frites 0.3.2

Framework of Information Theory for Electrophysiological data and Statistics

Description

FRITES = Framework for Information Theoretical analysis of Electrophysiological data and Statistics

Frites is a python package for analyzing neurophysiological brain data (i.e M/EEG, sEEG / iEEG / ECoG). The package is entirely based on information theoretic measures (such as mutual information (MI)) in order to perform analysis such as :

• “Correlation like” (I(c; c) = MI between two continuous variables)

• “Machine-learning like” (I(c; d) = MI between a continuous and a discrete variable)

• “Partial correlation like” (I(c; c | d) = MI between two continuous variables and removing the influence of a discrete one)

• Information-transfer about a specific feature

For a comprehensive (and extensive) review, see the paper of Robin AA Ince A statistical framework for neuroimaging data analysis based on mutual information estimated via a gaussian copula.

Frites also comes with embedded statistics which support fixed and random-effect analysis in combination with inferences either at the single time-point level or at the temporal cluster level.

Take a look at the online documentation and examples to start analyzing your data with Frites : https://brainets.github.io/frites/

Installation

Dependencies

The main dependencies of Frites are :

• Numpy

• Scipy

• MNE

• Joblib

In addition to the main dependencies, here’s the list of additional packages that you might need :

• Pandas and Xarray : additional output types

• Numba : speed computations of some functions

User installation

Frites can be installed (and/or updated) via pip with the following command :

pip install -U frites

Developer installation

For developers, you can install frites in develop mode with the following commands :

git clone https://github.com/brainets/frites.git
cd frites
python setup.py develop