kcpRS_workflow: KCP on the Running Statistics Workflow

In kcpRS: Kernel Change Point Detection on the Running Statistics

KCP on the Running Statistics Workflow

Description

Any of the four basic running statistics (i.e., running means, running variances, running autocorrelations and running correlations) or a combination thereof can be scanned for change points.

Usage

kcpRS_workflow(
  data,
  RS_funs = c("runMean", "runVar", "runAR", "runCorr"),
  wsize = 25,
  nperm = 1000,
  Kmax = 10,
  alpha = 0.05,
  varTest = FALSE,
  bcorr = TRUE,
  ncpu = 1
)

## S3 method for class 'kcpRS_workflow'
plot(x, ...)

## S3 method for class 'kcpRS_workflow'
print(x, ...)

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

Arguments

data	data N x v dataframe where N is the number of time points and v the number of variables
RS_funs	a vector of names of the functions that correspond to the running statistics to be monitored. Options available: "runMean"=running mean, "runVar"=running variance, "runAR"=running autocorrelation and "runCorr"=running correlation.
wsize	Window size
nperm	Number of permutations used in the permutation test
Kmax	Maximum number of change points desired
alpha	Significance level for the full KCP-RS workflow analysis if bcorr=1. Otherwise, this is the significance level for each running statistic.
varTest	If set to TRUE, only the variance DROP test is implemented, and if set to FALSE, both the variance test and the variance DROP tests are implemented.
bcorr	Set to TRUE if Bonferonni correction is desired for the workflow analysis and set to FALSE otherwise.
ncpu	number of cpu cores to use
x	An object of the type produced by kcpRS_workflow
...	Further plotting arguments
object	An object of the type produced by kcpRS_workflow

Details

The workflow proceeds in two steps: First, the mean change points are flagged using KCP on the running means. If there are significant change points, the data is centered based on the yielded change points. Otherwise, the data remains untouched for the next analysis. Second, the remaining running statistics are computed using the centered data in the first step. The user can specify which running statistics to scan change points for (see RS_funs and examples). Bonferonni correction for tracking multiple running statistics can be implemented using the bcorr option.

Value

kcpMean	kcpRS solution for the running means. See output of kcpRS for further details.
kcpVar	kcpRS solution for the running variances. See output of kcpRS for further details.
kcpAR	kcpRS solution for the running autocorrelations. See output of kcpRS for further details.
kcpCorr	kcpRS solution for the running correlations. See output of kcpRS for further details.

References

Cabrieto, J., Adolf, J., Tuerlinckx, F., Kuppens, P., & Ceulemans, E. (2019). An objective, comprehensive and flexible statistical framework for detecting early warning signs of mental health problems. Psychotherapy and Psychosomatics, 88, 184-186. doi:10.1159/000494356

Examples

phase1=cbind(rnorm(50,0,1),rnorm(50,0,1)) #phase1: Means=0
phase2=cbind(rnorm(50,1,1),rnorm(50,1,1)) #phase2: Means=1
X=rbind(phase1,phase2)

#scan all statistics

res=kcpRS_workflow(data=X,RS_funs=c("runMean","runVar","runAR","runCorr"),
wsize=25,nperm=1000,Kmax=10,alpha=.05, varTest=FALSE, bcorr=TRUE, ncpu=1)
summary(res)
plot(res)


#scan the mean and the correlation only
res=kcpRS_workflow(data=X,RS_funs=c("runMean","runCorr"),wsize=25,nperm=1000,Kmax=10,
    alpha=.05, varTest=FALSE, bcorr=TRUE, ncpu=1)
summary(res)
plot(res)

kcpRS documentation built on Oct. 25, 2023, 5:07 p.m.