statint: Compute the test statistics, D and W, for the...

In kerSeg: New Kernel-Based Change-Point Detection

Compute the test statistics, D and W, for the changed-interval alternatives.

Description

This function provides the test statistics, \textrm{D}(t_{1},t_{2}), \textrm{W}(t_{1},t_{2}), and the weighted \textrm{W}(t_{1},t_{2}) for the changed-interval alternatives.

Usage

statint(K, Rtemp, R0, r1, r2)

Arguments

K	A kernel matrix of observations in the sequence.
Rtemp	A numeric vector of k_{i.}, the sum of kernel values for each row i.
R0	The term R_{0}, defined in the paper.
r1	The constant in the test statistics \textrm{Z}_{W,r1}(t_{1},t_{2}).
r2	The constant in the test statistics \textrm{Z}_{W,r2}(t_{1},t_{2}).

Value

Returns a list of test statistics, \textrm{D}(t_{1},t_{2}), \textrm{W}(t_{1},t_{2}), \textrm{W}_{r1}(t_{1},t_{2}), and \textrm{W}_{r2}(t_{1},t_{2}).

Examples

## Sequence : change in the mean in the middle of the sequence.
d = 50
mu = 2
tau = 50
n = 100
set.seed(1)
y = rbind(matrix(rnorm(d*tau),tau), matrix(rnorm(d*(n-tau),mu/sqrt(d)), n-tau))
K = gaussiankernel(y) # Gaussian kernel matrix
R_temp = rowSums(K)
R0 = sum(K)
a = statint(K, R_temp, R0, r1=1.2, r2=0.8)

kerSeg documentation built on Aug. 23, 2023, 1:07 a.m.