a Joint Species Distribution Model for Presence-Only Data
This repository contains working examples for running (and fitting) the model described in the article: /A taxonomic-based joint species distribution model for presence-only data/ by Juan Escamilla-Molgora, Luigi Sedda, Peter Diggle and Peter Atkinson in 2020.
The model uses PyStan, Jupyter notebooks and Pandas.
The best way to create a Conda environment.
The files with the necessary libraries and packages are stored as: requirements.yml
To install the environment (with a working Anaconda installation) do:
conda env create -f requirements.yml
IMPORTANT: The library PyStan needs to be version 2.x. The notebooks would not work with version 3.x.
Due to Github policy restrictions on file sizes, some of the data needed are compressed in the folder 'data/'. Please unzip these files before.
The folder 'notebooks' contains the working examples of case studies 1 and 2.