Research Group Leader

Prof. Dr. Matthias Bethge
Prof. Dr. Matthias Bethge
Phone: +49 7071 29-89017
Fax: +49 7071 29-25015


Secretary: Heike König
Phone: +49 7071 29-89018
Fax: +49 7071 29-25015



A toolbox for circular or direction statistics, containg a wide selection of descriptive techniques for such data, as well as inferential methods ranging from simple tests for uniformity to complex to factor ANOVA like tests. It has been descriped in the paper

From the journal website, a tutorial can be downloaded as well.

A toolbox to efficiently sample from (1) correlated multivariate binary random variables (multivariate Bernoulli), (2) correlated multivariate Poisson random variables and (3) correlated random variables with arbitrary marginal statistics. Applications include modeling and generating of artificial neural data. It is based on the methods i

The file also contains a tutorial to reproduce most figures from the paper.
NISDET - A Natural Image Statistics Density Estimation Toolbox

The toolbox includes code for fitting Lp-spherically symmetric distributions as described in

It also includes code for fitting the quantitative ICA (QICA) as described in
Furthermore, it provides code for Lp-nested symmetric distributions as described in
In addition, it implements several whitening routines and allows for optimizing filter on the log-likelihood of the aforementioned models. A quick tutorial is provided explaining the usage of the toolbox.

NATTER - A Natural Image Statistics Density Estimation Toolbox for Python

The toolbox is an advancement of the NISDET toolbox. It includes more distributions, more functionality. In particular, it includes more functionality for Lp-nested symmetric distributions. It also includes the algorithms for radial factorization and nested radial factorization.

We provide a complete documentation of the functions here.

You can clone the lastest version of the NATTER from

Last updated: Wednesday, 27.02.2013