Skip to content

Grid cell model originally developed by Lukas Solanka and extended to include border cells.

Notifications You must be signed in to change notification settings

ikhaliq42/GridCellModel

Repository files navigation

Grid cell repository

This repository holds the source code of the Grid Cell model. It is based partly on Python (brian, numpy, scipy, matplotlib) and/or C++ (NEST simulator).

There are two versions therefore, that can (but it is not recommended) be intermixed:

Brian version
based on the Brian simulator. This is rather obsolete, Brian has become painfully slow with this model and as of current state does not allow for parallelisation of the simulation. It is however very flexible to use, because model descriptions can be easily manipulated by simply changing the differential equations and a few additional settings.
NEST version
Optimized version that implements the grid cell model as a module for the NEST simulator. It is about 10-20 times faster than brian version (theoretically, depending on processor type, thread support, etc.) but a little harder to comprehend when one needs to change single cell properties.

While the idea that a user accesses only a common interface is nice, this was not possible to achieve completely during this project. Therefore currently, both models, and especially the simulation scripts (simulation_*.py), are incompatible.

Installation

For installation information, see the INSTALL.rst file.

Repository content

The repository contains several folders, however the most important one is grid_cell_model. This directory contains the grid cell model and all its necessary components. All other folders are either experiments not very related to the actual model, or side-projects that could be useful in the future.

Cellular_GA
A very simple MPI implementation of a Cellular Genetic algorithm [KU1995]. This version, or the Evolving objects can be later used for more automatic parameter optimization.
Grid_Cells_ModelDB
A version of grid cells that will be prepared to ModelDB. This will be removed in the future.
cuda_test
A simple test of GPU computing. Nothing significant.
data
Data files useful in the simulations. Contain rodent tracking data and preprocessing scripts.
data_analysis
Lots of older MATLAB analysis scripts used in my MSc. thesis. No longer maintained, as all these analysis scripts have been ported to Python.
graphs
Graph theory scripts
grid_cell_model
Grid cell model simulation scripts and source files. This is the main directory. If you just want to work with the grid cell model, you can simply ignore all the other directories.
ideas
Ideas for future modeling that are written down
model_fitting
Scripts for estimating I-V curve data and subsequently fitting integrate and fire models onto the I-V curves. This is no longer maintained actively.
mult_bump_spiking_net
The first, simplified implementation of the spiking bump attractor network, based on [BURAK2009]. This is quite dead (but will be resurrected when I am writing up my thesis - soon)
nest_interface
A proposal. Possible work in progress to design a C++ NEST interface. God knows if it is useful anyhow yet.
oscillations
Some ancient MATLAB scripts dealing with simulation and analysis of gamma oscillations. This was a side-project that was later transferred to the theta-gamma oscillatory attractor model (grid_cell_model).
ramp_model
Another side project that dealt with Hugh Pastoll's ramp gamma model. Due to lack of time, this was stopped.

[KU1995]K. W. C. Ku, M. W. Mak, and W. C. Siu. A cellular genetic algorithm for training recurrent neural networks. In Proceedings of the International Conference on Neural Networks and Signal Processing, pages 140--143, 1995. <http://citeseer.comp.nus.edu.sg/295805.html>
[BURAK2009]Burak, Y., & Fiete, I. R. (2009). Accurate path integration in continuous attractor network models of grid cells. PLoS Computational Biology, 5(2), e1000291. doi:10.1371/journal.pcbi.1000291

About

Grid cell model originally developed by Lukas Solanka and extended to include border cells.

Resources

Stars

Watchers

Forks

Packages

No packages published