Skip to content

Rate model implementations for (adaptive) integrate-and-fire neurons based on the Fokker-Planck equation: (i) numerical (finite volume) solution of the full FP PDE, (ii) low-dim. ODE via spectral decomposition of the FP operator, (iii) low-dim. ODE via a linear-nonlinear cascade semianalytically fit to the FP model.

License

Notifications You must be signed in to change notification settings

nigroup/fokker-planck-based-spike-rate-models

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

50 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

fokker-planck-based-spike-rate-models

Implementations of spike rate models derived from networks of adaptive exponential integrate-and-fire models, as described in: Augustin*, Ladenbauer*, Baumann, Obermayer, Low-dimensional spike rate models derived from networks of adaptive integrate-and-fire neurons: comparison and implementation, PLOS Computational Biology 2017

  1. Numerical solution of the mean-field Fokker-Planck (FP) equation using a finite volume method with Scharfetter-Gummel flux approximation.
  2. Low-dimensional ordinary differential equations (ODE) derived from the spectral decomposition of the FP operator.
  3. Low-dimensional ODE based on a cascade of linear filters and a nonlinearity determined from the FP equation.

Furthermore: precalculation codes for the (look-up) quantities involved in (2) and (3).

For questions please contact us: Moritz Augustin and Josef Ladenbauer

Code Usage: change to folder adex_comparison and see the README.md file contained therein

About

Rate model implementations for (adaptive) integrate-and-fire neurons based on the Fokker-Planck equation: (i) numerical (finite volume) solution of the full FP PDE, (ii) low-dim. ODE via spectral decomposition of the FP operator, (iii) low-dim. ODE via a linear-nonlinear cascade semianalytically fit to the FP model.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%