Get predictions from a disaggregation model

This function is called by the pred_pred module and calls disaggregation::predict_model() and disaggregation::predict_uncertainty() and also generates predicted cases. SpatRasters are wrapped to enable asynchronous operation.

pred_pred(
  fit,
  cases = FALSE,
  aggregation = NULL,
  predict_iid = FALSE,
  uncertain = FALSE,
  N = 100,
  CI = 0.95,
  async = FALSE
)

Arguments

fit: disag_model. Object returned by disag_model function that contains all the necessary objects for generating predictions.
cases: logical. Whether to predictions of cases. Default FALSE
aggregation: SpatRaster. The aggregation raster. Default NULL. Must be supplied if cases is TRUE.
predict_iid: logical. Whether to generate predictions including the iid effect
uncertain: logical. Whether or not to generate upper and lower credible intervals
N: numeric. The number of realisations to use when generating uncertainty predictions. Default 100
CI: numeric. The credible interval to user when generating uncertainty predictions. Default 0.95
async: logical. Whether or not the function is being used asynchronously

Value

a list containing the predictions

Author

Simon Smart simon.smart@cantab.net

Examples

common <- readRDS(system.file("extdata", "fit-minimal.rds", package = "disagapp"))
common$fit$data$covariate_rasters <- unwrap_terra(common$fit$data$covariate_rasters)
prediction <- pred_pred(common$fit)