listCoefsCVBinREY: Cross-Validation Data for the Count Model ModelCountDetectBin

In ClementCalenge/caperpyogm: Estimation of the Number of Male Capercaillie in the Pyrenees Mountains

Description

This dataset contains the MCMC iterations for the K-fold cross-validation of the model "modelCountDetectBinREY"

Usage

data("listCoefsCVBinREY")

Format

An object of class "CVModelCount"

Details

Each dataset describes the K-fold cross validation results of the model "modelCountDetectBinREY". We show in the examples how this datasets was obtained.

Source

The original Dataset has been kindly provided by the Observatoire des Galliformes de Montagne:

Observatoire des Galliformes de Montagne. 90 impasse des daudes 74320 Sevrier, France.

References

Calenge C., Menoni E., Milhau B., Foulche K, Chiffard J., Marchandeau S. (in prep.). The participatory monitoring of the capercaillie in the French Pyrenees.

Examples

## We describe here the process to calculate the object
## listCoefCVBinREY

## We work on the dataset lekcounts
head(lekcounts)

## We prepare the dataset to fit the model with JAGS
dataList <- dataCount2jags(lekcounts$lek, lekcounts$period,
                           lekcounts$nbobs, lekcounts$nbmales,
                           lekcounts$gr, as.numeric(factor(lekcounts$type)),
                           lekcounts$natun, lekcounts$year)
dataList

## We define 10 groups of 33 leks
set.seed(980)
ooo <- sample(c(rep(1:10,each=33)))

## Performs K-fold validation. WARNING!! THIS CALCULATION TAKES
## SEVERAL HOURS!!!
## Not run: 
listCoefsCVBinREY <- kfoldCVModelCount(ooo, dataList, "modelCountDetectBinREY")

## End(Not run)

## To save time for the user, we have stored the result of this
## command in the dataset listCoefsCVBinREY (for the model
## modelCountDetectBinREY only. We could not include the results of
## cross-validation for other models due to the large object size, but
## we can send them on request).
listCoefsCVBinREY

## Finally, we can use LLCount to calculate the LPD of each lek counts
## for each MCMC iteration, under a model that was not fit using these
## counts.
## WARNING!!! THIS CALCULATION ALSO TAKES MORE THAN ONE HOUR!!!
## Not run: 
llcBinREY <- LLCount(dataList, listCVCoefBinREY, oo)

## End(Not run)

## And the result is stored in the dataset:
llcBinREY

## Expected LPD can be calculated with:
elpdLeks(llcBinREY)

ClementCalenge/caperpyogm documentation built on Sept. 14, 2021, 4:14 p.m.