neuro-kl contains a Python module and Matlab functions to compute an estimate of the entropy and Kullback-Leibler divergence of distribution of simultaneously recorded neural data.
The KL is estimated using a Bayesian method designed to deal with relatively large distributions (2^16 elements), and is described in the supplementary material of the paper
Berkes, P., Orban, G., Lengyel, M., and Fiser, J. (2011). Spontaneous cortical activity reveals hallmarks of an optimal internal model of the environment. Science, 331:6013, 83–87.
in the "Dissimilarity of neural activity distributions" section. The supplementary material is freely available at http://www.sciencemag.org/content/331/6013/83/suppl/DC1 .
neuro-kl is released under the GPL v3. See LICENSE.txt .
Copyright (c) 2011, Pietro Berkes, Dmitriy Lisitsyn. All rights reserved.