infeasible nodes in the solution algorithm

Answered

Iason Liagkas

• May 23, 2023 09:22

Hello guys, in a problem I run I get results. The algorithm used is dual simplex. Some repetitions produce infeasibility in the obj function but the final solution has a gap of 0%. This is the result of the code: 

As you can see on the second picture some objective values are infeasible. Should I be concerned? Is there any comment you have for my results? 

Kind regards
Iason

Related to

• Log Files

0

• 

  Marika Karbstein

  Gurobi Staff

  • May 23, 2023 13:27

  You can find some information about the log in the Gurobi Reference Manual, see MIP Logging:

  The Current Node section provides information on the specific node that was explored at that point in the branch-and-cut tree. It shows the objective of the associated relaxation..

  And a specific node in the branch-and-cut-tree can indeed be infeasible, so nothing to worry about here.

  However, you should check your solution. Gurobi prints warnings about a constraint violation of 1 and a general constraint violation of 2.8593e-03. You could use, e.g., Model.printQuality() to get information about the solution quality containing the constraint name of the most violated constraint.
  The second warning also gives a hint to check (and improve) variable bounds. If there are no or very large bounds on the variables, the approximation of general constraints can be poor.

  0
• 

  Iason Liagkas

  • May 23, 2023 23:13

  

  

  

  Thanks Marika

  0

Please sign in to leave a comment.