sortwvfrms

Sorts signals according to user specified criteria (3 options)

Sorts input matrix of data so that each column is re-ordered by
user-specified criteria. Options are (1) sort data by signal width (2)
sort data by energy localization (3) sort data by dissimilarity of
data to a normal probability distribution. The default option is to
return all 3 options, as vectors of indices so that W(:,Iw), for example,
gives the data ordered by signal width.

[varargout]=sortwvfrms(W,varargin)

INPUT
W: A matrix whose columns are time series to be sorted.
sortopt: (Optional) A string. Can be either 'width', 'erg' or
'pdf'. Default is all three.
plotopt: (Optional) A string. 'plot' produces a plot. Anything
else doesn't.
Inputting 'all' returns a plot of all 3 sorting results.

OUTPUT
I: a vector of indices that sorts the time series, so that W(:,I) has it's
columns ordered by the specified criteria.

Examples:
[Iw]=sortwvfrms(W,'width');
Wsort=W(:,Iw); %now contains the sorted signals.

[Iw,Ie,Ip]=sortwvfrms(W,'all');
%Returns indices that sort the input data by signal width, energy
localization, and 'least noisy'.

[Ip]=sortwvfrms(W,'pdf','plot');
%Now W(:,Ip) is ordered from least noisy to most noisy signal. A plot is
produced too.

NOTE: Code requires plotXmatrix as well. Download from file exchange, or
http://www.ess.washington.edu/~joshuadc