Originally posted by: open_ball

After getting a respond at https://www.ibm.com/developerworks/community/forums/html/topic?id=dcdacecd-180a-4ce2-b70d-a178a9d4c1ac&ps=25, I decided to create an array while reading my data from a CSV file.

 

I define tuple and fill it. 

        tuple XTuple
        {
        string a;
        string c;
        int t;
        }
        setof(XTuple) Xtup ={<a, c, t> | <c, c2, a> in locset, t in 1..time };

Then, I define an array.

int arrayX[Xtup];

Now, when I read my data set, I need to identify each <a,c,t> and start filling arrayX accordingly.

 

execute{
var f = new IloOplInputFile("X.csv");
while (!f.eof){
        var str=f.readline();
        var ar = str.split(",");
        if (ar.length==4)  {
                myX[] = Opl.intValue(ar[3]);
        }
}     
f.close();
}

Could you help me out how I should fill the inside of myX[] ? 

#DecisionOptimization
#OPLusingCPLEXOptimizer

Hi,

at https://www.ibm.com/developerworks/community/forums/html/topic?id=182a2977-89ac-48ac-9e3b-c4cf62434166&ps=25

I explained how to turn a tuple set into an array

regards

#DecisionOptimization
#OPLusingCPLEXOptimizer

Originally posted by: open_ball

Hi,

That is not what I asked. 

Suppose in OPL, I have the following tuple

A 1 

B 2 

D 5

K 9 

 

When I read data in excel, the rows might be in the following order.

K 9 8

D 5 7

A 1 19

B 2 4

 

Now, when I start reading CSV file, the first row is "K 9 8", but I should select  A 1 19 as the first row and pass that into my array as [<A,1>] =19.

I'd like to do this operation while reading the CSV file. 

 

Currently, I am using two inner for loop and if statement. In other words;

for(var i in OPLTuple)

  for(var j in CSVrows)

    if(i.first == j.first && i.second == j.first)

        myArray [ < j.first,  j.first>] = j.third

 

 

It runs extremely slow due to the size of my problem. Could you tell me if there is an efficient way of doing this operation in OPL?

#DecisionOptimization
#OPLusingCPLEXOptimizer

Instead of your slow loop you should use the function find in a tuple set

regards

#DecisionOptimization
#OPLusingCPLEXOptimizer

Originally posted by: open_ball

I am trying the following, but obviously I am doing wrong. Could you share a small example with me please?

 

for(var i in OPLTuple)

   if(CSVrows.contains(i.first, i.second))

        myArray [ <i.first, i.second] =?

I need to follow the OPL tuple order. On the other hand, CVS file has one column extra. That's why I could not figure out how to deal with it.

 

#DecisionOptimization
#OPLusingCPLEXOptimizer

Hi,

so let me give you a tiny example with 2 options:

tuple t2
{
string st;
int i;
}

tuple t3
{
key string st;
key int i;
int j;
}

{t2}s2=

 {<"A" ,1 >,
<"B", 2 >,
<"D", 5>,
<"K", 9  >};

{t3} s3;
int value[s2];

execute
{
s3.add("K",9, 8);
s3.add("D", 5, 7);
s3.add("A", 1 ,19);
s3.add("B", 2 ,4 );

for(var l in s3) value[s2.find(l.st,l.i)]=l.j;

}

int value2[k in s2]=item(s3,<k.st,k.i>).j;

execute
{
writeln(value);

writeln(value2);
}

 

gives

 

[19 4 7 8]
 [19 4 7 8]

regards

 

Many other how to with OPL at https://www.linkedin.com/pulse/how-opl-alex-fleischer/

#DecisionOptimization
#OPLusingCPLEXOptimizer