Constant reactants are added among the _rates only after substitution

[joefresna]

I don't think that constant reactant should be added to the MuMoTmodel._rates list. Although, in case this is necessary, it should be consistent before and after a substitute()

Shouldn't constant reactants be treated in the same way of other reactants?

[jarmarshall]

Please provide an example or description of why this functionality is problematic, and how a problem arises from using substitute()

[joefresna]

%%model
$
U -> \alpha : \gamma_1
(G) -> U : p
$

model1 = mmt.parseModel(In[2])
model2 = model1.substitute('p = q')
print(model1._rates)
print(model2._rates)

{gamma_{1}, p}
{gamma_{1}, G, q}

issue78.ipynb.zip

[jarmarshall]

OK, but why is this a problem? Where is _rates being used that this would cause a problem?

Arguably the constant reactants should function like rates and be shown as such, since they are parameters of the model

[joefresna]

it's a problem because:

• model1 and model2 are not consistent
• should constant reactants be specified as integer values or as fractions (as the other reactants)?
• should constant reactants appear ordered among other rate widgets or together with other reactant widgets in the 'advaced option' tab?

[jarmarshall]

Not sure this is a huge issue, can tidy up later but; showRates() functions correctly for both models.

Take a look at existing views for how to handle constant reactant widgets - these should already be handled by the base class, and, I believe, should be ordered as constant reactants first, in alphabetical order, then rates in alphabetical order. Check you're not reinventing the wheel!

Not sure how integer vs float for constant reactants; I think they need to be floats >= 0 ... what do you think @tbose1 ?

[joefresna]

constant rate widget is not shown in model1 but only in model2

should this value be considered as a proportion of the system size?

[tbose1]

When deriving the ODEs via Master equation and system size expansion they absorb the volume and will be expressed as fractions just as time-dependent reactants (in the book by van Kampen it seems to be the same). If you apply showODEs_vKE() to the Brusselator, for example, then the constant reactant is expressed as concentration. It is still shown as '(constant reactant)' on the slider - this needs to be changed I think.

[jarmarshall]

@joefresna OK that sounds like a problem and not what I expected @tbose1 I don't fully follow what happens / is recommended so suggest we discuss in person

[jarmarshall]

@jarmarshall to fix this, and make constant reactant slider labels have format \Phi_(X)