This is a collection of functions that allows one to perform thebehavioral change point analysis (BCPA) as described by Gurarie etal. (2009). The key features are estimation of discrete changes intime-series data, notably linear and turning components of gappyvelocity times series extracted from movement data.
The package has been onCRAN for a while. Theversion here is a minor update to address some technical R-check issues.Its development on the Git page will be minimal, mainly because asuperior tool for estimating behavioral changes from mechanisticcontinuous time movement models is available in thesmoovepackage: https://github.com/EliGurarie/smoove. Animal movement analystsare encouraged to switch to using those tools instead (see Gurarie etal. 2017).
That said,bcpa can be very useful for theidentification of structural changes of one-dimensional, Gaussian,irregularly sampled time-series, which can come up in many applications,not necessarily related to movement.
For example, here is a link detailing how to apply this tool directlyto a univariate time series:http://htmlpreview.github.io/?https://github.com/EliGurarie/bcpa/blob/master/examples/UnivariateBCPA.html.
Gurarie, E., R. Andrews and K. Laidre. 2009. A novel method foridentifying behavioural changes in animal movement data.EcologyLetters 12: 395-408.
Gurarie, E., C. Fleming, W.F. Fagan, K. Laidre, J. Hernández-Pliego,O. Ovaskainen. 2017. Correlated velocity models as a fundamental unit ofanimal movement: synthesis and applications.MovementEcology 5:13.