メインコンテンツへスキップ

Infeasibility Relaxation with Gurobi

回答済み

コメント

7件のコメント

  • 正式なコメント
    Simranjit Kaur
    • Gurobi Staff
    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.
  • Riley Clement
    • Gurobi Staff

    Hi Nitin,

    Note that in the documentation you linked it states the following for the vars parameter:

    vars: Variables whose bounds are allowed to be violated. 

    The function is expecting a collection of variables, such as a list, which is why vars=False is causing an error.

    - Riley

    0
  • Nitin Singh
    • Gurobi-versary
    • Conversationalist
    • First Question

    But what if I don't want the relaxation to play with the variable bounds, but only the RHS of the constraints? 

    0
  • Riley Clement
    • Gurobi Staff

    Passing an empty list as an argument is acceptable.

    - Riley

    1
  • Nitin Singh
    • Gurobi-versary
    • Conversationalist
    • First Question

    I just did that. Thanks.
    My bad that I thought it was a boolean.

    0
  • Nitin Singh
    • Gurobi-versary
    • Conversationalist
    • First Question

    Riley Clement
    I have a follow up question on this. For the same infeasible model (now I am not filtering down any constraint for this relaxation exercise):
    If I try model.feasRelax as follows:

    self.gurobi_model.feasRelax(
                  relaxobjtype=0, minrelax=False, vars=[], lbpen=None, ubpen=None,
                  constrs=self.gurobi_model.getConstrs(), rhspen=None
                )

    The model still remains infeasible, however if I use the model.feasRelaxS as:

    self.gurobi_model.feasRelaxS(
                  relaxobjtype=0, minrelax=False, vrelax=False, crelax=True
                )

    , I am able to arrive at a possible feasibility relaxation. 

    What is likely to cause this difference in the behavior?

    Thanks

    0
  • Riley Clement
    • Gurobi Staff

    Hi Nitin,

    The rhspen parameter needs values if you are providing constraints to be relaxed.

    rhspen: Penalty for violating a linear constraint. One entry for each constraint in argument constrs. 

    If you use

    rhspen=[1]*len(m.getConstrs())

    then you will get equivalent results to feasRelaxS.

    - Riley

    0

投稿コメントは受け付けていません。