getRecurrencePoints: A function to get recurrent points in a time series

In KarsTS: An Interface for Microclimate Time Series Analysis

Description

This function is the core of the function createSimpleRm. It finds the recurrence points in a time series, possibly embedded.

Usage

getRecurrencePoints(timSer, embedDim, lagDelay, threshold)

Arguments

timSer	The time series
embedDim	The embedding dimension
lagDelay	The delay for the embedding (in lags)
threshold	The threshold

Details

Two points in the phase space are recurrent when the distance between them (infinite norm) is less than the threshold.
In a recurrence matrix the X positions read rightwards and the Y positions read upwards. Recurrence matrices are symmetric and their diagonal (line X = Y) cannot contain not-recurrent points. For this reason, only the upper triangle is stored.

Value

A list containing the X and Y positions of the recurrent points in the recurrence matrix (upper triangle)

Author(s)

Marina Saez Andreu

Examples

# # Generate time series
timSer <- genTSExample(stationary = FALSE, InKTSEnv = FALSE)
graphics::plot(timSer)

# # Calculate recurrence matrix
RP <- getRecurrencePoints(timSer, 1, 0, threshold = 0.1)
X <- RP$recPointsX
Y <- RP$recPointsY

# # Recurrence plot
LT <- nrow(timSer)
graphics::par(pty = "s")
# Upper triangle
graphics::plot(timSer$time[X], timSer$time[Y], cex = 0.3, col = 4, xlab = "", ylab = "")
# Lower triangle
graphics::points(timSer$time[Y], timSer$time[X], cex = 0.3, col = 4)
# Diagonal
graphics::points(1:LT, 1:LT, cex = 0.3, col = 4)

graphics::par(pty = "m")

KarsTS documentation built on Jan. 16, 2021, 5:07 p.m.