Originally posted by: Torben-Ger-1988

Hey there,

i am trying to optimize a simple Form.

The picture shows you what i'd like to minimize.
Basicly there are two poissibilties how to do something. These possibilities are destribed with x_i
In different locations we have to check which possibility would be smaller. Therefor i added the letter "w".
y_i_w stands for the number of actings you got to do, to finish your work.

For excample:
Location A Possibilty 1: 300 actings
Locations A Possibilty 2: 350 actings

Location B Possibilty 1: 50 actings
Locations B Possibilty 2: 20 actings

I'd like to use the LP to Minimize the Problem. Therefor i added some constrains:

At the moment i entered my code like this:
My main Problem is how do i get my Data into the  objective funktion??

            for (int j = 0; j < w; w++) {
                for (int k = 0; k < i; k++) {
                        Kapa.addTerm(1.0, y[j][k]);
                        Deci.addTerm(1.0, x[j][k]);  

public static void solveMe() {
        
        int i = 2; //possibilities
        int w = 56; //locations
        
        final double M = 99999999;

        //model
        try {

            
            
            IloCplex cplex = new IloCplex();
            
            IloNumVar[][] y = new IloNumVar[w][];

            for (int j = 0; j < w; j++) {
                y[j] = cplex.numVarArray(i, 0, Double.MAX_VALUE);  //actings between 0 and max..
            }
            
            IloNumVar[][] x = new IloNumVar[w][];
            for (int j = 0; j < w; j++) {
                x[j] = cplex.boolVarArray(i);              //xiw either 0 or 1
            }
            
            // Objective Function: Minimize Kapazität
            IloLinearNumExpr Kapa = cplex.linearNumExpr ();
            IloLinearNumExpr Deci = cplex.linearNumExpr();
            for (int j = 0; j < w; w++) {
                for (int k = 0; k < i; k++) {
                        Kapa.addTerm(1.0, y[j][k]);
                        Deci.addTerm(1.0, x[j][k]);  
                }
            }
            
            cplex.addMinimize(cplex.sum(Kapa, Deci));
            
            
            //constraint between Xit and Yit
            for (int j = 0; j < w; j++) {
                for (int k = 0; k < w; k++)
                {
                    cplex.addLe(y[j][k], cplex.prod(M, x[j][k]));
                }
            }
            
        
            //solve model
            cplex.solve();
            //end
            cplex.end();
            
        } catch (IloException e) {
            e.printStackTrace();
            
        }
    }

I think there is a typo in your loop header

for (int j = 0; j < w; w++) {

should probably be (note: j++, not w++)

for (int j = 0; j < w; j++) {

The current loop will just keep adding terms until Java runs out of memory.

There seems to be a similar typo in the loop that sets up the constraints between Xit and Yit:: k should only run up to i, not to w, otherwise you will get an ArrayIndexOutOfBoundsException.

Originally posted by: Torben-Ger-1988

Hey Daniel,

thanks for your advise and fast reply :)

Have you got an idea how i fill in my data?

Maybe i'm describing my problem way too bad or its way too easy :D

How do i have to enter my Data that my cplex optimization implements it to its LP?

Cheerfully regards Torben

Sorry, I don't understand what your problem is. The code you presented looks like it creates the problem you want to create.

I am unclear what you mean by "Data" and what data you want to get into your model. Maybe you can describe what is currently missing from your code/model?

Originally posted by: Torben-Ger-1988

Hey Daniel,

Well my Problem is at the moment how i implement my data in cplex codes.

My Code needs the data of y[i][w] to decide if it takes alternative x[1][w] or x[2][w]

But somehow i need to get my Data into cplex that it's able to compare them.

For example i'd like to input the following data:

int y_1_1 = 300;

int y_2_1 = 250;

int y_1_2 = 100;

int y_2_2 = 250;

and the LP should print out the choosen alternatives (x)

Hope it helps :)

Many Thanks!!

I am still not sure I understand correctly. You want to fix (some of the) y variables and then solve the restricted problem?

In order to fix a variable you can fix its lower and upper bound to the same values:

y[0][0].setLB(300);
y[0][0].setUB(300);

Is that what you are looking for?

Btw, it looks like you are German? If I still did not get you correctly and you feel things are easier to explain in German, then feel free to drop an email in German to daniel(dot)junglas(at)de(dot)ibm(dot)com.