VARDetect (version 0.1.8)

second.step.detect: Backward elimination algorithm for screening in the second step

Description

Backward elimination algorithm function for screening

Usage

second.step.detect(
  data,
  pts,
  omega,
  lambda,
  mu,
  alpha_L = 0.25,
  verbose = FALSE
)

Value

A list object includes

pts

Final selected change points

ic

Values of information criterion

Arguments

data

a n by p dataset matrix

pts

a vector includes all candidate change points obtained by the first step

omega

tuning parameter for the information criterion function

lambda

tuning parameter for sparse component estimation

mu

tuning parameter for low rank component estimation

alpha_L

a numeric value, indicates the size of constraint space of low rank component

verbose

if TRUE, then it provides all information for current stage