Listening across coral reefs: dataframe and code

This deposit contains: - "acoustic_metrics_dataframe.csv": a CSV file presenting the values of 503 acoustic metrics calculated for each minute of cleaned soundscape recordings (i.e., minutes affected by boats, waves, and divers have been removed) across 98 sites located in Réunion Island, Mayotte Island, and Guadeloupe Island. - "metrics_selections.r": an R script containing the code used to classify sites based on acoustic metrics using Random Forest and Generalized Linear Models classification models, and to identify each situation (island and day period) discriminant acoustic metrics.本数据集包含以下文件： - "acoustic_metrics_dataframe.csv"：该CSV文件记录了留尼旺岛、马约特岛与瓜德罗普岛共98个监测点位的经预处理的声景录音（已剔除受船只、海浪及潜水员活动干扰的分钟时段）中，每一分钟对应的503项声学指标（acoustic metrics）数值。 - "metrics_selections.r"：该R脚本包含使用随机森林（Random Forest）与广义线性模型（Generalized Linear Models）分类模型，基于声学指标对监测点位进行分类，并识别各场景（岛屿与时段）下判别性声学指标的代码。