Methods for Nonparametric Changepoint Detection

Implements the multiple changepoint algorithm PELT with a nonparametric cost function based on the empirical distribution of the data. This package extends the changepoint package (see Killick, R and Eckley, I (2014) ).

An R package to detect changepoints using nonparametric methods. This package is an extension to the changepoint R package which looks at parametric methods for changepoint detection.

Currently the method uses PELT with a cost function based on the empirical distribution


J <- function(x){

n <- 1000
tau <- c(0.1,0.13,0.15,0.23,0.25,0.4,0.44,0.65,0.76,0.78,0.81)*n
h <- c(2.01, -2.51, 1.51, -2.01, 2.51, -2.11, 1.05, 2.16, -1.56, 2.56, -2.11)
sigma <- 0.5
t <- seq(0,1,length.out = n)
data <- array()
for (i in 1:n){
   data[i] <- sum(h*J(n*t[i] - tau)) + (sigma * rnorm(1))

out <-, penalty = "SIC",method="PELT",test.stat="empirical_distribution",
             class=TRUE,minseglen=2, nquantiles =4*log(length(data)))

#returns 100 130 150 230 250 400 440 650 760 780 810 as the changepoint locations.


alt tag

####### Heart Rate Example

This example uses heart rate data recorded using a wearable heart rate monitor whilst running. We use the CROPS penalty in this situation. The diagnostic plot gives us an idea of how many changes to choose (the point on the elbow).

cptHeartRate <-, penalty = "CROPS", pen.value = c(25,200), method="PELT",
   test.stat="empirical_distribution",class=TRUE,minseglen=2, nquantiles =4*log(length(data)))

plot(cptHeartRate, diagnostic = TRUE)
plot(cptHeartRate, ncpts = 11)

alt tag alt tag


Version 0.1.1:

*changed from NPPELT or EDPELT to just using PELT. *removed some of the functions copied from the changepointpackage since these have now been exported from changepoint.

Version 0.1.0: Original

I created this package for nonparametric changepoint detection initially using PELT with a cost function based on the empirical distribution

Kaylea Haynes STOR-i Lancaster University

Reference manual

It appears you don't have a PDF plugin for this browser. You can click here to download the reference manual.


1.0.3 by Daniel Grose, 9 months ago

Browse source code at

Authors: Kaylea Haynes [aut] , Rebecca Killick [aut] , Paul Fearnhead [ths, ctb] , Idris Eckley [ths] , Daniel Grose [ctb, cre]

Documentation:   PDF Manual  

Task views: Time Series Analysis

GPL license

Imports methods, zoo, Rdpack

Depends on changepoint

Depended on by changepoint.geo.

See at CRAN