Binary Segmentation Package

When dealing with time series, sometimes the trend in a signal seems to suddenly change at certain points. In many disciplines it is important to identify these changepoints in order to obtain useful information about the dataset. There are several techniques designed to accomplish this task. Among them, Binary Segmentation is an algorithm that computes changepoints in linearithmic time. Although the predicted changepoints are approximated, it is widely used since it is fast and efficient even for large datasets.

The aim of this project is to create a new package that implements the Binary Segmentation algorithm with support for several distributions and types of change. The algorithm was implemented in C++, by using an Object Oriented approach to create an easily extensible structure, and it was integrated with R via Rcpp. This will be useful for usRs with large datasets who want to obtain a quick segmentation model.