Originally posted by: SystemAdmin

[typhoon said:]

yes: it´s nearly weekend, but
no, i didn´t find the solution yet:

after the desperate effort of compiling my opl-code, the "journal of the motor" provided me with the following statement:

Infeasibility row 'id80':  0 <= -1.<br />Presolve time =    0.00 sec.

so, i checked every array with the "<=" sign by adding "-30" on the left side...to be sure (29 seemd not enough for me), but:<br />nothing changed. i continued searching in the forum-without results-and hope therefore, that one of you could tell me WHERE OR HOW TO FIND THE ROW WHICH CORRESPONDES TO THE MAGIC 'id80' !!!!

thank u very much for any statement!
for every useful hint, i propose a beer in paris as compensation...

nice weekend,

maximilian
#DecisionOptimization
#OPLusingCPLEXOptimizer

Originally posted by: SystemAdmin

[Didier Vidal said:]

Typhoon,

Which version of OPL do you have ? Recent OPL versions can find conflicting constraints. You will find explanations about this feature in doc section OPL IDE / IDE Tutorials / Relaxing infeasible models.

Didier, Paris (:-) )
#DecisionOptimization
#OPLusingCPLEXOptimizer

Originally posted by: SystemAdmin

[alain.chabrier said:]

Also, if you label your constraints in OPL, the rows in  CPLEX will have names... that might help!

Alain
#DecisionOptimization
#OPLusingCPLEXOptimizer

Originally posted by: SystemAdmin

[typhoon said:]

Thanks for your answers, i will try it out..!

Never the less, i don´t think that the problem is produced by a constraint: i disactivated every constraint one by one and it didn´t change the "error"-message (in the "journal of the motor").
BUT, when i changed the location of the objectif function, the message changed to "id68" and "id70" (the last one with abs.()-function, the first one without) , and i provoced "id79" and "id80" (same difference like before) for the objectif function initialized in the constraint-part of the programm.
that seems all very strange to me...even more when have to think about the reason WHY the right side of the equation endues a negative value°!??? 

thank u very much for your ideas, i´m realy glad of having found the forum!

a la prochaine,

max
#DecisionOptimization
#OPLusingCPLEXOptimizer

Originally posted by: SystemAdmin

[alain.chabrier said:]

If you have logical expressions in your objective then it can be an internal variable created during the linearization of it.
That would be strange however (not sure possible) that the linearization of the objective becomes infeasible.

Alain
#DecisionOptimization
#OPLusingCPLEXOptimizer

Originally posted by: SystemAdmin

[typhoon said:]

no,...there are no logical expressions in it.  maybe you could have a look on my code. but i don´t know what i could propose u as exchange...tell me what you can imagine!

max
#DecisionOptimization
#OPLusingCPLEXOptimizer