loo: Performs the leave-one-out analysis

In mchevalier2/crestr: A Probabilistic Approach to Reconstruct Past Climates Using Palaeoecological Datasets

Performs the leave-one-out analysis

Description

Repeat the reconstructions by removing one taxon at a time.

Usage

loo(x, climate = x$parameters$climate, verbose = TRUE)

Arguments

x	a crestObj produced by the crest.reconstruct or crest functions.
climate	A vector of the climate variables to extract. See accClimateVariables for the list of accepted values.
verbose	A boolean to print non-essential comments on the terminal (default TRUE).

Value

A crestObj object containing the reconstructions and all the associated data.

Examples

## Not run: 
  data(crest_ex)
  data(crest_ex_pse)
  data(crest_ex_selection)
  reconstr <- crest(
    df = crest_ex, pse = crest_ex_pse, taxaType = 0,
    climate = c("bio1", "bio12"), bin_width = c(2, 20),
    shape = c("normal", "lognormal"),
    selectedTaxa = crest_ex_selection, dbname = "crest_example"
  )
  reconstr <- loo(reconstr)

## End(Not run)
## example using pre-saved reconstruction obtained with the previous command.
data(reconstr)
lapply(reconstr$reconstructions$bio12$loo, head)
plot_loo(reconstr)

mchevalier2/crestr documentation built on Feb. 14, 2025, 6:31 p.m.