Warning: max constraint violation (1.0000e+00) exceeds tolerance

[Jason]

Hi,

I believe I am using the default values for the parameters for an MIP
problem (I never set any specifically). I am loading this MIP problem
from an MPS file that was generated by GLPK (could that affect the
parameter values for the gurobi model?).

It finds a solution which I believe isn't valid and spits out:
Warning: max constraint violation (1.0000e+00) exceeds tolerance

This means that the solution it found isn't valid I thought?

The status code is the same as if it had solved it correctly
though..... is there a way I can detect there has been a constraint
violation other than visually through this output warning?

Below is further detailed output.


Thanks in advance,

Jason


Read MPS format model from file testing_lp.mps
Reading time = 0.00 seconds
: 224 Rows, 202 Columns, 587 NonZeros
status:1
Optimize a model with 224 Rows, 202 Columns and 587 NonZeros
Presolve removed 203 rows and 186 columns
Presolve time: 0.00s
Presolved: 21 Rows, 16 Columns, 72 Nonzeros
Objective GCD is 1
Found heuristic solution: objective 2.0000000

Root relaxation: objective 8.888889e-01, 12 iterations, 0.00 seconds

Nodes | Current Node | Objective Bounds |
Work
Expl Unexpl | Obj Depth IntInf | Incumbent BestBd Gap | It/
Node Time

0 0 0.88889 0 10 2.00000 0.88889 55.6%
- 0s
H 0 0 1.0000000 0.88889 11.1%
- 0s

Explored 0 nodes (12 simplex iterations) in 0.00 seconds
Thread count was 2 (of 2 available processors)

Optimal solution found (tolerance 1.00e-04)
Warning: max constraint violation (1.0000e+00) exceeds tolerance
Best objective 1.0000000000e+00, best bound 1.0000000000e+00, gap 0.0%
status:2

[Greg Glockner]

> It finds a solution which I believe isn't valid and spits out:
> Warning: max constraint violation (1.0000e+00) exceeds tolerance
>
> This means that the solution it found isn't valid I thought?

Yes, this means that the solution is violating one or more constraints. This may arise if you have a large 'big-M' value. If so, try other model formulations or try to reduce the big-M value as much as possible. If your model does not contain any big-M values, please contact Gurobi support directly to investigate further.

[Jason]

I am not familiar with the term big-M value.

Does the fact it finds this solution which violates constraints imply
there is no solutions? Or could there still be and is there a way I
can enforce it to find those?


Thanks so much,

Jason

[Greg Glockner]

> I am not familiar with the term big-M value.

It means a very large coefficient value, such as:

x <= 10e9 y

> Does the fact it finds this solution which violates constraints imply
> there is no solutions? Or could there still be and is there a way I
> can enforce it to find those?

This simply means that the solution violates one or more constraint. It does not imply whether there exists another solution that may satisfy all constraints.

[Jason]

Thanks for the help. I don't have any large coefficient values
(largest ~4).

So is there a way I can have Gurobi find another solution when it does
return this one? How exactly are you supposed to handle a "max
constraint violation" warning.


Thanks again!

Jason