run_meta_loocv: Running LOOCV on hierarchical model outputs

In movedesign: Study Design Toolbox for Movement Ecology Studies

Running LOOCV on hierarchical model outputs

Description

This function runs hierarchical models on movement tracking data, for mean home range area (AKDE) or continuous-time speed and distance (CTSD) estimates for a sampled population. It leverages the ctmm R package, specifically the meta() function, to obtain population-level mean parameters. This function helps to assess outputs via leave-one-out cross-validation (LOOCV).

Usage

run_meta_loocv(
  rv,
  set_target = c("hr", "ctsd"),
  subpop = FALSE,
  trace = FALSE,
  .progress = FALSE,
  .only_max_m = TRUE,
  .lists = NULL
)

Arguments

rv	A list containing outputs, settings and data objects. Must not be NULL.
set_target	Character. Research target: "hr" for home range or "ctsd" for speed & distance.
subpop	Logical. If TRUE, will run meta-analyses with groups. Default is FALSE.
trace	Logical. If TRUE, prints progress messages. Default is FALSE.
.progress	Logical. If TRUE, will display a progress bar. Default is FALSE.
.only_max_m	Logical. If TRUE, will only run the maximum number of individuals. Default is FALSE.
.lists	A list containing already created meta inputs. Default is NULL.

Value

A data frame containing meta-analysis outputs, including estimates, errors, confidence intervals, and group information.

Author(s)

Inês Silva [e-mail]

Examples

if(interactive()) {
run_meta_loocv(rv, set_target = "hr")
}

movedesign documentation built on June 24, 2025, 9:07 a.m.