% function ent=shannon(z,bin,depth)%% calculate the approximate shannon entropy of a time series using a bin% bin encoding and depth symbol symbol sequences. The entropy is% given by %    sum P log P% where P is the probability of any given symbol sequence occuring, % and the summation is over all symbol sequences (of depth symbols% from a bin symbol alphabet).%% The binining is a uniform population binning.%% defualts:% bin = 2% depth = 3%% implementation is shannon.c%% Michael Small% 8/10/04