Neural decoding model

A Python and MATLAB package for decoding neural activity and predicting body position

A regression model that can decode whisker position and locomotion speed from neuronal spike times. There are both Python and MATLAB versions of this code.

Getting started

The folder raw_data contains some sample data to get started. I would suggest running this repository in python, as the codes written in python are mode advanced version and regularly updated.

Running in Python

Two files are required to run the decoder in Python

• runfile.ipynb is a notebook that allows you to run through the whole process of decoding from start to finish.
• pick_n_bins.ipynb is a notebook that allows you to decide how many surrounding bins around the current time t to choose for decoding.
• functions.py contains all the backend functions that are used in the notebook. To adjust parameters for the machine learning model, you might have to tweak through these.

Running in MATLAB

• runfile.m allows you to run through the whole process of decoding from start to finish.
• Rest of the .m files are the functions used in the main file.