Create a tidymodels workflow for H3-based SDMs

Combines a model specification and a prepared recipe into a single tidymodels workflow. This workflow is suitable for species distribution modeling using H3 hexagonal grids and can be directly fitted or cross-validated.

Usage

h3sdm_workflow(model_spec, sdm_recipe)

Arguments

model_spec

A tidymodels model specification (e.g., logistic_reg(), rand_forest(), or boost_tree()), describing the model type and engine to use for fitting.

sdm_recipe

A tidymodels recipe object, typically created with h3sdm_recipe(), which preprocesses the data and defines predictor/response roles.

Value

A workflow object ready to be used for model fitting with fit() or cross-validation.

Details

The function creates a workflow that combines preprocessing and modeling steps. This encapsulation allows consistent model training and evaluation with tidymodels functions like fit() or fit_resamples(), and is particularly useful when applying multiple models in parallel.

Examples

if (FALSE) { # \dontrun{
library(parsnip)
# Example: Create a tidymodels workflow for H3-based species distribution modeling

# Step 1: Define model specification
my_model_spec <- logistic_reg() %>%
  set_mode("classification") %>%
  set_engine("glm")

# Step 2: Create recipe
my_recipe <- h3sdm_recipe(combined_data)

# Step 3: Combine into workflow
sdm_wf <- h3sdm_workflow(model_spec = my_model_spec, sdm_recipe = my_recipe)
} # }