Objective function with Knitro

GAMS
1

Dear Gams world community,

I have a problem with knitro solver.
I have an MPEC program, that gives a locally optimal solution. However, when I check the output file, the level of the objective function equation (in SolEQU) is different than the level of the obj variable defined by this equation, as follow:


LOWER LEVEL UPPER MARGINAL

---- EQU objDSO 38654.568 38654.568 38654.568 1.000


LOWER LEVEL UPPER MARGINAL

---- VAR objD -INF 37828.887 +INF


ObjD is defined in the objDSO equation as follow:


objDSO…

objD =e= sum((t),sum((i),-…

Anyone has had this problem before?

BG,

2

Hello,

Why do you consider this to be a problem?

-Steve

3

Hello,
Because it should be zero, no?
As the sum of the two-sides variables in the equation.