Build the covariate grid for one month — covariate

The prediction surface a fitted model is projected onto: every covariate grid point for one (year, month), with derived covariates added so the grid carries the same predictors the model was trained on.

Usage

covariate_grid(env_dat, year, month, config)

Arguments

env_dat: covariate data from fetch_covariates()
year: year to extract
month: month to extract
config: a config list, as returned by load_config()

Value

a tibble with lon, lat, and one column per predictor, or NULL if the covariate data has no points for that year and month