fastcpd_lm: Find change points efficiently in linear regression models

In fastcpd: Fast Change Point Detection via Sequential Gradient Descent

Find change points efficiently in linear regression models

Description

fastcpd_lm() and fastcpd.lm() are wrapper functions of fastcpd() to find change points in linear regression models. The function is similar to fastcpd() except that the data is by default a matrix or data frame with the response variable as the first column and thus a formula is not required here.

Usage

fastcpd_lm(data, ...)

fastcpd.lm(data, ...)

Arguments

data	A matrix or a data frame with the response variable as the first column.
...	Other arguments passed to fastcpd(), for example, segment_count.

Value

A fastcpd object.

See Also

fastcpd()

Examples

if (requireNamespace("mvtnorm", quietly = TRUE)) {
  set.seed(1)
  n <- 300
  p <- 4
  x <- mvtnorm::rmvnorm(n, rep(0, p), diag(p))
  theta_0 <- rbind(c(1, 3.2, -1, 0), c(-1, -0.5, 2.5, -2), c(0.8, 0, 1, 2))
  y <- c(
    x[1:100, ] %*% theta_0[1, ] + rnorm(100, 0, 3),
    x[101:200, ] %*% theta_0[2, ] + rnorm(100, 0, 3),
    x[201:n, ] %*% theta_0[3, ] + rnorm(100, 0, 3)
  )
  result_lm <- fastcpd.lm(cbind(y, x))
  summary(result_lm)
  plot(result_lm)

  set.seed(1)
  n <- 600
  p <- 4
  d <- 2
  x <- mvtnorm::rmvnorm(n, rep(0, p), diag(p))
  theta_1 <- matrix(runif(8, -3, -1), nrow = p)
  theta_2 <- matrix(runif(8, -1, 3), nrow = p)
  y <- rbind(
    x[1:350, ] %*% theta_1 + mvtnorm::rmvnorm(350, rep(0, d), 3 * diag(d)),
    x[351:n, ] %*% theta_2 + mvtnorm::rmvnorm(250, rep(0, d), 3 * diag(d))
  )
  result_mlm <- fastcpd.lm(cbind.data.frame(y = y, x = x), p.response = 2)
  summary(result_mlm)
}

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