The prinia-project is a pipeline to analyse multispecies acoustic datasets in Python and MATLAB
It allows for easy visualisation and classification based tools to look for
- Acoustic Signal Partitioning in Note Space
- Complexity Analysis of Birdsong
- Repetition Rate Analysis
Methods
- We perform PCA to ordinate song notes as points in N-dimensional space
- Build Linear Discriminant Classifiers using scikitlearn (in Python) and MATLAB
- Perform Higherarchical Clustering to classify song notes
- Extract distributions of repetition rate for species with a single note type
The pipeline can be extended to look for clustering in multi-species community data
Check out our paper in Behavioural Ecology to find out more! (Data in this repository is also available on DRYAD)
Instructions on using binder: Upload your .csv files in the home section, change the filenames in the notebook (prinia_project.ipynb) and you are good to go!