Parameters to new synaptic model

[chuy...@gmail.com]

I followed the example DropOddSpike to develop a new synaptic model in nest.  Now this works in PyNN, but I can't seem to set any of the parameters from PyNN.  It sets the weight when I build the topology, and then the weight is changed via learning.  However, I can't figure out how to set internal synaptic parameters (like Wmax, the maximum weight) when I create the synapse.  I can do this in the predefined models.  Is there a special mechanism for setting the parameters?

[Andrew Davison]

Hi Chris,

> On 15 Jan 2018, at 15:05, chuy...@gmail.com wrote:
>
> I followed the example DropOddSpike to develop a new synaptic model in nest. Now this works in PyNN, but I can't seem to set any of the parameters from PyNN. It sets the weight when I build the topology, and then the weight is changed via learning. However, I can't figure out how to set internal synaptic parameters (like Wmax, the maximum weight) when I create the synapse. I can do this in the predefined models. Is there a special mechanism for setting the parameters?

If you run

nest.GetDefaults(‘my_model_name’)

do you see the parameters you want to set?


Cheers,

Andrew

[chuy...@gmail.com]

No, I get weight (but I don't get weight_ as it is in the file (oja_synapse.h) I don't get the other variables defined in my file.  I've intentionally picked one that another synapse type uses Wmax_.  Do I need to do something in nest to register these variables so pynn (or sli?) can see them?

[Andrew Davison]

> On 15 Jan 2018, at 17:05, chuy...@gmail.com wrote:
>
> No, I get weight (but I don't get weight_ as it is in the file (oja_synapse.h) I don't get the other variables defined in my file. I've intentionally picked one that another synapse type uses Wmax_. Do I need to do something in nest to register these variables so pynn (or sli?) can see them?

Yes, I think so. Maybe take a look at some of the existing NEST synapse models to see what they do? Otherwise I suggest asking on the NEST users mailing list.