CV.search.DP.poly: Grid search for dynamic programming to select the tuning...

In changepoints: A Collection of Change-Point Detection Methods

Grid search for dynamic programming to select the tuning parameter through Cross-Validation.

Description

Perform grid search for dynamic programming to select the tuning parameter through Cross-Validation.

Usage

CV.search.DP.poly(y, r, gamma_set, delta)

Arguments

y	A numeric vector of observations.
r	An integer scalar order of polynomials.
gamma_set	A numeric vector of candidate tuning parameter associated with the l0 penalty.
delta	A positive integer scalar of minimum spacing.

Value

A list with the following structure:

cpt_hat	A list of vector of estimated change points locations (sorted in strictly increasing order)
K_hat	A list of scalar of number of estimated change points
test_error	A list of vector of testing errors
train_error	A list of vector of training errors

Author(s)

Haotian Xu

References

Yu and Chatterjee (2020) <arXiv:2007.09910>

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))
plot.ts(y)
gamma_set = 3:9
DP_result = CV.search.DP.poly(y, r = 2, gamma_set, delta = 5)
min_idx = which.min(DP_result$test_error)
cpt_init = unlist(DP_result$cpt_hat[min_idx])
local.refine.poly(cpt_init, y, r = 2, delta_lr = 5)

changepoints documentation built on Sept. 4, 2022, 5:06 p.m.