Save runtime in a list

回答済み

Lorenz Wagner

2023年07月07日 13:27

• 
• 
• 

Hi, I would like to run my model through different loops, always storing the runtime of each pass in a list. I have managed to do this, but there are two things I would like to adjust to make it more like what I want.

 

1) If the model is infeasible, then the list should say "infeasible".

2) If the model takes longer than e.g. 5000 seconds to solve, then the next iteration should be started and entered in the 5000 list.

 

I use m.Runtime to store the duration.

The list is called sec = []

Thanks for t

he help

0

• 

  Maliheh Aramon

  • Gurobi Staff

  2023年07月08日 01:34 編集済み

  Hi Lorenz, 

  What does each iteration represent here? 

  Assuming at each iteration, you are solving a model independent of other iterations, the script below solves model A with different seed values \(n\) times and saves the result as you described in your post:

  runtimes = []
  for i in range(n):
      # Set the Seed
      model.params.Seed = i
      # Set the time limit
      model.params.TimeLimit = 5000
      model.optimize()
      
      # If the model solves to optimality, save the runtime
      if model.status == GRB.OPTIMAL:
          runtimes.append(model.Runtime)
      # If the model times out, save the time limit as the runtime
      elif model.status == GRB.TIME_LIMIT:
          runtimes.append(model.params.TimeLimit)
      # If the model terminates with any other status including infeasible status, 
      # save the model status
      else:
         runtimes.append(model.status)

  Best regards,

  Maliheh

  0

サインインしてコメントを残してください。