Creating variable for each linear expression

[aurelije]

Hi linopyers, What is the recommended way to create variable for each linear expression? The context is that operation has been done on multidimensional variable resulting in LinearExpresssion object. I need to set soft constraint for that expression so I need variable per each linear expression to create constraint.

For me it was logical to do something like this:

soft_constraint_deviation_var = model.add_variables(
    0, 
    100, 
    coords=linear_expression.coords, 
    name='soft_constraint_deviation_var')

Variable gets created but if I try to print it I get KeyError: '_term'

The thing that worked (but it looks like too much of work) is:

soft_constraint_deviation_var = model.add_variables(
    linear_expression.assign(lower=0).data.lower, 
    linear_expression.assign(upper=100).data.upper, 
    name='soft_constraint_deviation_var')

Maybe in api that is made more for OR people and less for XArray people, we could have an method to generate soft constraint and/or deviation varaible.

[FabianHofmann]

good question! (EDIT HERE, previous answer was wrong) how does linear_expression.coords look like?

[FabianHofmann]

ah, you could take linear_exprssion.indexes instead of .coords

[aurelije]

It worked, just it is a bit misleading that for parameter named coords you need to send indexes as an argument

[FabianHofmann]

I know, it would be nice if that worked. but I would not know how to realize that. I am closing this issue, but if you don't mind you are could to document your use cases so that we can incorporate them into the documentation later.