https://github.com/euanmcgonigle/cptnonpar

Development version of the CptNonPar R package
https://github.com/euanmcgonigle/cptnonpar

change-point-detection cran moving-sum nonparametric package r segmentation time-series

Last synced: 7 months ago
JSON representation

Development version of the CptNonPar R package

• Host: GitHub
• URL: https://github.com/euanmcgonigle/cptnonpar
• Owner: EuanMcGonigle
• Created: 2023-05-12T14:08:15.000Z (about 3 years ago)
• Default Branch: main
• Last Pushed: 2025-08-23T07:17:09.000Z (9 months ago)
• Last Synced: 2025-10-22T05:59:27.298Z (7 months ago)
• Topics: change-point-detection, cran, moving-sum, nonparametric, package, r, segmentation, time-series
• Language: R
• Homepage: https://CRAN.R-project.org/package=CptNonPar
• Size: 133 KB
• Stars: 4
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.Rmd
  • Changelog: NEWS.md

Awesome Lists containing this project

README

          
---

output: github_document

---

```{r, include = FALSE}

knitr::opts_chunk$set(

  collapse = TRUE,

  comment = "#>",

  fig.path = "man/figures/README-",

  out.width = "100%"

)

```

# CptNonPar

[![R-CMD-check](https://github.com/EuanMcGonigle/CptNonPar/actions/workflows/R-CMD-check.yaml/badge.svg)](https://github.com/EuanMcGonigle/CptNonPar/actions/workflows/R-CMD-check.yaml)

[![Codecov test coverage](https://codecov.io/gh/EuanMcGonigle/CptNonPar/branch/main/graph/badge.svg)](https://app.codecov.io/gh/EuanMcGonigle/CptNonPar?branch=main)

Nonparametric change point detection for multivariate time series. Implements the NP-MOJO methodology proposed in [McGonigle and Cho (2025)](https://doi.org/doi:10.1093/biomet/asaf024).

## Installation

You can install the released version of `CptNonPar` from [CRAN](https://CRAN.R-project.org) with:

```

install.packages("CptNonPar")

```

You can install the development version of `CptNonPar` from [GitHub](https://github.com/) with:

```

devtools::install_github("https://github.com/EuanMcGonigle/CptNonPar")

```

## Usage

For further examples, see the help files within the package. We can generate an example for change point detection as follows.

We generate a univariate time series of length 1000, with a mean change at time 300, and an autocovariance (but not marginal) change at time 650. Then, we perform the multi-lag NP-MOJO algorithm with lags 0 and 1, and print the estimated change points and the associated clusters:

```{r example}

library(CptNonPar)

n <- 1000

set.seed(123)

noise1 <- stats::arima.sim(model = list(ar = -0.5), n = n, sd = sqrt(1 - 0.5^2))

noise2 <- stats::arima.sim(model = list(ar = 0.5), n = n, sd = sqrt(1 - 0.5^2))

noise <- c(noise1[1:650], noise2[651:n])

signal <- c(rep(0, 300), rep(0.7, 700))

x <- signal + noise

x.c <- np.mojo.multilag(x, G = 166, lags = c(0, 1))

x.c$cpts

x.c$cpt.clusters

```