Get posterior predictions from a flocker model
predict_flocker(
flocker_fit,
draw_ids = NULL,
history_condition = FALSE,
new_data = NULL,
mixed = FALSE,
allow_new_levels = FALSE,
sample_new_levels = "uncertainty"
)A `flocker_fit` object
Vector of indices of the posterior draws to be used. If `NULL` (the default) all draws are used in their native order.
Logical indicator of whether to directly condition the posterior for the occupancy state on the observed detection histories. For example, at sites with at least one detection, the true occupancy state conditioned on the history is one with absolute certainty. Without directly conditioning on the history, the occupancy state is controlled exclusively by the posterior distribution for the occupancy probability psi.
Optional new data at which to predict. If `NULL`, predictions are given at the data points used for model fitting. Otherwise, must be a flocker_data object produced by `make_flocker_data`.
When `new_data` is not provided, should random effect levels be drawn from their posteriors (`FALSE`, the default) or re-sampled from their fitted hyperparameters (`TRUE`). The latter can be useful for mixed predictive checking. Note that setting to TRUE is not available for grouping terms involved in phylogenetic random effects or other random effects with specified covariance structures.
Should new_data be allowed to contain new levels for random effects?
If new_data is provided and contains random effect levels not present in the original data, how should predictions be handled? Passed directly to brms::prepare_predictions, which see.
An array of posterior predictions in the same shape as the observations passed to `make_flocker_data()` with posterior iterations stacked along the final dimension.
# \donttest{
unconditioned_preds <- predict_flocker(example_flocker_model_single)
conditioned_preds <- predict_flocker(
example_flocker_model_single,
history_condition = TRUE
)
# }