Species Distribution Modeling
Fit explainable habitat occupancy models to presence–absence surveys and evaluate predictions on held-out spatial blocks.
Application guideVerifying your session…
Using Species Distribution Modeling
Fit explainable species–habitat relationships and test out-of-block predictions and model sensitivity.
Presence–absence observations · Regularized logistic fit · Spatial cross-validation · Prediction spread
- Inputs and working data
- Spatially blocked, source-labelled presence/absence observations, moisture and canopy covariates, prediction grid and model parameters.
- Working in the editor
- Plot and grid import, penalty and classification settings, probability curves, fold diagnostics and coefficient report.
- What can I produce?
- Quadrat predictions, fold ranges, coefficients, training scales, source-linked held-out probabilities and model comparison.
- What are the limits?
- Two environmental covariates, observed presence/absence and bounded data. Sampling effort and imperfect detection are not inferred.
- Example
- Compare regularization on independent survey blocks, check reduced habitat contrast and restore the model.