I deal with mono-channel breathing signals sampled at 30[Hz] which are
non-linear and non-stationary.
My goal is to classify the signals according to common breathing patterns
Trend remotion is necessary for cluster analysis but quite challenging. In
fact, quasi-periodic patterns that span a number of
consecutive breathing cycles should not be naively removed as they carry a lot
of useful information.
In advance of applying any trend removing technique, I have to extract the main
period and save it.
I thought it would be an easy task through functions "spec" and "specprop".
But I cannot get them to work for me.
Here is my simple exercise:
> length(amp)
[1] 3885
> amp.sp <- spec(amp,f=0.033)
> specprop(amp.sp,f=0.033,str=TRUE,plot=1)
Error in specprop(amp.sp, f = 0.033, str = TRUE, plot = 1) :
Frequency resolution is to high (<0.5 hz)
> specprop(amp.sp,f=0.033,str=TRUE)
Error in specprop(amp.sp, f = 0.033, str = TRUE) :
Frequency resolution is to high (<0.5 hz)
I cannot get the meaning of the message error unless there is a constraint on
the frequency resolution deal with
by these functions ... in this case the error message should read "...
Frequency resolution is TOO high (<0.5 hz)"
Even so, I am still puzzled beause the default window is 512 wide ==> 1/512 =
0.001953125
The signal length is 3885 ==> 1/3885 = 0.0002574003
Your help is very welcome.
Thank you in advance.
Kind regards,
Maura
tutti i telefonini TIM!
tutti i telefonini TIM!
[[alternative HTML version deleted]]
______________________________________________
[email protected] 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.
