psipeter/detailed_neurons

Methods for training encoders, decoders, and filters in Nengo networks containing detailed neuron models implemented in NEURON. Test the performance of these methods on several feedforward and recurrent dynamical systems.

detailed_neurons: build nengo networks with complex neurons

Includes methods for training encoders, decoders, and filters in nengo networks containing detailed neuron models.

Test the performance of these methods on several dynamical systems, including: communication channel, f(x)=x^2, integrator, 2d oscillator, Lorenz attractor

Clone the repository

git clone https://github.com/psipeter/detailed_neurons.git

cd detailed_neurons

Install virtual environment

pip3 install pipenv

pipenv shell

Install NEURON

(follow instructions at https://neuron.yale.edu/neuron/download/compile_linux)

cd detailed_neurons/detailed_neurons/NEURON

module load mpi

/pwd/x86_64/bin/nrnivmodl

(pwd could be /usr/local/nrn/, or an install directory in your home or user folder)

Install detailed_neurons

cd ../..

pipenv --three install '-e .'