The article introduces the changepoint package, an R package designed for detecting multiple change points in time series data. The package offers a variety of search methods, including the Binary Segmentation, Segment Neighbourhood, and Pruned Exact Linear Time (PELT) algorithms, along with several test statistics. The authors highlight the importance of identifying multiple change points in various applications such as climatology, bioinformatics, and finance. They provide a detailed background on changepoint detection, explaining the likelihood-based framework and the challenges of multiple change point detection. The package is structured around an S4 class called `cpt`, which stores analysis results and provides methods for summarization, plotting, and parameter retrieval. The article includes simulated and practical examples to demonstrate the use of the package, focusing on changes in mean, variance, and both mean and variance. The PELT algorithm is noted for its computational efficiency, making it suitable for large datasets. The package is available on CRAN and is useful for both practitioners and researchers.The article introduces the changepoint package, an R package designed for detecting multiple change points in time series data. The package offers a variety of search methods, including the Binary Segmentation, Segment Neighbourhood, and Pruned Exact Linear Time (PELT) algorithms, along with several test statistics. The authors highlight the importance of identifying multiple change points in various applications such as climatology, bioinformatics, and finance. They provide a detailed background on changepoint detection, explaining the likelihood-based framework and the challenges of multiple change point detection. The package is structured around an S4 class called `cpt`, which stores analysis results and provides methods for summarization, plotting, and parameter retrieval. The article includes simulated and practical examples to demonstrate the use of the package, focusing on changes in mean, variance, and both mean and variance. The PELT algorithm is noted for its computational efficiency, making it suitable for large datasets. The package is available on CRAN and is useful for both practitioners and researchers.