[R] Fwd: Re: Wavelet reconstruction

From: Elizabeth Lawson <lizzylaws_at_yahoo.com>
Date: Wed 14 Dec 2005 - 06:23:05 EST


  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]]



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 Received on Wed Dec 14 08:11:27 2005

This archive was generated by hypermail 2.1.8 : Wed 14 Dec 2005 - 09:34:31 EST