[R] Stuck with FFT

Rthoughts seb.brockel at northampton.ac.uk
Tue Nov 18 10:42:39 CET 2008



Hi,

I have more questions about the fft. The application in Excel is very
limited.

In Excel I can adjust graphs and calibrate the x and y-axis. The input and
process, however, is limited compared to R.

With a Dataset table where one column is the hour difference and the second
are the values with hidden frequencies, how is it possible to plot a
frequency spectrum with the x-axis calibrated to the hour difference where
the left should be infinite frequency and half way to the right should be
the midpoint.

Below is an example dataset with a 24 hour sinusoid that I would like to fft
for assisting with my ability to interpret my output. How do I calibrate the
plot so that the x-axis displays the cycles per that run and the y-axis is
frequency? In the command line, (Dataset$X10, inverse = FALSE, frequency =
24), what does the frequency line specify and and it is related to number of
data points? By doing this am I generating frequencies rather than revealing
them?

Secondly, how to do generate a spectrogram on R? I believe that this one is
where the x-axis is the cycles per data point with the left being infinite
cycles to 0.5 being one cylce per 2 data points and the y-axis is the
amplitude.

Thank you very much in advance if you can help and share your experience/
knowledge. I am progressing with fourier transforms in my research but I am
new to it so that doesn't help.

Thirdly, I came across in my reading that it is a good idea to generate a
correlogram to determine if it is valid to do an fft on the dataset. How
would I go about doing this and how is frequency/ lag determined?

00/01 00:00	4.330
00/01 01:00	3.536
00/01 02:00	2.500
00/01 03:00	1.294
00/01 04:00	0.000
00/01 05:00	-1.294
00/01 06:00	-2.500
00/01 07:00	-3.536
00/01 08:00	-4.330
00/01 09:00	-4.830
00/01 10:00	-5.000
00/01 11:00	-4.830
00/01 12:00	-4.330
00/01 13:00	-3.536
00/01 14:00	-2.500
00/01 15:00	-1.294
00/01 16:00	0.000
00/01 17:00	1.294
00/01 18:00	2.500
00/01 19:00	3.536
00/01 20:00	4.330
00/01 21:00	4.830
00/01 22:00	5.000
00/01 23:00	4.830
01/01 00:00	4.330
01/01 01:00	3.536
01/01 02:00	2.500
01/01 03:00	1.294
01/01 04:00	0.000
01/01 05:00	-1.294
01/01 06:00	-2.500
01/01 07:00	-3.536
01/01 08:00	-4.330
01/01 09:00	-4.830
01/01 10:00	-5.000
01/01 11:00	-4.830
01/01 12:00	-4.330
01/01 13:00	-3.536
01/01 14:00	-2.500
01/01 15:00	-1.294
01/01 16:00	0.000
01/01 17:00	1.294
01/01 18:00	2.500
01/01 19:00	3.536
01/01 20:00	4.330
01/01 21:00	4.830
01/01 22:00	5.000
01/01 23:00	4.830
02/01 00:00	4.330
02/01 01:00	3.536
02/01 02:00	2.500
02/01 03:00	1.294
02/01 04:00	0.000
02/01 05:00	-1.294
02/01 06:00	-2.500
02/01 07:00	-3.536
02/01 08:00	-4.330
02/01 09:00	-4.830
02/01 10:00	-5.000
02/01 11:00	-4.830
02/01 12:00	-4.330
02/01 13:00	-3.536
02/01 14:00	-2.500
02/01 15:00	-1.294
02/01 16:00	0.000
02/01 17:00	1.294
02/01 18:00	2.500
02/01 19:00	3.536
02/01 20:00	4.330
02/01 21:00	4.830
02/01 22:00	5.000
02/01 23:00	4.830
03/01 00:00	4.330
03/01 01:00	3.536
03/01 02:00	2.500
03/01 03:00	1.294
03/01 04:00	0.000
03/01 05:00	-1.294
03/01 06:00	-2.500
03/01 07:00	-3.536
03/01 08:00	-4.330
03/01 09:00	-4.830
03/01 10:00	-5.000
03/01 11:00	-4.830
03/01 12:00	-4.330
03/01 13:00	-3.536
03/01 14:00	-2.500
03/01 15:00	-1.294
03/01 16:00	0.000
03/01 17:00	1.294
03/01 18:00	2.500
03/01 19:00	3.536
03/01 20:00	4.330
03/01 21:00	4.830
03/01 22:00	5.000
03/01 23:00	4.830
-- 
View this message in context: http://www.nabble.com/Stuck-with-FFT-tp20192942p20556317.html
Sent from the R help mailing list archive at Nabble.com.



More information about the R-help mailing list