Extend LEMS with *optional* max/min/step for Parameters/Constants/StateVariables

[pgleeson]

Motivation: to allow TVB/RateML (and other simulators) to specify recommended max/min for parameters in a valid LEMS file, e.g.
https://github.com/the-virtual-brain/tvb-root/blob/a2d64896bb8fa9ea7da3c10d57b177bcb0889101/scientific_library/tvb/rateML/XMLmodels/kuramoto.xml#L6

This would probably result in something like:

OLD: <Parameter name="global_speed" dimension='-1.0, 1.0'/>
NEW: <Parameter name="global_speed" min_recommended_value='-1.0' max_recommended_value='1.0'/>

also for Constants:

<Constant name="x0"  value="-1.6" description="..." min_recommended_value='-2.0' max_recommended_value='2.0' recommended_step='0.1'/>

and StateVariables:

<StateVariable name="y1" exposure="None" min_expected_value='-2.0' max_expected_value='2.0'/>

Note:

• These should be treated in implementations as optional metadata, and can be ignored by the simulators, etc. hence "recommended" and "expected".
• They will be particularly useful for generating GUIs etc. to give recommended values for sliders, setting bounds in plots etc.
• I will aim to add these to the Java impl to throw warnings if the bounds are exceeded.

FYI @DeLaVlag @maedoc

[DeLaVlag]

I created a PR for adding the field to pylems definitions. I have setup a draft for rateml models using native lems here: https://github.com/the-virtual-brain/tvb-root/tree/rateml_lems/tvb_library/tvb/rateML , following the showcase models depicted here: https://github.com/OpenSourceBrain/TheVirtualBrainShowcase/tree/master/TVB_LEMS

[pgleeson]

Thank you for this @DeLaVlag! We're just coming out of an intense grant writing period any I'll make some time to look at this in the next few days. TVB and RateML support are still very much on our radar!