RateML 2023

[maedoc]

Thanks @DeLaVlag for initiating the a discussion about the future of RateML at the beginning of the year.

We could discuss some issues/wishes that have been mentioned before:

• move it closer to native LEMS
• Make it more dynamic using template per distinguishable model section
• Adding functionality
  • accommodation for larger models
  • targeting other backends
• The dependency of TVB ontology.

two points in the issue at https://req.thevirtualbrain.org/browse/TVB-3063

• fix state variable limits in generated TVB models?
• not committing generated code to Git repo

Feel free to add other points @i-Zaak @sdiazpier @aperezmartin

[maedoc]

There may be potential interaction with code generation for data fitting with PyMC3/PyTorch if @dionperd or @JohnGriffiths want to participate?

[DeLaVlag]

For the closer to native LEMS discussion: I have set up a repo using LEMS definitions (https://github.com/the-virtual-brain/tvb-root/tree/rateml_lems/tvb_library/tvb/rateML) with the minval, maxval field additions as discussed here (LEMS/jLEMS#102), following the showcase (https://github.com/OpenSourceBrain/TheVirtualBrainShowcase/tree/master/TVB_LEMS)

And for the dynamic templating, is this in the direction of that: https://github.com/the-virtual-brain/tvb-root/tree/master/tvb_library/tvb/simulator/backend/templates ?