Execution Errors after starting from Scratch

[Tanzeel]

Dear All,

Please let me know about the errors as I am unable to find the causes. The model before starting from scratch is working well.

[Vikrant]

Tanzeel,

Please post the *.LST file.

[Tanzeel]

Tanzeel,

Please post the *.LST file.

I don't know that what's happening here. I can't attach the *.LST file here.

[Antti-L]

I think at least *.TXT, *.DOC*, *.XLS* and *.ZIP files can be directly uploaded. For uploading other files, like listing files, I would suggest to ZIP-compress the file(s) first, and then upload the ZIP file. The size of a listing file will be much smaller when zipped.

[Tanzeel]

I think at least *.TXT, *.DOC*, *.XLS* and *.ZIP files can be directly uploaded. For uploading other files, like listing files, I would suggest to ZIP-compress the file(s) first, and then upload the ZIP file. The size of a listing file will be much smaller when zipped.

Right!

Please check it.

[Tanzeel]

I think at least *.TXT, *.DOC*, *.XLS* and *.ZIP files can be directly uploaded. For uploading other files, like listing files, I would suggest to ZIP-compress the file(s) first, and then upload the ZIP file. The size of a listing file will be much smaller when zipped.

Right!

Please check it.

Thanks. Adjusted now. Any good reference to debug model from LST file.

[Antti-L]

Hi. The listing file tells us that the error is the model being "unbounded or infeasible". In addition, it tells that you have some input data errors.  These two issues could be either linked together or not. Anyway, you should correct the errors in your input data, shown at the top of the file (lines 20–90).

If your model remains infeasible, you can find the infeasibilities by enabling the Cplex Feasibility Relaxation option (feasopt 1), or by enabling the infeasibility finder option (iis 1).  See the Cplex manual included in your GAMS distribution. I am not sure how you could best diagnose an unbounded condition with Cplex.

Anyway, if the model was earlier running ok, I think you should best diagnose the problem by comparing the data only GDX files between the earlier run and the new failing run, to find the crucial differences in your input data, which are causing the problem.

[Tanzeel]

Hi. The listing file tells us that the error is the model being "unbounded or infeasible". In addition, it tells that you have some input data errors.  These two issues could be either linked together or not. Anyway, you should correct the errors in your input data, shown at the top of the file (lines 20–90).

If your model remains infeasible, you can find the infeasibilities by enabling the Cplex Feasibility Relaxation option (feasopt 1), or by enabling the infeasibility finder option (iis 1).  See the Cplex manual included in your GAMS distribution. I am not sure how you could best diagnose an unbounded condition with Cplex.

Anyway, if the model was earlier running ok, I think you should best diagnose the problem by comparing the data only GDX files between the earlier run and the new failing run, to find the crucial differences in your input data, which are causing the problem.

Thank You Antti-L.

I check this workflow more closely. As now, I am doing all these pure manually. Relaxing constraints one by one to resolve in-feasibility rather to do systematic debugging. It is really seem hard to find "Presolve found the problem infeasible or unbounded". It takes a lot of time. I check these options.