Recoded wmtsa routines
Joel Hasbrouck
July 5, 2013

wmtsa (Cornish, C., 2006. The WMTSA wavelet toolkit for MATLAB Department of Atmospheric Sciences, University of Washington, Seattle.)
is a collection of matlab routines that does wavelet computations described in:
Percival, D.B., Walden, A.T., 2000. Wavelet Methods for Time Series Analysis. Cambridge University Press, Cambridge;
(Also see the web notes for 
Percival, D.B., 2012. Class Materials: Wavelets: Data Analysis, Algorithms and Theory (Stat/EE 530). University of Washington.)

In the wmtsa distribution, there are two functions (modwtj and imodwtj) which are written in c. These need to be compiled into MEX files.
In order to avoid that step, I've recoded them in Matlab. (I use matlab's circular shift functions to keep looping to a minimum,
so they should be relatively efficient.) The new routines are contained in the wmtsa/hasbrouck/hasbrouckFunctions

The function waveTest.m in /hasbrouck is a short driver program to illustrate the modwt and inverse modwt transforms
on some simulated data. There is also a small function (waveletCovariance.m) that calculates wavelet covariances.

The files in the subdirectory /wmtsa-matlab-0.2.6 are copies of the current wmtsa distribution. I'm including them in the zip file
for convenience.

The wmtsa notation and data structures correspond to the usage in P&W's WMTSA book.
I believe that Mathematica uses the same conventions in its wavelet routines.
(Transforms done under wmtsa and Mathematica agree.)

The conventions used in Matlab's own wavelet package, though, differ from wmtsa/Mathematica.
I haven't been able to work out a simple correspondence.
