Repository navigation
First release of mlsdm package on PyPI.
Capabilities
- Raster Processing: Clip rasters to presence extents, stack rasters, and convert to DataFrames.
- Pseudo-Absence Sampling: Generate pseudo-absence points using PCA+KDE or random sampling methods.
- Machine Learning: Train and evaluate ensemble SDModels (RandomForest, ExtraTrees, XGBoost, LightGBM).
- Hypertuning: Hypertune each individual model
- Visualization: Plot rasters, partial dependence plots, and presence vs. pseudo-absence points.
- Feature Analysis: Analyze feature importance and perform recursive feature elimination (RFE) with Cross-validation (CV).
Edit: Zenodo integration is added