Time-Varying Seasonal Strength — seasonal.strength.curve • fdars

Computes seasonal strength at each time point using a sliding window, allowing detection of how seasonality changes over time.

Usage

seasonal.strength.curve(
  fdataobj,
  period,
  window_size = NULL,
  method = c("variance", "spectral")
)

Arguments

fdataobj: An fdata object.
period: Known or estimated period.
window_size: Width of the sliding window. Recommended: 2 * period.
method: Method for computing strength: "variance" or "spectral".

Value

An fdata object containing the time-varying seasonal strength curve.

Examples

# Signal that transitions from seasonal to non-seasonal
t <- seq(0, 20, length.out = 400)
X <- ifelse(t < 10, sin(2 * pi * t / 2), rnorm(length(t[t >= 10]), sd = 0.5))
X <- matrix(X, nrow = 1)
fd <- fdata(X, argvals = t)

# Compute time-varying strength
ss <- seasonal.strength.curve(fd, period = 2, window_size = 4)
# plot(ss)  # Shows strength declining around t = 10