gaussbandpass: Gaussian kernel bandpass and smoothing
In krehfeld/nest: Non-equally spaced time series analysis

Description Usage Arguments Value Warning Author(s) References See Also Examples

Bandpassing, smoothing and detrending of irregular time series using a gaussian kernel smoother.

1 2	gaussbandpass(X, per1, per2,prune=FALSE) gaussdetr(X, tsc.in = mean(diff(index(X)))*10,prune=FALSE)

`X`	Input time series (`zoo`-object)
`per1`	Timescale 1 for lowpass
`per2`	Timescale 2 for highpass
`prune`	Logical; prune `(per1/mean(diff(index(X))))` points at the edges of the time series
`tsc.in`	Timescale for detrending in `gaussdetr`

A list consisting of

`trend`	Trend (used for highpass)
`smoothed`	Smoothed time series, used for lowpass
`filt`	Filtered time series (filtered=smoothed-lowpass

No elaborate edge-treatment yet

Kira Rehfeld krehfeld@awi.de

Rehfeld, K., Marwan, N., Heitzig, J. and Kurths, J. (2011) Comparison of correlation analysis techniques for irregularly sampled time series, Nonlinear Processes in Geophysics, 18 (3), pp. 389-404. doi:10.5194/npg-18-389-2011

tsc_dep_var

## Generate one gamma-distributed and one regular time axis
tx<-generate_t(dt=1,tmin=0,tmax=100,method="gamma")
ty<-generate_t(dt=1,tmin=0,tmax=100,method="linear")
## Simulate one coupled AR1 process (see reference for details)
Proc<-car(tx,ty,coupl_strength=0.5,phi=0.5,lag=0,nsur=1)
## Bind the results to zoo time series
x<-zoo(Proc$x,order.by=tx)
y<-zoo(Proc$y,order.by=ty)


plot(x)
lines(gaussbandpass(x,10,50)$trend,lwd=2)
lines(gaussbandpass(x,10,50)$smoothed,lwd=2,col="limegreen")
lines(gaussbandpass(x,10,50)$filt,col="red2",lwd=2)
colrs<-c("black","black","limegreen","red")
legend("bottom",c("original ts","trend","smoothed","filtered"),col=colrs,lty=1,lwd=c(1,2,2,2))