Rethomics, a framework for high-throughput behaviour analysis in R
2022-02-03
Introduction
Only if we share a common data structure can we use a common set of tools
rethomics
is a free academic software, don’t hesistate to cite rethomics if you use it in your research.
In the last few years, there has been growing interests in ethomics – that is, the analysis of large behavioural data sets. Many software and hardware solutions have been proposed to record different behavioural variables on several model organisms. Although subsequent analysis and visualisation share many similarities, each method tends to provide its own output format and, in practice, its own restricted analysis software. This results in a lot of replicated work but also limits extension and collaboration.
Rethomics attempts to unify analysis of behaviour by providing several packages:
behavr
tables – a flexible and universal structure to handle very large behavioural data setsdamr
,scopr
, … – to load data from DAMS, ethoscopes and others intobehavr
tablesggetho
– based onggplot2
, to produce high-quality representations of behavioural datasleepr
,zeitgebr
, … – to analyse behavioural data (sleep analysis, circadian rhythm, …)
This document is a tutorial intended for experimenters as well as data analysts. It provides a suite of both conceptual explanations and very concrete examples.