Learn R Programming

nmcdr (version 0.3.0)

isp: Initial Screening Procedure for Detecting Multiple Change-points

Description

Initial screening procedure to exclude most of the inrelevant points based on the Cramer-Von Mises two-sample test statistics, thus NMCD is implemented in a much lower dimensional space.

Usage

isp(x, n = length(x), an = as.integer(log(n)^1.5/2))

Arguments

x
the data vector.
n
the length of the data.
an
an appropriate integer which is the length of each subsequence of the data.

Value

a list is returned with elements:
ncp
the number of candidate change-points
cpp
the position of candidate change-points

Examples

Run this code
t <- isp(Blocks)
t
plot(t)

Run the code above in your browser using DataLab