README.md

In bklimowski/tsar: Nonlinear Time Series Analysis in R

ntsar

Nonlinear Time Series Analysis with R is package dedicated to provide tools derived from dynamical system theory for purpose of analysis nonlinear properties of time series. Developed version will contain things like:

• Estimation of embedding dimension and time lag
• Attractor reconstruction with machine learning
• Complex network approaches - visibility graph, transfer and recourrence networks
• Measures of fractal dimension, entropy production and irreversibility

Installation

devtools::install_github("bklimowski/ntsar")

Example

Creation of recurrence matrix from time series.

library(ntsar)
set.seed(448)
n = 500
periodic_series <- (sin(1:n/16) + cos(1:n/10)/2 + tan(1:n/20)) 

plot(periodic_series, type = 'l')

recurrence_matrix(periodic_series, 0.5) %>% 
  image(col=gray.colors(2, start = 1, end = 0),
        xaxt='n',
        yaxt='n')

bklimowski/tsar documentation built on Oct. 12, 2020, 7:34 a.m.