Unsupervised, multivariate, binary clustering for meaningful annotation of data, taking into account the uncertainty in the data. A specific constructor for trajectory analysis in movement ecology yields behavioural annotation of trajectories based on estimated local measures of velocity and turning angle, eventually with solar position covariate as a daytime indicator, ("Expectation-Maximization Binary Clustering for Behavioural Annotation").
- Corrected a bug in the post-smoothing function (thanks to Simona Imperio for noting it). Also, we have added a method in the post-smoothing function for binClstStck instances.
Some core functionalities have been implemented using Rcpp. This results in a significant increase in computational speed when using large datasets.
With the aim of enhancing the EMbC R-package as a real general purpose algorithm, the dependency with respect to the move R-package has been dropped. Thus, potential users coming from domains other than ecology are not forced to install packages that are not desired. However, Move objects from the move R-package can still be used directly as input data.
The command view was depicting the trajectory without taking into account the proportionality of the axes. This has been corrected in version 1.9.4.
In the multivarite case, commands view() and sctr() depict plots with a light-grey background colour to enhance the visibility of the data points. We added a bg parameter to allow changing this default behaviour.