Feature selection, Random Forest, and SEM workflow for simulated stand dataset

This dataset and R script provide the workflow for analyzing stand structure and productivity of Chinese fir (Cunninghamia lanceolata) stands using simulated data. The dataset includes key stand, site, and climate variables. The R script demonstrates the following steps:Feature selection with the Boruta algorithmRandom Forest modeling with cross-validation and stepwise variable eliminationMulticollinearity diagnostics using VIFStructural Equation Modeling (SEM) and effect decomposition with piecewiseSEM and semEffNotesThe dataset is simulated for reproducibility and does not include raw inventory data.Users may adapt the script to their own datasets for similar analyses in forestry and ecology.