Skip to content

Integration between TVB and NEST, for multiscale (co)simulations .

License

Notifications You must be signed in to change notification settings

dionperd/tvb-multiscale-archive

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

TVB multiscale

This package currently provides a solution to interface between TVB and NEST, ANNarchy or NETPYNE (NEURON) spiking networks for multiscale co-simulations.

It tries to offer a generic way to interface between them and the TVB simulator.

Project structure

At the top-level, we have the following folders:

docker

Set of files used for building and distributing the module.

docs

Here is were you can find some documentation about the module. In several forms: pdf or a jupyter notebook with documented steps.

examples

Set of scripts and jupyter notebooks that act as demos on how to use the API with different use-cases.

tests

Unit and integration tests for the module.

tvb_multiscale

This holds the main codebase.

Description of sub-folders:

core

Contains the base code that is considered generic/abstract enough to interface between a spiking network simulator and TVB (inside spiking_models and interfaces).

Here, we also keep I/O related code (read/write from/to H5 format and plots) and some data_analysis related classes.

tvb_nest

Code for interfacing with NEST - depends on core and extends the classes defined there in order to specialize them for NEST (inside nest_models and interfaces).

tvb_annarchy

Code for interfacing with ANNarchy - depends on core and extends the classes defined there in order to specialize them for ANNarchy (inside annarchy_models and interfaces).

tvb_netpyne

Code for interfacing with NETPYNE - depends on core and extends the classes defined there in order to specialize them for NETPYNE (inside netpyne_models and interfaces).

tvb_elephant

Code that interfaces with Elephant for spike train generation and analysis functionality (used also for transformations of interfaces between TVB and spiking simulators), as well as for generating spiking train stimuli for TVB.

tvb_pyspike

Core that interfaces with pyspike package for spike train synchronization analysis.

Acknowledgments

This research has received funding from the European Union’s Horizon 2020 Framework Programme for Research and Innovation under the Specific Grant Agreement Nos. 785907 (Human Brain Project SGA2), 945539 (Human Brain Project SGA3), ICEI 800858, VirtualBrainCloud 826421 and ERC 683049.

About

Integration between TVB and NEST, for multiscale (co)simulations .

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Python 63.8%
  • C++ 30.0%
  • CMake 5.1%
  • Other 1.1%