Originally posted by: arjaree

Hi there,

I have a problem in using summation. One index of my 3D-decision variable is depending on another index. And I need to use summation for that variable. Is it possible to code that?

decision variable = Inv[i][b][t]

b in 1..M[i]

therefore,

forall (i in I, t in T)

sum(b in M[i]) (Inv[i][b][t]) == demand[i][t] + order[i][t] 

Hi,

let me give you an example:

range I=1..2;
range T=1..2;
{int}   M[I]=[{1,2}, {2}];

{int} B=union (i in I) M[i];

int demand[I][T];
int order[I][T];

dvar int Inv[I][B][T];

subject to
{

forall (i in I, t in T)

sum(b in M[i]) (Inv[i][b][t]) == demand[i][t] + order[i][t] ;
}

regards

Originally posted by: arjaree

Hi,

Thanks a million for your help, Alex. Now I'm facing some more problems.

1. I can't use      forall (b in M[i]) ....what is the correct way of coding that?

2. Let   {int} X[I]=[{2}, {3}];

             {int} C=union (i in I) X[i];

    then I can't use    sum(i in I, t in T) (Inv[i][X[i]][t];         

3. If i have one extra array dimension (makes it to be 4D) and this b-index is depending on both the i-index and this new index-r as well...will it also works the same way?

Thanks for your help in advance

Kind Regards

Originally posted by: arjaree

for clarification,

in 2...X[I] contains maximum numbers from each M[I]

thanks