Search found 18 matches

by Tom Close
Mon Mar 16, 2015 9:10 pm
Forum: Adding new mechanisms and functions to NEURON
Topic: Rounding error for function in derivative block
Replies: 6
Views: 8481

Re: Rounding error for function in derivative block

Hi Michael, Thanks, I switched to derivimplicit and the results are the same now. I think it would be really helpful for nrnivmodl to spit out warnings if the integration method is not applicable/optimal for the provided mechanism. I will create a ticket so that this also gets updated in the PyNN ve...
by Tom Close
Fri Mar 13, 2015 7:56 pm
Forum: Adding new mechanisms and functions to NEURON
Topic: Rounding error for function in derivative block
Replies: 6
Views: 8481

Rounding error for function in derivative block

I appear to be getting a rounding error when encapsulating a time derivative in the derivative block within a function, which I don't understand. Using the PyNN izhikevich mod file as an example, could you explain to me why : Izhikevich artificial neuron model from : EM Izhikevich "Simple Model of S...
by Tom Close
Fri Mar 13, 2015 7:47 pm
Forum: NEURON + Python
Topic: MWE for apparent bug discovered by PyNN format
Replies: 4
Views: 1200

Re: MWE for apparent bug discovered by PyNN format

Thanks Michael, I will let Andrew know so that he can rename 'a_' back to 'a' in the next release of PyNN :)
by Tom Close
Wed Mar 11, 2015 12:18 am
Forum: NEURON + Python
Topic: MWE for apparent bug discovered by PyNN format
Replies: 4
Views: 1200

Re: MWE for apparent bug discovered by PyNN format

renaming the parameters in the mod file doesn't seem to help. Are 'a', 'b', 'c', 'd' hard-coded somewhere?
by Tom Close
Tue Mar 10, 2015 10:46 pm
Forum: NEURON + Python
Topic: MWE for apparent bug discovered by PyNN format
Replies: 4
Views: 1200

Re: MWE for apparent bug discovered by PyNN format

Sorry, that should have been

Code: Select all

NameError: a, the mechanism does not exist at PySec_0x*(0.5)
(I reproduced this behaviour for both 'a' and 'b' and I assume it would occur for all parameters used in the izhikevich model)
by Tom Close
Tue Mar 10, 2015 10:44 pm
Forum: NEURON + Python
Topic: MWE for apparent bug discovered by PyNN format
Replies: 4
Views: 1200

MWE for apparent bug discovered by PyNN format

Hi, I am getting a strange bug when using the built-in PyNN Izhikevich model (outside of PyNN), which I assume is caused by a buffer overrun somewhere in the NEURON Python interface. The following snippet is a MWE import neuron import pylab as plt from neuron import h, nrn, load_mechanisms from math...
by Tom Close
Thu Jul 17, 2014 12:09 am
Forum: Getting started
Topic: psection-like dump of global parameters
Replies: 9
Views: 2450

Re: psection-like dump of global parameters

Thanks Patoorio, Ted, Michael, this method works nicely. However, is there a straightforward way to extend it to also print out the names of the secctions in each of the the ModelViewParmSubsets?
by Tom Close
Sun Jun 29, 2014 5:35 am
Forum: NEURON + Python
Topic: Playing vector into SEClamp
Replies: 4
Views: 1977

Re: Playing vector into SEClamp

I'm pretty sure the key here is to set the duration of the SEClamp (as in Ted's example) to something very long such as 1e9 and then it should work.
by Tom Close
Fri Feb 08, 2013 8:01 pm
Forum: Parallel NEURON
Topic: Minimal Working Example of MPI Gap Junctions (in python)
Replies: 15
Views: 5441

Re: Minimal Working Example of MPI Gap Junctions (in python)

Thank Michael, I will send you an email with the code, but actually you can reproduce the problem on 3 processes with the code I have posted above (adding in the dummy_section to keep the record_t function happy on the middle and otherwise unused process).
by Tom Close
Fri Feb 08, 2013 11:04 am
Forum: Parallel NEURON
Topic: Minimal Working Example of MPI Gap Junctions (in python)
Replies: 15
Views: 5441

Re: Minimal Working Example of MPI Gap Junctions (in python)

I think it might be something to do with that because I was able to replicate the error:

nrn_timeout t=0.125

with my MWE with > 2 processes (<= 2 processes works fine).

(I also had to add "dummy_section = h.Section()" somewhere near the top to keep the rest of the code happy).
by Tom Close
Fri Feb 08, 2013 10:42 am
Forum: Parallel NEURON
Topic: Minimal Working Example of MPI Gap Junctions (in python)
Replies: 15
Views: 5441

Re: Minimal Working Example of MPI Gap Junctions (in python)

Okay, that makes sense. I double checked and I am calling "setup_transfer()" on all processes. It doesn't make any assumptions about there being source_var and target_var connections being present on every process does it? Because I seem to get this timeout error when there is at least one node that...
by Tom Close
Fri Feb 08, 2013 6:43 am
Forum: Parallel NEURON
Topic: Minimal Working Example of MPI Gap Junctions (in python)
Replies: 15
Views: 5441

Re: Minimal Working Example of MPI Gap Junctions (in python)

Thanks for the tips. I am actually writing a gap junction connector for pyNN so if I understand what you mean correctly the gap junctions will be independent of the number of processes (is there any specific reason why it is useful by the way?). I have now progressed from my simple "minimal working ...
by Tom Close
Wed Jan 30, 2013 1:18 am
Forum: Parallel NEURON
Topic: Minimal Working Example of MPI Gap Junctions (in python)
Replies: 15
Views: 5441

Re: Minimal Working Example of MPI Gap Junctions (in python)

I take your point that only providing one side of the gap junction is not equivalent to a rectifying electrical synapse (which having had a think about it, would require the conductance to vary with the voltage difference). However, whether or not it is a realistic representation of a real gap junct...
by Tom Close
Tue Jan 29, 2013 8:07 pm
Forum: Parallel NEURON
Topic: Minimal Working Example of MPI Gap Junctions (in python)
Replies: 15
Views: 5441

Re: Minimal Working Example of MPI Gap Junctions (in python)

Okay, so we found out that if you insert a second gap junction mechanism (going back from section2 to section1) to form a bi-directional gap junction, the problem with the nrn_timeout goes away. So maybe using MPI introduces some sort of instability that isn't a problem when the gap junctions are bi...
by Tom Close
Tue Jan 29, 2013 3:07 am
Forum: Parallel NEURON
Topic: Minimal Working Example of MPI Gap Junctions (in python)
Replies: 15
Views: 5441

Re: Minimal Working Example of MPI Gap Junctions (in python)

Thanks for your reply Ted. 1. Yes 2. Okay, that was silly. So now I am using the same as Lazarewicz's Gap NMODL code, with the slight exception that 'vgap' is declared as a RANGE variable instead of a POINTER (as per Hines' instructions at http://www.neuron.yale.edu/phpBB/viewtopic.php?f=31&t=2743 )...