Originally posted by: char507

Hi,

I asked a question similar to this a while back but I thought I would make everything more clear here. A vast simplification is that our model is trying to assign assets to customers. Each customer starts with an asset, then the model looks at the assets available and determines if a different asset would be more efficient. The data is defined as follows:

tuple Asset
{
key string id
float cost
int numberAvailable
}

tuple Customer
{
key string id
string currentAssetId
}

There are a number of places in the model where we currently link tuple objects by Id and then use filtering the constraints to link the objects together. My guess is that this model would be more efficient if we defined tuples inside of tuples and eliminated the filtering by constraints using a definition like so:

tuple Customer
{
key string id
Asset currentAsset
}

{Customer} Customers = ...;

My problem is that I do not know how to instantiate the Customer objects using a .NET CustomDataSource. Once I have loaded the list of assets through the data handler, I move to the customer data and write the following:

handler.StartElement("Customers");
handler.StartSet();
foreach(Customer c in customerList)
{
handler.StartTuple();
handler.AddStringItem(c.Name);
handler.AddStringItem(c.CurrentAsset.Id);
}
handler.EndSet();
handler.EndElement();

My question is: What goes in place of "handler.AddStringItem(c.CurrentAsset.Id)"? When I define a tuple inside of a tuple, how do I instantiate the data with a link to the sub-tuple object?

I would be happy to answer any clarifying questions and I would really appreciate some help!
#DecisionOptimization
#OPLusingCPLEXOptimizer

Originally posted by: SystemAdmin

Have you tried:

handler.StartElement("Customers");
handler.StartSet();
foreach(Customer c in customerList)
{
handler.StartTuple();
handler.AddStringItem(c.Name);
 
handler.StartTuple();
handler.AddStringItem(c.CurrentAsset.Id);
handler.AddNumItem(c.CurrentAsset.Cost); //assuming you can access cost like this
handler.AddIntItem(c.CurrentAsset.numberAvailable); 
handler.EndTuple();
 
handler.EndTuple();
}
handler.EndSet();
handler.EndElement();

Regards,
Faisal
#DecisionOptimization
#OPLusingCPLEXOptimizer

Originally posted by: char507

Sorry it took me so long to get back to you. Turns out there was a bug in one of my constraints once I made that formulation. Looks like that method works now.

Just for completeness of the thread and for my understanding of OPL:

• If I have already added all of the assets into the "Assets" tuple set: this method (adding the asset again into the Customer tuple) simply looks up the already existing item in the Asset set and creates a reference to it? I want to be sure i'm not duplicating data.

• If i'm accessing data that belongs to the customer's asset in my constraints, is this method more efficient than using an AssetID string and linking the customer and asset in that fashion?

#DecisionOptimization
#OPLusingCPLEXOptimizer