Hurst exponent via rescaled-range (R/S) analysis

Estimates the Hurst exponent \(H\): a long-memory measure of a time series. \(H = 0.5\) indicates uncorrelated (Brownian) increments; \(H > 0.5\) indicates persistent (trending) behaviour; \(H < 0.5\) indicates anti-persistent (mean-reverting) behaviour.

Usage

morie_hurst_r(x)

Arguments

x

Numeric vector (time series).

Value

List with H (numeric) and interpretation ("persistent"/"anti-persistent"/"random").

Examples

# \donttest{
if (requireNamespace("pracma", quietly = TRUE)) {
  set.seed(1)
  x <- cumsum(rnorm(2048)) # Brownian motion, expected H ~ 0.5
  res <- morie_hurst_r(x)
  res$interpretation
}
#> [1] "persistent"
# }