project {RTisean}R Documentation

Projective nonlinear noise reduction

Description

Performs locally projective nonlinear noise reduction of a time series.

Usage

project(series, m, q, r, k, i = 1, l, x = 0, c = 1)

Arguments

series a vector or a matrix.
m embedding dimension.
q dimension of manifold.
r radius of neighbourhoods.
k minimal number of neighbours.
i number of iterations.
l number of values to be read.
x number of values to be skipped.
c column to be read.

Value

A matrix containing the filtered time series in the first column and the difference between the original and the filtered time series in the second column.

References

P. Grassberger, R. Hegger, H. Kantz, C. Schaffrath, and T. Schreiber, On noise reduction methods for chaotic data, Chaos 3, 127 (1993); Reprinted in: E. Ott, T. Sauer, and J. A. Yorke, eds., Coping With Chaos, Wiley, New York (1994)

See Also

ghkss

Examples

## Not run: 

x <- 1:500
y <- cos(x/100)^2 - cos(x/200)+ rnorm(500)/10
filteredy <- project(y,m=7,q=2,k=10,r=1)
plot(x, y, t="l", xlab="Time", ylab="Time series", main="Projective nonlinear noise reduction")
lines(x, filteredy[,1], col=2,lwd=1.5)
legend(350,0, c("Noisy data","Filtered data"),fill=c(1,2), bty="n")

## End(Not run)

[Package RTisean version 3.0.10 Index]