Unbounded Ray

Answered

Mina Valaei

November 06, 2023 18:36

• 
• 
• 

Hello,

I have a problem and I am using Bender's Decomposition to solve it. This is my subproblem which gets unbounded:

I want to get the unboundedness ray to add it to the master problem's feasibility cut as below:

Note that I = 238, J = 19, K = 15. So I have 4522 q(ij)s, 4522 p(ij)s, and 19 r(j)s. (Totally 9063 variables)

I am using UnbdRay function as below:

InfUnbdInfo = 1

a = np.array(sp1.UnbdRay)

 

The dimension of a vector is 1 * 9063, which I think is the sum of all decision variables of my subproblem (q, p and r). I do not know how to use this function to add the unbounded ray to the feasibility cut to my master problem. Can you help me with the way of using this function?

 

Thanks,

Mina

 

 

Related to

• Unbounded
• Bender's

0

• Official comment

  

  Simranjit Kaur

  • Gurobi Staff

  November 13, 2026 01:35

  This post is more than two 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.
• 

  Mario Ruthmair

  • Gurobi Staff

  November 09, 2023 11:32

  Hi Mina,

  I guess your question is how to obtain the unbounded ray for one particular variable, right?
  You can access the UnbdRay attribute directly for each variable via

  var.UnbdRay

  In your code, it seems you want to access the UnbdRay attribute for the whole variable array (sp1?) in one step.
  How did you create your variables? If you created them via addVars() like

  r = model.addVars(J, ...)
  p = model.addVars(I, J, ...)
  q = model.addVars(I, J, ...)

  with I and J being the index sets for your constraints, then you can access the attribute via

  q[i,j].UnbdRay

  for one particular index i and index j. With this, you can build your feasibility cut.

  Best regards,
  Mario

  0

Post is closed for comments.