mdFnn: Find optimal embedding dimension of a multi-dimensional...

In crqa: Unidimensional and Multidimensional Methods for Recurrence Quantification Analysis

Find optimal embedding dimension of a multi-dimensional dataset.

Description

Computes the percentage of false nearest neighbors for multidimensional time series as a function of embedding dimension.

Usage

 mdFnn(data, tau, maxEmb, numSamples, Rtol, Atol) 

Arguments

data	The matrix of data to estimate FNN.
tau	Time delay for embedding.
maxEmb	Maximum number of embedding dimensions considered
numSamples	Number of randomly drawn coordinates from phase-space used to estimate FNN
Rtol	First distance criterion for separating false neighbors
Atol	Second distance criterion for separating false neighbors

Value

It returns the percentage of false neighbors for each embedding.

Author(s)

Sebastian Wallot, Max Planck Insitute for Empirical Aesthetics Dan Moenster, Aarhus University, Moreno I. Coco, University of East London

References

Kennel, M. B., Brown, R., & Abarbanel, H. D. (1992). Determining embedding dimension for phase-space reconstruction using a geometrical construction. Physical review A, 45, 3403. Wallot, S., and Moenster, D. (2018). Calculation of average mutual information (AMI) and false-nearest neighbors (FNN) for the estimation of embedding parameters of multidimensional time-series in Matlab. Front. Psychol. - Quantitative Psychology and Measurement

See Also

mdDelay, optimizeParam

Examples

tau = 1; maxEmb = 10; numSamples = 500; Rtol = 10; Atol = 2

data(crqa) ## load the data

handset = handmovement[1:300, ] ## take less points

mdFnn(handset, tau, maxEmb, numSamples, Rtol, Atol)

crqa documentation built on Nov. 27, 2023, 5:10 p.m.