10 Nov
2021
10 Nov
'21
12:30 p.m.
Dear Aadhar,
Enable and Disable functions for structural plasticity apply to the
whole network. What I do to disable subnetworks is to set their growth
rate to 0 and back to the original value when you want to activate them.
This is possible because you can assign different growth rules to each
subnetwork and/or to each synapse type which can grow between specific
synaptic elements.
Below you can see an example for NEST 2.20.
All the best,
Sandra
**********
# At network definition
growth_curve = {'growth_curve': "gaussian",
'growth_rate': 0.0000, # (elements/ms)
'continuous': False,
'eta': -0.005,# Ca2+
'eps': 0.075, # Ca2+
}
synaptic_elements_e = {'Den_ex': growth_curve, 'Axon_ex':
growth_curve}
neurons = nest.Create('iaf_psc_alpha', 2,{'synaptic_elements':
synaptic_elements_e})
ps = nest.Create('poisson_generator_ps', params={'rate': 1000.})
nest.Connect(ps, neurons)
nest.Connect(neurons, neurons)
nest.SetKernelStatus({'structural_plasticity_synapses':
{'exsynapse':
{'model': 'static_synapse',
'post_synaptic_element': 'Den_ex',
'pre_synaptic_element': 'Axon_ex',
}}})
# Make SP ON
growth_curve_on = {
'growth_rate': 0.005,
'eta': -0.005,
'eps': 0.03,
}
synaptic_elements_on = {'Den_ex': growth_curve_on, 'Axon_ex':
growth_curve_on}
nest.SetStatus(neurons, 'synaptic_elements_param',
synaptic_elements_on)
nest.Simulate(300)
print("SP ON: {}\n".format(nest.GetStatus(neurons,
'synaptic_elements')))
# Make SP OFF
growth_curve_off = { 'growth_rate': 0.0, }
synaptic_elements_off = {'Den_ex': growth_curve_off, 'Axon_ex':
growth_curve_off}
nest.SetStatus(neurons, 'synaptic_elements_param',
synaptic_elements_off)
nest.Simulate(300)
print("SP OFF: {}\n".format(nest.GetStatus(neurons,
'synaptic_elements')))
**********
On 09.11.21 20:43, Aadhar Sharma wrote:
Hi all,
I want to bring my previous mail to notice. As I mentioned, I wish to
enable/disable structural plasticity at neuron (sub-population) level
rather than at the kernel level. I have tried a bunch of stuff (as
mentioned in the previous email), but nothing seems to really work as
desired.
*I would really appreciate if anyone can suggest potential solutions.*
Many thanks!
Best,
Aadhar
On Wed, 03 Nov 2021 01:20:46 +0100 "Aadhar Sharma" wrote:
Hi all,
I am trying to simulate networks where two (or more) sub-populations
are structurally plastic (structural plasticity module). My
simulation protocol requires that I enable/disable structural
plasticity on a sub-population level rather than at kernel level
(nest.EnableStructuralPlasticity() /
nest.DisableStructuralPlasticity()).
[Q1] Is anyone aware of a way to do this at the sub-population level?
One workaround that I have tried is setting the growth_rate parameter
of synaptic elements to 0 and restoring it to the original value at
the desired simulation time
SP_OFF_dict = {'growth_rate': 0., 'eps': eps}
nest.SetStatus(neurons, {"synaptic_elements":{"Den_exc":
SP_OFF_dict, "Axon_exc":SP_OFF_dict}})
However, this crashes the Kernel with the following error:
terminate called after throwing an instance of
'nest::KernelException' what(): Last update of the calcium
concentration does not match the last update of the synaptic
element.
I don't fully know why the Kernel does not like this SetStatus()
call.
[Q2] Is there something that I may be doing wrong? Can anyone please
recommend a solution?
I am currently using NESTv2.20 for all my simulations on HPC. But, I
am happy to move to NESTv3.1 too.
Many thanks for the support!
Best, Aadhar
_______________________________________________
NEST Users mailing list -- users(a)nest-simulator.org
To unsubscribe send an email to users-leave(a)nest-simulator.org