I agree that excessive interpolation might cause problems. Maybe Lomb-Scargle periodogram could be used (spectral analysis on unevenly spaced data). Another option would be to use Kalman filtering to interpolate data. I belive that both are implemented in R.
Milos Zarkovic ----- Original Message ----- From: "Thomas Petzoldt" <[EMAIL PROTECTED]> To: "Francisco J. Zagmutt" <[EMAIL PROTECTED]> Cc: <[email protected]> Sent: Thursday, September 08, 2005 8:17 AM Subject: Re: [R] Interpolating / smoothing missing time series data > Francisco J. Zagmutt wrote: >> I don't have much experience in the subject but it seems that >> library(akima) >> should be useful for your problem. Try library(help="akima") to see a >> list >> of the functions available in the library. >> >> I hope this helps >> >> Francisco > > Yes, function aspline() of package akima is well suited for such things: > no wiggles like in spline() and less variance reducing than approx(). > But in any case: excessive interpolation will definitely lead to biased > results, in particular artificial autocorrelations. > > If ever possible, David should look for methods, capable of dealing with > missing data directly. > > Thomas P. > > ______________________________________________ > [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 ______________________________________________ [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
