Rescale future climate predictors using current model coefficients

Standardises future predictors using the mean and SD from current climate, then applies the glmnet beta weights. This is the same weighting applied upstream in RescaleRasters — here we just do it against a different raster stack.

Usage

rescaleFuture(
  model,
  future_predictors,
  current_predictors,
  training_data,
  pred_mat
)

Arguments

model

glmnet model object from elasticSDM_noPCNM()$Model.

future_predictors

SpatRaster of future climate variables. Names must match those retained in model.

current_predictors

SpatRaster of current climate (provides the standardisation parameters).

training_data

the same data that went into the glmnet model, this is used for calculating variance which is required for the scaling process. From elasticSDM

pred_mat

the Prediction matrix from elasticSDM

Value

SpatRaster with rescaled future predictors (one layer per non-zero coefficient).