Warm start of MIP solve using R API

Gregory Hunter

• May 15, 2019 21:00

Greetings, all. First, I will say that I am new-ish to Gurobi, and definitely new to deep use of the solver and tuning.

I need to solve ~1600 problems in every execution of my script. (This is all running through an R session.) I am trying to first solve for a single feasible solution, then use that solution to warm start to run for a fixed amount of time or acceptable MIP gap level. I am modifying the model object with a Start parameter between the two calls to gurobi(), but it does not seem that the second call is using that information correctly. Did I miss an example of how to do this somewhere?

0

• 

  Sonja Mars

  Gurobi Staff

  • August 22, 2019 06:34

  Hi Gregory,

  Can you share a log file showing this behavior?

  Thanks,

    Sonja

  0

Please sign in to leave a comment.