Train SDM models

This function is a wrapper to fit models in caret using caretSDM data.

Usage

train_sdm(occ,
          pred = NULL,
          algo,
          ctrl = NULL,
          variables_selected = NULL,
          parallel = FALSE,
          ...)

get_tune_length(i)

algorithms_used(i)

get_models(i)

get_validation_metrics(i)

mean_validation_metrics(i)

add_models(m1, m2)

Arguments

occ: A occurrences or a input_sdm object.
pred: A predictors object. If occ is a input_sdm object, then pred is obtained from it.
algo: A character vector. Algorithms to be used. For a complete list see (https://topepo.github.io/caret/available-models.html) or in caretSDM::algorithms.
ctrl: A trainControl object to be used to build models. See ?caret::trainControl and details.
variables_selected: A vector of variables to be used as predictors. If NULL, predictors names from pred will be used. Can also be a selection method (e.g. 'vif').
parallel: Should a paralelization method be used (not yet implemented)?
...: Additional arguments to be passed to caret::train function.
i: A models or a input_sdm object.
m1: A models object.
m2: A models object.

Value

A models or a input_sdm object.

Details

The object algorithms has a table comparing algorithms available. If the function detects that the necessary packages are not available it will ask for installation. This will happen just in the first time you use the algorithm. caret::trainControl holds multiple resources for validation and model tuning. Make sure to understand its parameters beforehand. As it is a key function in the modeling process, we also implemented spatial crossvalidation on it. You can set methods to be cv_spatial or cv_cluster and train_sdm will detect that and apply the method according to blockCV package.

get_tune_length return the length used in grid-search for tunning.

algorithms_used return the names of the algorithms used in the modeling process.

get_models returns a list with trained models (class train) to each species.

get_validation_metrics return a list with a data.frame to each species with complete values for ROC, Sensitivity, Specificity, with their respectives Standard Deviations (SD) and TSS to each of the algorithms and pseudoabsence datasets used.

mean_validation_metrics return a list with a tibble to each species summarizing values for ROC, Sensitivity, Specificity and TSS to each of the algorithms used.

Author

Luíz Fernando Esser (luizesser@gmail.com) https://luizfesser.wordpress.com

Examples

# Create sdm_area object:
sa <- sdm_area(parana, cell_size = 100000, crs = 6933)
#> ! Making grid over study area is an expensive task. Please, be patient!
#> ℹ Using GDAL to make the grid and resample the variables.

# Include predictors:
sa <- add_predictors(sa, bioc) |> select_predictors(c("bio1", "bio12"))
#> ! Making grid over the study area is an expensive task. Please, be patient!
#> ℹ Using GDAL to make the grid and resample the variables.

# Include scenarios:
sa <- add_scenarios(sa)

# Create occurrences:
oc <- occurrences_sdm(occ, crs = 6933) |> join_area(sa)

# Create input_sdm:
i <- input_sdm(oc, sa)

# Pseudoabsence generation:
i <- pseudoabsences(i, method = "random")

# Custom trainControl:
ctrl_sdm <- caret::trainControl(method = "repeatedcv",
                                number = 2,
                                repeats = 1,
                                classProbs = TRUE,
                                returnResamp = "all",
                                summaryFunction = summary_sdm,
                                savePredictions = "all")

# Train models:
i <- train_sdm(i, algo = c("naive_bayes"), ctrl = ctrl_sdm) |>
suppressWarnings()