Datasets and R code supporting the study "Predicting vegetation change for adaptive management under complex disturbance regimes using monitoring data and expert judgements"
Description
Repository contents This repository contains two ZIP archives associated with the study: Sinclair et al. "Predicting vegetation change for adaptive management under complex disturbance regimes using monitoring data and expert judgements." 1. Test_data_and_code.7z Contains the primary dataset, scenario dataset, and R scripts used for data preparation, model fitting, hyperparameter tuning, model evaluation, ensemble modelling, and scenario prediction. The modelling framework evaluates both single-target and multi-target machine-learning approaches using single-year and multi-year datasets. Modelling methods include tree-based models and artificial neural networks. 2. Shiny_app.zip Contains the source code and example input files for the Scenario Elicitation Tool (SET), a web-based application developed using the R package Shiny. The SET was used to collect expert estimates relating to ecological change in grasslands, which contributed to the expert-elicited component of the primary dataset. Example input files are provided and can be modified to support future applications. A separate README file is included within the archive to describe its file structure and contents.