blockedCV function

blockedCV: run spatial blocked cross-validation on the integrated model.

This function is used to perform spatial blocked cross-validation with regards to model selection for the integrated model. It does so by leaving out a block of data in the full model, running a model with the remaining data, and then calculating the deviance information criteria (DIC) as a score of model fit.

blockedCV(
  data,
  options = list(),
  method = "DIC",
  predictName = NULL,
  datasetCombs = NULL
)

Arguments

• data: An object produced by either startISDM of startSpecies. Requires the slot function, .$spatialBlock to be run first in order to specify how the data in the model is blocked.
• options: A list of INLA or inlabru options to be used in the model. Defaults to list().
• method: Which cross-validation method to perform. Must be one of 'DIC' or 'Predict'. If 'DIC' then the DIC values for each block are obtained. If 'Predict' then predictions are made on a dataset in the left out block. For this to work, please specify the argument methodOptions.
• predictName: Name of the dataset to predict onto if method = 'Predict'.
• datasetCombs: A list of vectors containing dataset combinations to include in each model run if method = 'Prediction'. If NULL then all combinations of the dataset will be estimated.

Returns

An object of class blockedCV, which is essentially a list of DIC values obtained from each iteration of the model.

Examples

## Not run:

 if(requireNamespace('INLA')) {
   
 #Get Data
 data("SolitaryTinamou")
 proj <- "+proj=longlat +ellps=WGS84"
 data <- SolitaryTinamou$datasets
 mesh <- SolitaryTinamou$mesh
 mesh$crs <- proj
 
 #Set model up
 organizedData <- startISDM(data, Mesh = mesh,
                            responsePA = 'Present',
                            Projection = proj)
 
 #Set up spatial block
 organizedData$spatialBlock(k = 2, rows = 2, cols = 1)
 
 #Run spatial block cross-validation
 blocked <- blockedCV(organizedData)
 
 #Print summary
 blocked
   
 }
## End(Not run)