Parameters defined by ODE: Parameter Values or Initial Conditions

[PavelBal]

I know that we discussed it again, but I cannot remember what was the solution:

Do parameters that are defined by an ODE belong to the list of Parameter Values or Initial Conditions?

It is more logical to list them under Parameter Values, as those are, well, parameters. The only problem is the scale factor - in theory it should be possible to change the scale factor for them, too. However, it was not possible until now, so it would be OK to let it out in the new version?

@Yuri05 @msevestre @StephanSchaller

[StephanSchaller]

It is more logical to list them under Parameter Values

No, I disagree, for a modeller, if a parameter is defined as a state-variable, the "initial value" you then have to specify should also be listed as "initial condition".

We might need to aligning the nomenclature "initial value" / "initial condition"

[PavelBal]

In this case: if an entry for molecule A in container B is not present in the list, then this molecule has status "not present" in this container. What about the parameter? How does this work together with the idea of only listing parameter whose values are changed?

[msevestre]

Didn't we say we didn't have IC anymore? I think no one complained about scale factor for Parameter and it's safe to assume than no one really missed it

On Tue., Jan. 28, 2020, 5:54 a.m. Pavel Balazki, notifications@github.com wrote:

In this case: if an entry for molecule A in container B is not present in the list, then this molecule has status "not present" in this container. What about the parameter? How does this work together with the idea of only listing parameter whose values are changed?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/Open-Systems-Pharmacology/OSMOSES/issues/15?email_source=notifications&email_token=AAH6GVMIOEQMUTJN2CGLQ43RAAFFRA5CNFSM4KMP4HEKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEKC3S7A#issuecomment-579189116, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAH6GVNYHPRGPL6YFB3EGFDRAAFFRANCNFSM4KMP4HEA .

[msevestre]

Or are we renaming MSV to IC?

On Tue., Jan. 28, 2020, 7:38 a.m. Michael Sevestre, michael@design2code.ca wrote:

Didn't we say we didn't have IC anymore? I think no one complained about scale factor for Parameter and it's safe to assume than no one really missed it

On Tue., Jan. 28, 2020, 5:54 a.m. Pavel Balazki, notifications@github.com wrote:

In this case: if an entry for molecule A in container B is not present in the list, then this molecule has status "not present" in this container. What about the parameter? How does this work together with the idea of only listing parameter whose values are changed?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/Open-Systems-Pharmacology/OSMOSES/issues/15?email_source=notifications&email_token=AAH6GVMIOEQMUTJN2CGLQ43RAAFFRA5CNFSM4KMP4HEKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEKC3S7A#issuecomment-579189116, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAH6GVNYHPRGPL6YFB3EGFDRAAFFRANCNFSM4KMP4HEA .

[StephanSchaller]

I guess this is a technical issue. This simply means the "initial value" is no longer defined at the parameter (NaN, or there simply is no corresponding UI element), but in the "initial conditions" BB.

Also, the R-function "setSpeciesInitialValues" would need to handle parameters that are StateVariables. Again, we might need to change Nomenclature to be precise (e.g. call it "setInitialConditions")

[StephanSchaller]

Or are we renaming MSV to IC?

Correct

[msevestre]

I disagree with the technical issue. Having parameter values potentially in two places is highly confusing don't you think?

On Tue., Jan. 28, 2020, 7:53 a.m. esqlabs GmbH, notifications@github.com wrote:

Or are we renaming MSV to IC?

Correct

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/Open-Systems-Pharmacology/OSMOSES/issues/15?email_source=notifications&email_token=AAH6GVP72ALDABEYKUJXJDDRAATC5A5CNFSM4KMP4HEKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEKDF4CI#issuecomment-579231241, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAH6GVLSXXJZKTQ3TSRKCFTRAATC5ANCNFSM4KMP4HEA .

[StephanSchaller]

Having parameter values potentially in two places is highly confusing don't you think?

Depends who you are asking, I guess (modeller vs lay person). These "parameters" are anyway of a special case... and technically are not (input) parameters at all... but state variables of the model (just like species) So, are "Initial Values" then really Parameter Values? Because for a Parameter defined as State Variable, its (current) value is defined by the differential equation.

[Yuri05]

my vote: Parameter Values. Apart from arguments already given above, if we are going to keep the possibility to switch between "normal" parameter and "ODE"-parameter just via selecting/deselecting a checkbox: having ODE parameter initial value in IC would mean that it must be moved between PV and IC every time user activates/deactivates this option. Not impossible, but seems confusing from users perspective and also means some additional implementation effort.

[PavelBal]

I also vote for Parameter Values, especially with Yuris argument.

[StephanSchaller]

Technically, the (non-ODE) Parameter does not appear in the PV BB by default, but only if changed. Also, I don't think users would switch back and forth between "normal" and "ODE"...

… okay... if you all vote like this… but are the ICs of these "ODE" parameters handled with "setSpeciesInitialValue" at the moment?

[PavelBal]

Also, I don't think users would switch back and forth between "normal" and "ODE"...

I am doing it often...

but are the ICs of these "ODE" parameters handled with "setSpeciesInitialValue" at the moment?

At the moment? You mean, in the "old" r-toolbox?

[StephanSchaller]

At the moment? You mean, in the "old" r-toolbox?

Jop

[PavelBal]

Remains as it is - start value of a state variable parameter is stored in the PV BB (if changed, otherwise in the source BB).