changeTest: A test showing whether two datasets have similar linear...

In changedetection: Nonparametric Change Detection in Multivariate Linear Relationships

Description

The function performs a nonparametric test showing whether two datasets have similar linear structure or not. The test is based on applying energy distance \insertCiterizzo-szekely10changedetection to residuals estimated for each dataset separately but only by one model (either first or second). It is implemented as a permutation test with R rounds and corresponding pzero \insertCitegorskikh17changedetection.

Usage

changeTest(x1, y1, x2, y2, l = NULL, R = 1000, pzero = 0.05,
  alpha = 1)

Arguments

x1	matrix of first period regressors with variables in columns and observations in rows
y1	matrix of first period responses with variables in columns and observations in rows
x2	matrix of second period regressors with variables in columns and observations in rows
y2	matrix of second period responses with variables in columns and observations in rows
l	approximate number of contributing variables (default: overall number of regressors)
R	number of bootstrap rounds (default: '1000')
pzero	trust level for bootstrap (default: '0.05')
alpha	parameter for energy distance formula (default: '1')

Value

TRUE or FALSE

References

\insertAllCited

Examples

T<-60
change<-35
x<-rnorm(n=T, m=0, sd=1)
e<-scale(rt(n=T,3), scale=FALSE)
y1<-5*x[1:(change-1)]+e[1:(change-1)]
y2<--2*x[change:T]+e[change:T]
y<-c(y1,y2)

testResult <- changeTest(x1=as.data.frame(x[1:T]),
                         y1=as.data.frame(y[1:T]),
                         x2=as.data.frame(x[31:T]),
                         y2=as.data.frame(y[31:T]),
                         R=200)

changedetection documentation built on June 17, 2019, 5:03 p.m.