In the case of muitivariate, from the documentation it looks like I can compare more than two signals at a time. Each column of the input matix seem to accommodate a signal. The problem is that my signals do NOT have the same number of samples (length). They were all collected at 30Hz so the sampling time interval is roughly 0.033[s]. Some signals have about 5000 samples and other ones have more than 8000. The R routine "spectrum" expect the multivariate to be a matrix ... Any idea how to overcome such an obstacle ? Padding with zeros would alter (I think) the phenomen being studied that is breathing patterns. Is there a way to feed the "spectrum" function with the signal spectrum (power density) instead of the time domain signal ? Since the sampling interval is equal for all the signal, so is the Nyquist frequency. I can easily get the power spectrum defined over the domain [0, Nyquist-frequency] which does not have the problem of different lengths ... ???
Thank you so much. Maura On Wed, Apr 30, 2008 at 8:56 AM, stephen sefick <[EMAIL PROTECTED]> wrote: > $names > [1] "freq" "spec" "coh" "phase" "kernel" "df" > [7] "bandwidth" "n.used" "orig.n" "series" "snames" "method" > [13] "taper" "pad" "detrend" "demean" > > $freq and $spec are used to plot the power spectrum. freq is the x-axis > and spec is the y-axis. $coh is the squared coherency between the two > signals in your case and I believe that this is also plotted against > frequency. This is your "correlation" strength. Phase I haven't been able > to figure out- I think that it is some sort of estimator for the phase > shift. to get either phase or coherency plot add the plot.type argument to > your plot command > > x <- spectrum(yourdata, log="no") #this will plot it without a log scale I > find it useful to look at both the no log plot and then the logscale plot > (just remove the log="no") > > plot(x, plot.type="marginal") #this is the default type (the > powerspectrum) > plot(x, plot.type="phase") > plot(x, plot.type="coherency") > > also just look at > > ?spectrum > schumway is a good book - I think it is something like time series > analysis with examples in R > > hope this helps > > stephen > > > On Tue, Apr 29, 2008 at 8:54 PM, Maura E Monville < > [EMAIL PROTECTED]> wrote: > > > I am reading some documentation about Cross Spectrum Analysis as a > > technique > > to compare spectra. > > My understanding is that it estimates the correlation strength between > > quasi-periodic structures embedded in two signals. I believe it may be > > useful for my signals analysis. > > > > I was referred to the R functions that implement this type of > > analysis. I > > tried all the examples which generated a series of fancy plots. But I > > need > > to work on the numerical results. > > > > I have read that the following info is available through Cross Spectra > > analysis: > > *Cross-periodogram, Cross-Density, Quadrature-density, Cross-amplitude, > > Squared > > Coherency, Gain, and Phase Shift* > > I went through a couple of the two-series (bivariate) cross-spectrum > > analysis examples with R. > > I also printed out the attributes of the analysis (see the following). I > > cannot quite match the above quantities with the attributes/features > > output > > of cross-spectra analysis with R. > > I would greatly appreciate some explanation (which is what) and seeing > > some > > more worked out examples. > > > > > attributes(mfdeaths.spc) > > $names > > [1] "freq" "spec" "coh" "phase" "kernel" "df" > > [7] "bandwidth" "n.used" "orig.n" "series" "snames" > > "method" > > [13] "taper" "pad" "detrend" "demean" > > > > $class > > [1] "spec" > > > > > > Thank you so much. > > > > Yours Faithfully, > > -- > > Maura E.M > > > > [[alternative HTML version deleted]] > > > > ______________________________________________ > > R-help@r-project.org mailing list > > https://stat.ethz.ch/mailman/listinfo/r-help > > PLEASE do read the posting guide > > http://www.R-project.org/posting-guide.html > > and provide commented, minimal, self-contained, reproducible code. > > > > > > -- > Let's not spend our time and resources thinking about things that are so > little or so large that all they really do for us is puff us up and make us > feel like gods. We are mammals, and have not exhausted the annoying little > problems of being mammals. > > -K. Mullis -- Maura E.M [[alternative HTML version deleted]] ______________________________________________ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.