pile.up {DecomposeR}

Repeat and stack a signal in central and line symmetry

Description

Repeats and stacks a signal duplicated in central (even) and line (odd) symmetry to apply Ensemble Empirical Mode Decomposition (EEMD) on one single vector following the simple boundary rule of Zeng and He (2004). This allows to avoid the iterations that are typical of EEMD. A complete set of signal is added by default at the upper and lower part of the stack, to be removed in the end process.

Usage

pile.up(xy, dt, n, warn = TRUE)

Arguments

xy

the signal

dt

the depth/time positions of each xy

n

the number of replicates you want. It has to be a multiple of two, as you will generate two stacks: the even and the odd one.

warn

whether you want to be annoyed

Value

a dataframe of the original dt (odt), the stack-modified dt (ndt), the inversion factor to change the even stack into the odd one and vice-versa (invert), the even xy stack (even) and the odd one (odd)

Examples

set.seed(42)

n <- 200
t <- seq_len(n)

p1 <- 25
p2 <- 75

xy <- (1 + 0.6 * sin(t*2*pi/p2)) * sin(t*2*pi/p1)  + 2 * sin(t*2*pi/p2) +
  rnorm(n, sd = 0.5)

inter_dt <- round(runif(length(xy), min = 0.5, max = 1.5),1)
inter_dt[20] <- 20

dt <- cumsum(inter_dt)

opar <- par()$mfrow
par(mfrow = c(1,1))

res <- pile.up(xy, dt, 4)

par(mfrow = c(2,1))
plot(res$ndt, res$even, type = "l", col = "blue")
plot(res$ndt, res$odd,  type = "l", col = "red")

par(mfrow = c(opar))