Adaptive filtering techniques for gravitational wave interferometric
data: Removing long-term sinusoidal disturbances and oscillatory
transients.

E. Chassande-Mottin and S. V. Dhurandhar
gr-qc 0003099
Phys.Rev. D63 (2001) 042004

[IMPORTANT: these codes have been used to produce the figures of
the preprint, not the final paper. I'll put the final codes when
I get a minute to do it]

The codes are distributed under the terms of the GNU General Public 
License (GPL) as published by the Free Software Foundation.

Wed Feb 23 18:57:26 MET 2000
____________________________________________________________________

This package contains all Matlab files necessary to reproduce all
figures included in the preprint. Matlab version 5.xx, the
Time-Frequency Toolbox (v. 2, available for free at
http://www.obs-nice.fr/ecm/tftb) and the Signal
Processing Toolbox (v3.0 or +) are required to be able to run these
scripts. The scripts that are operating on the Caltech 40meter data,
also require the GRASP package (v. 1.9.x, available at
http://www.lsc-group.phys.uwm.edu/~ballen/grasp-distribution/), and
especially its Matlab interface.

Here is the directory list:

adaptsignal.m
adapttest1.m
adapttest1b.m
adapttest2.m
* adpt.m
adptvw.m
gr40mload.m
mafi.m
plchirp2.m
+ rmhist.m
rmmafi.m
rmpspec.m
rmsignal.m
* rmviri.m
rmwaveform.m
welch.m


All files are self-documented (in the header and in the code). The
denoising algorithm consists essentially in the two files indicated
stars (*).

(+) rmhist needs also the Statistics Toolbox (v. 2.1.1).
But the script can be easily change to avoid that.
____________________________________________________________________

Figure 1: LaTeX drawing

Figure 2: Applying the ALE to a sinusoidal signal: approach to locking and steady state 
Matlab command line: adaptest1;  

Figure 3: Applying the ALE to a sinusoidal signal: convergence time
Matlab command line: adaptest1b

Figure 4: Applying the ALE to oscillating transients: testing the
ringdown removal algorithm
Matlab command line: adaptsignal;adapttest2;

Figure 5: Illustration of the denoising procedure on Caltech
proto-type data
Matlab command line: rmsignal;rmviri;
(flags setting in rmviri.m: graph=1;printok=1;)

The next four figures require to run the following commands first.
Matlab command line: rmsignal;rmviri;
Three files (step0.mat, step1.mat and step2.mat) containing the result
of the denoising algorithm will be created during the execution, and
used in the following.

Figure 6: ``Caltech signal only'': comparison between power spectra of
ALE input/output signals
Matlab command line: rmpspec

Figure 7: ``Caltech signal only'': comparison between histograms of
ALE input/output signals
Matlab command line: rmhist

Figure 8:``Caltech+inspiral'' signal: matched filter response before
and after denoising
Matlab command line: rmmafi

Figure 9:``Caltech+inspiral'' signal: close view after denoising
Matlab command line: rmwaveform