Data from: Case report: Using accelerometers for tracking loggerhead and green sea turtle behaviour.

Data and R Scripts for labelling Green turtle and Loggerhead turtle accelerometer data, calculating summary metrics (for different smoothing windows and sampling frequencies) and running a random forest model. X.1-Label the accelerometer data with behavioural labels from BORIS X.2-Create summary metrics and resample frequencies. This includes code adapted from (Clark, 2019; Clark et al., 2022). https://ore.exeter.ac.uk/repository/handle/10871/120152, https://www.int-res.com/abstracts/meps/v701/p145-157/ X.2 produces labelled summary metrics (2hztrainingmetrics200p2seglength_loggerheadTurtles.csv and 2hztrainingmetrics200p2seglength_greenTurtles.csv to input into X.3) X.3-Run random forest for each frequency and smoothing window (here only final model (2Hz, 2 second smooth and third scute) has been supplied. Please contact [e-mail] for additional datasets/models. Final Models pos2_2HzRF200smooth_bestRF_greenTurtles.RDS is the final random forest model for green turtles (2Hz, 2 second smooth, third scute). pos2_2HzRF200smooth_bestRF_loggerheadTurtles.RDS is the final random forest model for Loggerhead turtles (2Hz, 2 second smooth, third scute). Breakdown of all 2 second smoothed model results can be found in: combined_behaviouraccuracy.csv