Eliminating numerical problems in the model

Answered

Hur Butun

• May 29, 2019 08:54

Hello,

I have a very large MILP problem. When I try to solve it using Gurobi with AMPL I receive the following message. 

Gurobi 8.0.0: outlev = 1
mipgap = 0.2
basis = 0
Optimize a model with 281296 rows, 377144 columns and 1629986 nonzeros
Variable types: 352635 continuous, 24509 integer (24509 binary)
Coefficient statistics:
Matrix range [1e-03, 7e+10]
Objective range [2e-01, 9e-01]
Bounds range [5e-01, 7e+07]
RHS range [4e-03, 1e+09]
Warning: Model contains large matrix coefficient range
Consider reformulating model or setting NumericFocus parameter
to avoid numerical issues.

 

If I set numericfocus=3 solution time gets significantly longer, so at this point, I think my only option is to reformulate the model. However, the model is quite elaborate and it is difficult to spot what is causing the numerical problems. Is there any way I can see where the minimum and maximum coefficients come from? 

Hur

0

• Official comment

  

  Kostja Siefen

  Gurobi Staff

  • June 28, 2019 09:09

  Please take a look at our Numerical Guidelines: https://www.gurobi.com/documentation/current/refman/numerics_gurobi_guidelines.html
• Official comment

  

  Simranjit Kaur

  Gurobi Staff

  • December 06, 2024 00:58

  This post is more than three 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 why not try our AI Gurobot?.

Post is closed for comments.