local.refine.univar {changepoints}

Local refinement of an initial estimator for univariate mean change points detection.

Description

Perform local refinement for univariate mean change points detection.

Usage

local.refine.univar(cpt_init, y)

Arguments

cpt_init

An integer vector of initial change points estimation (sorted in strictly increasing order).

y

A numeric vector of univariate time series.

Value

An integer vector of locally refined change point estimation.

Author(s)

Haotian Xu

References

Wang, Yu and Rinaldo (2020) <doi:10.1214/20-EJS1710>.

Examples

set.seed(0)
cpt_true = c(20, 50, 170)
y = rnorm(300) + c(rep(0,20),rep(2,30),rep(0,120),rep(2,130))
gamma_set = 1:5
DP_result = CV.search.DP.univar(y, gamma_set, delta = 5)
min_idx = which.min(DP_result$test_error)
cpt_hat = unlist(DP_result$cpt_hat[min_idx])
Hausdorff.dist(cpt_hat, cpt_true)
cpt_LR = local.refine.univar(cpt_hat, y)
Hausdorff.dist(cpt_LR, cpt_true)