Protocol workflow for manuscript 'A novel protocol for exploratory analysis of unknown sound-types in large acoustic datasets'
Creators
- 1. Griffith University
- 2. Queensland University of Technology
- 3. CSIRO Environment
Description
This repository contains the R code associated with the publication "A novel protocol for exploratory analysis of unknown sound-types in large acoustic datasets", published in Methods in Ecology and Evolution. The code implements a semi-automated protocol for clustering environmental sound events using beta acoustic indices and kmeans clustering. While the published example focuses on underwater recordings from river ecosystems, the protocol is broadly applicable to ecoacoustic data from any ecosystem type and is scalable to datasets of most sizes. The scripts include functions for calculating acoustic indices, generating dissimilarity matrices, performing clustering, evaluating cluster validity, and optionally sampling sound events for manual review. A sample dataset is included for demonstration.
Files
data.zip
Additional details
Dates
- Accepted
-
2025-07-23Code made publicly available to accompany the Methods in Ecology and Evolution publication
Software
- Programming language
- R
- Development Status
- Active