[R] Filtering the time series

Spencer Graves spencer.graves at pdf.com
Thu Feb 2 05:16:57 CET 2006

	  The Nyquist frequency is shockingly easy:  Google just produced for 
me about 766,000 hits.  The first two I checked said the Nyquist 
frequency is half the sampling rate (e.g., 
http://en.wikipedia.org/wiki/Nyquist_frequency).  Therefore, with daily 
numbers, the Nyquist frequency is 0.5 days, which means you can detect 
patterns that occur on alternate days.

	  If it were my problem, I would spend my time trying to correlate any 
cycles with known physical phenomena, like the annual weather patterns 
and the phases of the moon.  I hope some spectral wizard will enlighten 
me, but the primary value I've gotten from things like spectral analysis 
is to push me and others to think of other phenomena that occur at the 
special identified frequencies.

	  This logic pushes me to ask why you think there might be oscillations 
with periods of 10-20 days?

	  Beyond this, I suggest you consider the functions arima and 
KalmanLike in the Stats package, the time series chapter in Venables and 
Ripley (2002) Modern Applied Statistics with S, 4th ed. (Springer).  If 
you do not have reasonable access to this latter book, I encourage you 
to tell your management that it comes most highly recommended, and that 
the purchase of a copy will be an exceptionally good investment, in my 

	  Hope this helps.
	  spencer graves

anil kumar rohilla wrote:

> Hi List,
>     I have a time series of 122 values, actualy it is 
a time series of daily indian monsoon rainfall. now i want to
filter this time series for a particular oscilation say 10 to
20days oscilation. i want to find out what amount of variance
is explained by this mode. Which package is available in R for
this purpose. and how to calculate nquest frequancy of this
series. any help is much appriciated.
> thanks in advance
> anil 
> PUNE-411005 INDIA
> MOBILE +919422023277
> anilkumar at imdpune.gov.in
> 	[[alternative HTML version deleted]]
> ______________________________________________
> R-help at 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

More information about the R-help mailing list