I realized that I may not have answered the question you were asking and that no
one else has responded. I can across a similar problem and may have an answer
to your question now. If you have both the wavelet coefficients and the scaling
coefficients then create a fake sequence of the same length as the original and
decompose that sequence using wd form wavethersh with the same wavelet family
and filter that was used to decompose the data. Then you can replace the
wavelet coefficients and the scaling coefficients using putC and putD from
wavethresh. This will leave you with a wd object that you can then reconstruct
using wr from wavethresh. I hope this works and is still useful to you.
Elizabeth Lawson
Elizabeth Lawson <lizzylaws@yahoo.com> wrote:
Date: Wed, 19 Oct 2005 08:04:02 -0700 (PDT)
From: Elizabeth Lawson <lizzylaws@yahoo.com>
Subject: Re: [R] Wavelet reconstruction
To: Amir Safari <amsa36060@yahoo.com>
Using wavethresh you can recompose a decompose signal using wr.
Here is an example of decomposing, thresholding and recomposing a signal.
library(wavethresh)
brain<-c(0,0,0.5,1,-0.75,-0.25,1.833333333,-3,0.416666667,1.083333333,-1.833333333,
-0.583333333,2.166666667,-4.083333333,5.75,9.75,1.583333333,0.75,15.83333333,
16.66666667,7.666666667,8.166666667,1.333333333,-3.333333333,-2.75,-2.083333333,
-1.75,0.416666667,1.25,8.583333333,-4.583333333,0.666666667,-6.416666667,3.583333333,
3.416666667,-3.333333333,-7.25,-1.833333333,-1.5,-0.083333333,-2.333333333,7.75,5,
-2.333333333,12,10.5,-1.333333333,-3.333333333,-3.416666667,14.08333333,5.166666667,
5.166666667,2.25,-0.083333333,-1.25,-1,0.083333333,1.666666667,1,-1.333333333,0.416666667,
-0.166666667,-0.25,-0.166666667)
x<-(
c(1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,
27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,
52,53,54,55,56,57,58,59,60,61,62,63,64))
x<-x/64
par(mfrow=c(1,1))
plot(x,brain,xlab="Voxel",ylab="Activity",main="fMRI
Data")
wdbrain<-wd(brain,4,family="DaubExPhase",
bc="periodic")
thres2<-threshold(wdbrain,levels=3:(wdbrain$nlevels-1),
type="soft",
policy="manual", by.level=FALSE, value=7.32032, dev=var,
boundary=FALSE,
verbose = getOption("verbose"), return.threshold=F)
thr2 <- wr(thres2)
plot(x,brain, col =
"slateblue",xlab="Voxel",ylab="Activity",main="Wavelet
Regression")
mtext("N=4, Threshold=7.32032")
lines(x, thr2, col= "violetred" , lwd=2,type="l")
Good Luck!!
Elizabeth Lawson
Amir Safari <amsa36060@yahoo.com> wrote:
Hi There, I tried to find a function in {waveslim} or {wavethresh} in order to
reconstruct the decomposed signals. As far as I found there is no function in
{waveslim} to reconstruct decomposed data. The function wr{wavethresh}
reconstructs the results of wd function. Apart from its limitations ( for
example the length of vector must be power of 2 ) it apparently doesn't work
with the functions and objects in waveslim.
What could help ?
So many thanks for your any idea.
Amir Safari
---------------------------------
[[alternative HTML version deleted]]
______________________________________________
R-help@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
---------------------------------
---------------------------------
[[alternative HTML version deleted]]
