Speed up MIP.

回答済み

Shesha Sreenivasamurthy

• 2024年10月08日 06:44

Hi,

I have a model that has following characteristics

Optimize a model with 1630 rows, 23392 columns and 4329 nonzeros
Model fingerprint: 0x34ba5f74
Model has 21 quadratic constraints
Model has 43782 general constraints
Variable types: 1082 continuous, 22310 integer (21890 binary)
Coefficient statistics:
  Matrix range     [1e+00, 4e+01]
  QMatrix range    [2e+00, 2e+00]
  Objective range  [2e+00, 4e+01]
  Bounds range     [1e+00, 2e+02]
  RHS range        [1e+00, 2e+02]
  GenCon rhs range [1e+00, 6e+01]
  GenCon coe range [1e+00, 4e+01]
Presolve added 60847 rows and 23431 columns
Presolve time: 1.64s
Presolved: 62477 rows, 46823 columns, 186456 nonzeros
Variable types: 152 continuous, 46671 integer (46380 binary)

Hardware

CPU model: Intel(R) Core(TM) i7-9750H CPU @ 2.60GHz
Thread count: 6 physical cores, 12 logical processors, using up to 12 threads

16GB RAM

However, it is very slow. Takes closer to an hour to complete. I am new to LP and followed some suggestions and played with the following params.

# opt_mod.Params.FeasibilityTol = 1e-2
# opt_mod.Params.NoRelHeurTime = 5
# opt_mod.Params.BranchDir = 1
# opt_mod.Params.NodeMethod = 2
# opt_mod.Params.NumericFocus = 2

None of them seem to improve the speed. Do you have any suggestions?

関連するエンティティ

• Performance

0

• 

  Simranjit Kaur

  Gurobi Staff

  • 2024年10月09日 03:40

  Hi Shesha,

  Could you please share the full log of a run? This will help us to identify bottlenecks in the solution process.

  Best regards,
  Simran

  0
• 

  Shesha Sreenivasamurthy

  • 2024年10月10日 07:21

  

  

  I had some issues in the model and therefore, I will fix my issues and will reach out if I have issues. Thank you for your support.

  0

サインインしてコメントを残してください。