Reference point when resolving relative object paths in event assignments

[Yuri05]

At the moment, when the object references in event assignments formulas are resolved, the reference assignment itself is used as the reference point (and not the assigned entity).
I think this is wrong and should be changed.

E.g. check this example and try to create a simulation: Event_Assignment_Path_Resolution.zip

In my spatial structure, I have 2 containers under Organism : Organ1 and Organ2.
Both containers have parameters P1 and P2.

I define a new event, which should set P1 := P2 in both organs (with P2 from the same organ as P1).
So I define a formula P2 with P2 referenced as ..|P2 and use this formula in both event assignments:

Now when trying to create a simulation, I get 2 errors below (1 error per event assignment), which show that the object reference "..|P2" is being resolved relative to the event assignment and not relative to Organism|OrganX|P1

So now I would need to define 2 different formulas - one for each event assignment, where I reference the parameter P1 e.g. with its absolute path (Organism|Organ1|P1 for the 1st assignment and Organism|Organ2|P1 for the 2nd).

This is very inconvenient and should be changed (e.g. with the status quo we cannot avoid formulas duplication described in Open-Systems-Pharmacology/PK-Sim#2856)

[PavelBal]

Is it a use case for a new keyword? :D

I think the current behavior (assignment is the local reference point) follows the logic of the relative path concept within the suite. Changing this behavior would also break many things, I guess. For example, adding the amount of water of an oral application to the total liquid volume in stomach is given by the formula:

Liquid + Water with Water = ..|..|ProtocolSchemaItem|Amount of water the amount of water in the particular administration. This would not work any more if the local reference point were the changed entity Organism|Lumen|Stomach|Liquid.

I see how the solution you proposed can make certain things easier. So maybe we should indeed introduce a new keyword, like ENTITY or CHANGED_ENTITY?

[PavelBal]

So definitely not a bug for me.

[Yuri05]

Is it a use case for a new keyword? :D

I think the current behavior (assignment is the local reference point) follows the logic of the relative path concept within the suite. Changing this behavior would also break many things, I guess. For example, adding the amount of water of an oral application to the total liquid volume in stomach is given by the formula:

Liquid + Water with Water = ..|..|ProtocolSchemaItem|Amount of water the amount of water in the particular administration. This would not work any more if the local reference point were the changed entity Organism|Lumen|Stomach|Liquid.

I see how the solution you proposed can make certain things easier. So maybe we should indeed introduce a new keyword, like ENTITY or CHANGED_ENTITY?

yeah, you are right, these are 2 different use cases and the current use case would not work anymore.
Maybe a new keyword indeed.. need to think more about it.