Currently moose.Function variables must be named as x0, x1, ...., or y0, y1, etc.. I suggest that in addition to this, moose.Function should also accept any keyword prefixed with var_ or v_ (or something similar) as variables. This will make expressions more readable e.g. var_CAMKII over x0 representing CAMKII etc.
Anonymous
This had been thought about and discussed before. Implementing this requires that either (1)
Cinfoallows dynamic creation of named fields on any element, or (2) the entries in aFieldElementbe allowed different names.The constants can have arbitrary names because they are local to the
Functionelement and a map can be used for translating the identifier to an entry in the 'c' field. However, this scheme poses a problem withValueFinfos looked up from other objects (y) andDestFinfos filled in by other objects (x) because a connection needs to be setup usingaddMsg, a function which knows only the Cinfo of the class and is agnostic of the local variables of the elements it is connecting. So even if a map was used to translate between variable names and x/y index, the user will have to connect messages for them by index - which is even more cumbersome than just translating the formula by renaming all variables x{i}/y{i}.You are welcome to further discuss and improve the implementation.
Last edit: Subhasis Ray 2015-03-28