Error for not supported parameter names

[shoepfl]

Dear antimony team,

we currently conduct a reproducibility study with students. One team found out that the parameter name "gamma" is not working with antimony. However, the error message is very unspecific and it took us a while to figure out what the problem was:

Exception: Antimony: Error in model string, line 14: syntax error, unexpected '=', expecting '.' or element name or "'is'"

There should be some warning or specific error message for names that produce this error.

Attached you can find a working model without the parameter gamma (called "g" instead) and the antimony string that produces the error:
SERID antimony.zip

Best,
Sebastian

[luciansmith]

Thank you for your report! This is actually kind of hard to support, because of design decisions I made 15 years ago that weren't great, but I'll see what I can do. The problem is that anything predefined in Antimony as one thing can't then be used as anything else, so you can't have a compartment named 'compartment' because 'compartment' is a keyword, for example. For this, 'gamma' is the name of a predefined distribution function, so now you can't use it elsewhere.

It's perfectly fine in SBML to mix things like this, and some day I might be able to recode the parser to be more flexible. In the mean time, I'll see if there's any tweaks I can make to the parser to at least be a little more clear what the problem is.

[shoepfl]

Hi Lucian,

thanks for your answer. Of course, I understand that not all predefined names can simply be given freely. With compartment etc. I would have thought of it sooner, but gamma was not on my mind at first. I think a simple warning or error message indicating which reserved names exist would be sufficient.