objective -Infinity

[Ying WANG]

Hi all,

The same code run smooth in a smaller scale system, but cannot work for a larger scale system. The objective is -Infinity. Have you ever met similar problem?

Thanks,

Ying

CPLEX 12.4.0.0: mipgap 0.005

mipdisplay 2

timing 1

timelimit 3600

treememory 2048

MIP Presolve eliminated 38761 rows and 17480 columns.

MIP Presolve modified 47881 coefficients.

Reduced MIP has 138559 rows, 98251 columns, and 524393 nonzeros.

Reduced MIP has 10296 binaries, 0 generals, 0 SOSs, and 0 indicators.

Probing fixed 0 vars, tightened 5687 bounds.

Probing time =    0.27 sec.

MIP Presolve eliminated 1082 rows and 479 columns.

MIP Presolve modified 5539 coefficients.

Reduced MIP has 137477 rows, 97772 columns, and 519990 nonzeros.

Reduced MIP has 10383 binaries, 0 generals, 0 SOSs, and 0 indicators.

Found feasible solution after 5.23 sec.  Objective = -1.#INFe+000

Probing time =    0.08 sec.

Clique table members: 4098.

MIP emphasis: balance optimality and feasibility.

MIP search method: dynamic search.

Parallel mode: deterministic, using up to 4 threads.

Root relaxation solution time =    6.07 sec.

        Nodes                                         Cuts/

   Node  Left     Objective  IInf  Best Integer    Best Bound    ItCnt     Gap

*     0+    0                     -1.#INF0e+000                  10652     --- 

      0     0        cutoff       -1.#INF0e+000                  10652    0.00%

Root node processing (before b&c):

  Real time             =    8.19

Parallel b&c, 4 threads:

  Real time             =    0.00

  Sync time (average)   =    0.00

  Wait time (average)   =    0.00

                          -------

Total (root+branch&cut) =    8.19 sec.

Times (seconds):

Input =  0.604

Solve =  13.156

Output = 0.524

CPLEX solution status 12 with fixed integers:

phase II objective limit exceeded

IBM ILOG CPLEX Optimization Studio Commercial Edition                        

CPLEX 12.4.0.0: optimal integer solution; objective -Infinity

10652 MIP simplex iterations

0 branch-and-bound nodes

Status recovering solution: phase II objective limit exceeded

"option abs_boundtol 2.842170943040401e-14;"

or "option rel_boundtol 3.513190288059828e-16;"

will change deduced dual values.

[Robert Fourer]

It appears that CPLEX found a feasible solution to your constraints that gives a value of "minus infinity" for your objective function. Probably this means, as a practical matter, that the objective value is minus some number that is so large that it can't be represented in the computer.

To check this possibility, look closely at your solution and your objective function to see what might be causing a (negative) very large value. Note that the cause could be very large values of some variables, or very large coefficients in the objective; to check the objective, you can use AMPL's "solexpand" command -- for example, "solexpand TotalCost;" -- to see the actual coefficients that AMPL sent to the solver.

Bob Fourer
am...@googlegroups.com

=======

From: am...@googlegroups.com [mailto:am...@googlegroups.com] On Behalf Of Ying WANG
Sent: Wednesday, June 14, 2017 2:37 AM
To: AMPL Modeling Language
Subject: [AMPL 14215] objective -Infinity

The same code run smooth in a smaller scale system, but cannot work for a larger scale system. The objective is -Infinity. Have you ever met similar problem?

---

[Ying WANG]

Thank you Bob.

I tried it. But it did not show any large coefficient. Could it be the dual problem has -infinity objective?

 

Ying

在 2017年6月15日星期四 UTC+8下午10:14:01，Robert Fourer写道：

[Robert Fourer]

If the coefficients are not large, then look at the values of the variables to see if they are large. If neither the coefficients nor the variables are large, then give the command "display ...;" where ... is replaced by the name of your objective function.

-- If this command shows -Infinity for the value of the objective function, then you will need to look more closely at the values of the coefficients and the values of the variables to understand why the objective has an infinite value.

-- If this command does not show -Infinity, then we may need to examine the problem more closely. In this case we suggest that you send us model and data files that we can use to reproduce the situation on our computers. If they are not too large, you can attach these files to an email, either to this group or to sup...@ampl.com.