Code for the simulations in Simon N Weber & Henning Sprekeler: 'Learning place cells, grid cells and invariances with excitatory and inhibitory plasticity'. eLife 2018.

Please cite the paper if you use any part of this package.

The package is made available under the GNU General Public License v3.0.

Copyright 2018 Simon N. Weber

Clone this repository using

git clone --recursive https://github.com/sim-web/spatial_patterns

The recursive flag is necessary to download the content of the git submodules general_utils and gridscore.

For all the imports to work, make sure that the parent directory of spatial_patterns is part of your PYTHONPATH. If it isn't, use the following command in your terminal to add it:

export PYTHONPATH="${PYTHONPATH}:<path to parent directory of spatial_patterns>"

Example: So say you cloned spatial_patterns to /Users/joe/workspace/spatial_patterns, then use

export PYTHONPATH="${PYTHONPATH}:/Users/joe/workspace"

If you want /Users/joe/workspace to pe permantly part of your python path, add the export line above to your .bashrc (Linux) or .bash_profile (Mac).

You find example experiments in experiment.py.

Go there and select a predefined set of parameters, e.g., parameters.params_test_2d for a quick test of 2 dimensional simulations.

Run it.

A plot of the time evolution of the grid pattern should appear.

You find the predefined parameter sets in parameters.py. You can also define your own set of parameters there. The given parameter sets are identical to those used in the paper. See the parameter tables in the paper to find more interesting parameter combinations.

Contact me over GitHub if you encounter problems.

• Python 3, NumPy, SciPy, matplotlib