WIP: AdEx model update

[i-Zaak]

The purpose of this pull request is to consolidate the various implementations of the AdEx model, create relevant demos, and change the approach taken to derived parameters for this model.

[i-Zaak]

For start, I've updated the first order Zerlaut model according to the EBRAINS gitlab repository which is linked from the paper:

• https://gitlab.ebrains.eu/kancourt/tvb-adex-showcase3-git/-/blob/master/tvb_model_reference/src/Zerlaut.py#L10

In particular, this meant splitting the p_connect parameter into p_connect_e and p_connect_i. In addition the tvb-adex-showcase3 variants have additional noise state variable - this is something we need to discuss, as it would be cleaner to have the (colored) noise handled by the Noise module.

@rcofre will lead this from here.

[maedoc]

isn't the OU something you could do with the colored noise support?

[i-Zaak]

Well, not really unfortunately. The issue is that the noise drive enters same as the other external inputs (e.g. connectivity coupling), and is then dragged through the transfer functions:

https://github.com/the-virtual-brain/tvb-root/blob/6d075f377d2411743534e27bfccb47fc754968e0/tvb_library/tvb/simulator/models/zerlaut.py#L714)

[i-Zaak]

One thing I've noticed is that the first order model will need a more careful review (at least the dfun). To be done together with the examples reproducing plots from papers .

[rcofre]

Thanks Jan for this first update.

Best,

R.

On Mon, 27 Jun 2022 at 22:48, Jan Fousek @.***> wrote:

For start, I've updated the first order Zerlaut model according to the EBRAINS gitlab repository which is linked from the paper:

https://gitlab.ebrains.eu/kancourt/tvb-adex-showcase3-git/-/blob/master/tvb_model_reference/src/Zerlaut.py#L10

In particular, this meant splitting the p_connect parameter into p_connect_e and p_connect_i. In addition the tvb-adex-showcase3 variants have additional noise state variable - this is something we need to discuss, as it would be cleaner to have the (colored) noise handled by the Noise module.

@rcofre https://github.com/rcofre will lead this from here.

— Reply to this email directly, view it on GitHub https://github.com/the-virtual-brain/tvb-root/pull/573#issuecomment-1167877759, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADCMHUFFMYWEAG3LLEFK2V3VRIHTTANCNFSM5Z7WAJKA . You are receiving this because you were mentioned.Message ID: @.***>

[i-Zaak]

I've consolidated the notebooks and reviewed the changes to the model. If there are no objections, I'll merge early next week. Further potential improvements and fixes to be done in separate pull requests.