MATLAB code for implementing the Population Tracking neural data analysis method.
The model is explained in this paper: http://biorxiv.org/content/early/2016/07/19/064717
No setup required. The following function files are included:
fitPopTrack : fit the parameters of the Population Tracking model to binary neural population data. get_jsdiv : estimate the Jensen-Shannon divergence between two fitted distributions given the estimated population tracking model parameters. get_kldiv : estimate the Kullback-Leibler divergence between two fitted distributions given the estimated population tracking model parameters. samplePopTrack: generate samples from the fitted model, given the parameters. compute_ak: calculate the set of normalization constants, needed for exact pattern probabilities. compute_px: calculate the predicted probability of a specific neural activity pattern, given the model parameters. stir_binom: computes binomial coefficient using Stirling's appoximation (useful for large n, k).
See the help functions of each file for more details.