月09日 03:57

• 
• 
• 

Hi,

Recently, i set up a model in Gurobi. It shows that the model is infeasible.

In order to troubleshoot the error. I hope to use the following codes to generate the ilp file :

m.computeIIS()
m.write("model.ilp")

However, It got stuck for long time with no response (more than 10 hours). May i know what's happening? Or Could anybody suggest if there are other methods to fix an infeasible model.

Many thanks!  

0

• 正式なコメント

  

  Simranjit Kaur

  • Gurobi Staff

  2025年11月16日 23:25

  This post is more than three years old. Some information may not be up to date. For current information, please check the Gurobi Documentation or Knowledge Base. If you need more help, please create a new post in the community forum, or try Gurobot, our chatbot interface offering instant, expert-level support.
• 

  Jaromił Najman

  • Gurobi Staff

  2022年11月09日 09:07 編集済み

  Hi,

  Computing an IIS can be very expensive. It is in general not easier than solving the original model.

  You could try removing chunks of your constraints by hand in order to reduce the size of the model while keeping it infeasible (you might also find the constraints responsible for infeasibility by doing so). This might reduce the time to compute an IIS.

  Instead of computing an IIS, you could construct a feasibility relaxation, cf. How do I determine why my model is infeasible? and documentation of feasRelax. A feasibility relaxation is a relaxation of your infeasible model that it guaranteed to be feasible. You can solve the feasibility relaxation and check which auxiliary variables are non zero, i.e., which changes would have to be made in order to make your model feasible.

  Best regards, 
  Jaromił

  0

投稿コメントは受け付けていません。