Adding events macro to reaction network DSL

[tstrey]

One possible way of doing this is the following:

rn = @reaction_network rxs begin
     @variables t
     @parameters k
     @species A(t) B(t)

     k, A --> B
     @discrete_events begin
          (t == switch_time) => [B ~ 0.0]
     end
     @continuous_events begin
          (A < switch_time) => [B ~ 3.0]
     end
end

However, I think by having the macro inside @reaction_network you loose some flexibility. For example, if you wanted to use rn with different events. It could also be argued that conceptually a reaction network is how the species interact with each other but, the events, being external added boundary conditions, merging the two would not be a good idea.

[isaacsas]

I think this looks good. Can't one have multiple events within the begin/end blocks? Or by "different events" do you mean if you want to change the events in an already created system?

The general design, which we admittedly don't always adhere to, is that systems should be immutable. So if you want to change something in a system you need to generate a new system with the desired properties. ReactionSystems right now, encompass a full model for a system that involves reactions. This includes coupled events, algebraic equations, and differential equations (along with variables that might not represent chemical species). So they are more a container for a model than just reactions.

Even from the current symbolic interface, where events are created as kwargs to ReactionSystem, you can't really change them once the system is created.

But maybe I've misunderstood your point?

[isaacsas]

Also, one can always create a rn without events, and then extend it with a system that has the desired events to generate a merged system of reactions and events. So in that way you can keep around an event-free ReactionSystem.

[TorkelE]

now works on master