mutual_information_naive.m

% MUTUAL_INFORMATION_NAIVE
% A mutual information estimate from samples.
%
% I = mutual_information_naive(S)
% I = mutual_information_naive(S, 'key', value, ...)
%
% where
%
% S is a real (m x n)-matrix that contains n samples of an
% m-dimensional signal.
%
% I is a real (m x m)-matrix where I(i, j) contains a mutual information
% estimate between S(:, i) and S(:, j).
%
% Optional input arguments in 'key'-value pairs:
%
% BINS ('bins') is an integer that determines the number of 
% bins to use for 1d distribution estimation. Default: 100.
%
% Type 'help tim' for more documentation.

% Description: Naive mutual information estimation
% Documentation: mutual_information.txt

function I = mutual_information_naive(S, varargin)

import([tim_package, '.*']);

% Optional input arguments.
bins = 100;
eval(process_options({'bins'}, varargin));

concept_check(nargin, 'inputs', 1);
concept_check(nargout, 'outputs', 0 : 1);

pastelmatlab.concept_check(...
    S, 'real_matrix', ...
    bins, 'integer', ...
    bins, 'positive');

I = tim_matlab('mutual_information_naive', S, bins);