Originally posted by: pickup

i have a problem when i execute my model, syntax error in the objective function : this is not an array type

the objectiv function is : sum v in V sum i in I sum  j in J  ( C_v * DIJ_ij * XIJV_ijv )

in cplex : 

 

#DecisionOptimization
#OPLusingCPLEXOptimizer

Hi,

 

can you attach your model ?

 

regards

#DecisionOptimization
#OPLusingCPLEXOptimizer

Originally posted by: pickup

hi, 

this is my model, thank you 

regards 

#DecisionOptimization
#OPLusingCPLEXOptimizer

Hi

if you add teh key keyword in

tuple distance{
 key int i;
 key int j;
 int valeur;
}

 

and

 

tuple coutveh{
  key int v;
   int  coutvehicule;
}

 

then if you write

minimize
sum (v in V)
  sum (i in I)
    sum ( j in J)
item(CV,v).coutvehicule*item(DIJ,<i.i,j.j>).valeur*Xijv[i][j][v];

subject to
{
}

 

the model will work fine

 

regards

 

 

#DecisionOptimization
#OPLusingCPLEXOptimizer

Originally posted by: pickup

 

thank you Mr. Alex ,

and how we can define this two set in a graph G=( N, A)

 SEGP (i) = {j in N : (i, j) in A} : is all successors of node i

 SEGM (i) = {j in N : (j, i) in A} :all predecessors of node i

in my model I defined as : 

 

 

but i have a problem when i execute my model, syntax error in the contraints: 

1) The type Ssegplus {} can not be used to userFunction.

2) The type Ssegmoin{} can not be used to userFunction.

 

sum (v in V, j in SEGP[i]) Xijv[i][j][v]==1;

 

#DecisionOptimization
#OPLusingCPLEXOptimizer

Hi,

are sets.

SEGP[i] is an array indirection.

Instead of

sum (v in V, j in SEGP[i]) Xijv[i][j][v]==1;

have you tried anything like

sum (v in V, j in SEGP:j.n==i) Xijv[i][j.s][v]==1;

regards

#DecisionOptimization
#OPLusingCPLEXOptimizer

Originally posted by: pickup

hi,

yes i try it but i have problem : 

sum (v in V, j in SEGP:j.n==i) Xijv[i][j.s][v]==1;

Not available for int operator == <i:int>.

 

regards

 

#DecisionOptimization
#OPLusingCPLEXOptimizer

Ok then

sum (v in V, j in SEGP:j.n==i.i) Xijv[i][<j.s>][v]==1;
 

#DecisionOptimization
#OPLusingCPLEXOptimizer

Originally posted by: pickup

hi 

thank you it works now,

and for this 2 constraint please :

c4: forall ( v in V)  ( sum of  i in SEGP (0) ) X_0iv=1 ; to say all successors of node  0 ( 0 is the index of dépot) 

c5: forall ( v in V)  ( sum of  i in SEGM (0) ) X_i0v=1 ; to say all predecessors of node  0 ( 0 is the index of dépot) ,

IN CPLEX :

 

syntax error :

 

regards

 

#DecisionOptimization
#OPLusingCPLEXOptimizer

Hi

 

the syntax is ok for

c4 : forall (v in V)
sum (i in SEGP:i.n==0)Xijv[item(ICF,0)][<i.s>][v]==1;

so I will let you adapt it so that is does what you need.

 

Regards

 

 

#DecisionOptimization
#OPLusingCPLEXOptimizer