d2 {RTisean}R Documentation

Dimension and entropy estimation

Description

Estimates the correlation sum, the correlation dimension and the correlation entropy of a given, possibly multivariate, time series.

Usage

d2(series, l, x = 0, d = 1, M, c, t = 0, R, r, scale = 100, N = 1000, E = FALSE, pretty=FALSE)

Arguments

series a vector or matrix.
l number of data points to be used.
x number of lines to be ignored.
d delay for the delay vectors.
M number of components, maximal embedding dimension
c columns to be read.
t theiler window.
R maximal length scale.
r minimal length scale.
scale number of epsilon values.
N maximal number of pairs to be used.
E use data that is normalized to [0,1] for all components.
pretty clean ouput for pretty printing

Details

The parameter pretty must be set to FALSE if the output of d2 is meant to be post-processed by av_d2, c2d,cdg or c2t.

Value

A list of lists, each composed by as many matrices as the treated length scales and embedding dimensions. The first column of each matrix contains the values of epsilon; the second column contains, according to the list item:

.c2 the correlation sums.
.d2 the local slopes of the logarithm of the correlation sum, the correlation dimension.
.h2 the correlation entropies.

See Also

c1

Examples

## Not run: 

dat <- henon(10000)
d2output <- d2(dat, pretty=TRUE)
cordim <- d2output$.d2
plot(cordim[[1]],t="l",ylim=c(0,7),col=2,xlab="Epsilon",
ylab=expression(D[2](m,epsilon)),log="x", main="Correlation Dimension Plot")
for (a in 2:10) 
        lines(cordim[[a]],col=2)

## End(Not run)

[Package RTisean version 3.0.10 Index]