The ‘social.distance’ clustering matrix is now weighted by the length of time spent interacting with each subject.
The metrics ‘mean.cage.sharing’ and ‘mean.distance.from.all’ now weighted by the length of time spent interacting with each subject.
Bug fixed where daylight saving/summertime changes affected day assignments in edge cases.
Renamed ‘chasing’ to ‘following’ to prefer more neutral terminology.
Ethogram plotting now more flexible for continuous 24-hour plots.
Changed the chase clustering metrics to be ‘row-dominant’ (i.e. actors/sources in rows and recipients/targets in columns). This makes the output more compatible with several packages for hierarchy analysis and graph analysis.
Modified the density kernel (cosine) and bandwidth (smaller) to better identify peak activity times and activity blocks.
Fixed an error in the calculation of pairwise social interactions.
The redundant clustering metric ‘cage.share’ has been removed.