Originally posted by: NicoGDFSUEZ

I have the following code in my .mod file :
{string} C = ...;
{string} P[C] = ...;

Then I want to create a variable v which is indexed by C and P[C] like :
dvar float+ v[c in C, p in Pc];
But I have the error message "Variable indexer size not allowed for a generic array".

I found two ways for solving this problem :

• using tuple pair but it is not very convenient
• using another data "{string} P1 = ...;" which contains all the string present in P[C] and then define v as "dvar float+ vC,P1;". The problem of this solution is that it create a lot of useless variables and I would like to avoid this !!!

Does anyone have a "miracle" solution ????
#DecisionOptimization
#OPLusingCPLEXOptimizer

Originally posted by: NicoGDFSUEZ

sorry for the type mistake :
I would like to write : dvar float+ v[c in C, p in P c ; and dvar float+ v C,P1 ;
#DecisionOptimization
#OPLusingCPLEXOptimizer

Originally posted by: DinuN

Hi,
I'm currently facing the same problem, I am trying to define a variable with an index which is varying based on the previous index, and I keep getting the same error "Variable indexer size not allowed for a generic array."

I want to define a variable- dvar boolean x[i in I][j in J][k in K[j]];.

Did you happen to find any solution to the problem? Please let me know if you did. Thank you.

Dinakar
#DecisionOptimization
#OPLusingCPLEXOptimizer

Originally posted by: SystemAdmin

Hi,

what you should try is to define a tupe set "someSet" that contains
all combinations of "i in I, j in J, k in K[j]" and than define
your variable set over "someSet, i.e. you have something like

tuple someTuple{
  int i;
  int j;
  int k;
}
{someTuple} someSet = {<i,j,k> | i in I, j in J, k in K[j]};
dvar boolean x[s in someSet];

(code is not tested)

Hope this will help.

Best regards
Norbert
#DecisionOptimization
#OPLusingCPLEXOptimizer

Originally posted by: DinuN

Thanks Norbert. I did not get any compilation errors, which is good. I'm trying to figure out if it is doing what I want to.
Thanks a lot for your help again.
#DecisionOptimization
#OPLusingCPLEXOptimizer

Originally posted by: SystemAdmin

Hi Nobert,

I faced the same issue, but with a data matrix.
I had

int fa in Ab in B = ... ;
range Ca in Ab in B = 1..f[a][b];
float T a in Ab in B[c in C[a]b] = ... ;

But the programme have given me an error on both last lines.
So I change my structure for the one you suggested.
Maybe, in a case of data arrays, there is a better solution, but now, I have :

tuple TripleIndex{int a;int b;int c;}
{TripleIndex} Index = {<a,b,c> | a in A, b in B, c in 1..C[a][b]};
float TIndex = ... ;

But how do I write down data now that I have only one dimension ?
In other words, how do I say which a,b,c correspond to the floats in my array ?

Because the structure T = [ [ 1 2 3 4] [ 5 6 7 8] ] ; does not work, of course....

Thanks a lot

Adrien
#DecisionOptimization
#OPLusingCPLEXOptimizer

Originally posted by: SystemAdmin

Hi,
I think I found the answer on my on finally...
That kind of writing should work :

T = #[
<a,b,c> : k
<a',b',c'> : k'
...
]#;

That's it...
But there may be smarter ways to it...

I'm here listening if that's so...
#DecisionOptimization
#OPLusingCPLEXOptimizer