summary.SNSeg_Multi: Summary of SN-based change-point estimates for multivariate...

In SNSeg: Self-Normalization(SN) Based Change-Point Estimation for Time Series

Summary of SN-based change-point estimates for multivariate time series with dimension no greater than 10

Description

Summary method for objects of class SNSeg_Multi

Usage

## S3 method for class 'SNSeg_Multi'
summary(object, ...)

Arguments

object	a SNSeg_Multi object
...	not in use

Details

Provide information about estimated change-point locations, the parameter tested by SN-based procedures, the confidence level, the grid_size, and the critical value of the SN-based test.

Examples

# Please run this function before simulation
exchange_cor_matrix <- function(d, rho){
  tmp <- matrix(rho, d, d)
  diag(tmp) <- 1
  return(tmp)
}

# simulation of multivariate time series
library(mvtnorm)
set.seed(10)
d <- 5
n <- 600
nocp <- 5
cp_sets <- round(seq(0, nocp+1 ,1)/(nocp+1)*n)
mean_shift <- rep(c(0,2),100)[1:(length(cp_sets)-1)]/sqrt(d)
rho_sets <- 0.2
sigma_cross <- list(exchange_cor_matrix(d,0))
ts <- MAR_MTS_Covariance(n, 2, rho_sets, cp_sets = c(0,n), sigma_cross)
ts <- ts[1][[1]]

# Test for the change in multivariate means
# grid_size defined
result <- SNSeg_Multi(ts, paras_to_test = "mean", confidence = 0.99,
                      grid_size_scale = 0.05, grid_size = 45)
# summary method
summary(result)

SNSeg documentation built on June 22, 2024, 10:50 a.m.