Wang, Buzsaki 1996's model running issue

[breakwave922]

Hi,
I downloaded the model file--Gamma oscillations in hippocampal interneuron networks (Wang, Buzsaki 1996), from modelDB. Then I compiled it and tried to run it on my local computer. When I click the jnsci16_6402F3.hoc file, it kept giving me error, saying intsall_vecst undefined function.
Do anyone have experience on running this model?
Here is the link for downloading the model http://senselab.med.yale.edu/ModelDB/Sh ... =\wang1996

Any help would be appreciated.

[tom_morse]

Start with the mosinit.hoc file instead if that is OK.

I will help you get the model running if the mosinit.hoc doesn't have the part you would like. 1)Which platform (mswin, mac os x, unix/linux) and which 2) version of NEURON (7.3, alpha, hg) are you using?

[breakwave922]

Thank you very much, Tom.
I started with the mosinit.hoc file. Yes, it works. It gave the GUI to ask for choosing Fig 1A or Fig 3A.
I use Windows 8 and Neuron version 7.3

[breakwave922]

Start with the mosinit.hoc file instead if that is OK.

I will help you get the model running if the mosinit.hoc doesn't have the part you would like. 1)Which platform (mswin, mac os x, unix/linux) and which 2) version of NEURON (7.3, alpha, hg) are you using?

Hi, Tom,
I just realized that I couldn't even compile it. When I compile it in NEURON7.3, it gave me error:"vecst.c : 162:23: fatal error: sys/time.h: No such file or directory".
Sorry for ignoring this important message before.

[ted]

Ah, the perils of assuming specific paths for key files. On whatever machine that Bill Lytton used, time.h lived in a particular location. Your task, if you want to reuse his code, is to locate where time.h lives on your own machine (if you're using MSWindows, you'll have to grope around in cygwin's directory structure, or maybe someone else knows where to point you). Then you'll want to identify the mod files that include time.h and revise those include statements to point to the correct location.

[tom_morse]

Changing the include file from sys/time.h to just time.h got past this mknrndll error. However there were other errors with the
drand48 not being found.
I found another model in ModelDB that has some clues about how to convert this to windows from linux:
https://senselab.med.yale.edu/modeldb/S ... 7\rand.mod
I will work on this and get back to you soon.

[tom_morse]

I updated the model in modeldb to run on mswin as well as unix/linux. Please download and try this new version:

http://senselab.med.yale.edu/modeldb/Sh ... odel=26997

[breakwave922]

Hi, Tom,
Yes. It works now. Thank you so much.

[RBett]

Hi all,

I've come accross the same problem in the Pinsky-Rinzel reduced Traub model (1994)

https://senselab.med.yale.edu/modeldb/S ... 358#tabs-1

I used the fix that Tom suggested for sys/time.h to time.h but have got stuck with drand48. The relevant mod file is vecst. Any ideas?


(running version 7.4x86_64 on 64-bit windows 7 if you're curious)

[RBett]

but have got stuck with drand48.

...got past it. Followed directly the modifications that have been made to the Buzsaki Wang model and now Pinsky Rinzel seems to be running fine.

Is it worth updating the version on modelDB do you think?

[tom_morse]

Yes it is worth putting in ModelDB. Please email me the code (will send separate email to you) and I will upload it!

[tom_morse]

The model in modeldb has been updated with a vecst.mod file RBett sent. Thanks RBett!