[R] false.nearest on tseriesChaos

[Jeff Newmiller]

This theory question doesn't seem on-topic here. Read the Posting Guide mentioned below... contact the package maintainer if the package description doesn't inform you where to go for support.

On January 3, 2021 11:07:48 PM PST, "Norma Elizabeth Quiroz Pérez" <nquirozp using hotmail.com> wrote:
>Dear all:
>
>I am new in R and in Analysis of Chaotic Time Series. I have seen that
>the package tseriesChaos
>can estimate the embedding dimension using the FNN. In this program
>false.nearest does the work,
>but to use it, I need the escape factor(rt) and the neighborhood
>diameter (eps = sd(x)/10).
>My problem is that I do not understand the definitions of these
>parameters.
>I have read Kennel, Brown, and Abarbanel�s paper. There the authors
>suggest two criteria for the embedding dimension looking at the FNN.
>One is the ratio between the distances in d and d+1 dimensions,
>respectively. If such ratio is bigger than a threshold R_tol, then one
>has a false neighbor.
>The second criterio is that R_{d+1} > A_{tol} x fractal size then one
>has a FNN.
>What these thresholds  R_{tol} and A_{tol}  have to do  with the escape
>factor and the eps?
>
>
>I hope someone can help me.
>Thanks
>
>	[[alternative HTML version deleted]]

-- 
Sent from my phone. Please excuse my brevity.