fastcpd_arma: Find change points efficiently in ARMA(p, q) models

In fastcpd: Fast Change Point Detection via Sequential Gradient Descent

Find change points efficiently in ARMA(p, q) models

Description

fastcpd_arma() and fastcpd.arma() are wrapper functions of fastcpd() to find change points in ARMA(p, q) models. The function is similar to fastcpd() except that the data is by default a one-column matrix or univariate vector and thus a formula is not required here.

Usage

fastcpd_arma(data, order = c(0, 0), ...)

fastcpd.arma(data, order = c(0, 0), ...)

Arguments

data	A numeric vector, a matrix, a data frame or a time series object.
order	A vector of length two specifying the order of the ARMA model.
...	Other arguments passed to fastcpd(), for example, segment_count.

Value

A fastcpd object.

See Also

fastcpd()

Examples

set.seed(1)
n <- 300
w <- rnorm(n + 3, 0, 3)
x <- rep(0, n + 3)
for (i in 1:200) {
  x[i + 3] <- 0.1 * x[i + 2] - 0.3 * x[i + 1] + 0.1 * x[i] +
    0.1 * w[i + 2] + 0.5 * w[i + 1] + w[i + 3]
}
for (i in 201:n) {
  x[i + 3] <- 0.3 * x[i + 2] + 0.1 * x[i + 1] - 0.3 * x[i] -
    0.6 * w[i + 2] - 0.1 * w[i + 1] + w[i + 3]
}
result <- suppressWarnings(
  fastcpd.arma(
    data = x[3 + seq_len(n)],
    order = c(3, 2),
    segment_count = 3,
    lower = c(rep(-1, 3 + 2), 1e-10),
    upper = c(rep(1, 3 + 2), Inf),
    line_search = c(1, 0.1, 1e-2),
    beta = "BIC",
    cost_adjustment = "BIC"
  )
)
summary(result)
plot(result)

fastcpd documentation built on May 29, 2024, 8:36 a.m.