Compute MESS (Multivariate Environmental Similarity Surface)

Measures how similar each cell is to the training environment using the full distribution of reference values. Negative MESS values indicate novel (non-analog) environments.

Usage

maxent_mess(env_grids, reference_values, feature_names)

maxent_mess(env_grids, reference_values, feature_names)

Arguments

env_grids: List of external pointers to Grid<float> objects.
reference_values: List of numeric vectors with reference values for each variable (e.g. values at training sites).
feature_names: Character vector of variable names.

Value

A named list with:

mess_grid: External pointer to Grid<float> with MESS values
mod_grid: External pointer to Grid<float> with Most Dissimilar Variable index (1-based)

A named list with:

mess_grid: External pointer to Grid<float> with MESS values
mod_grid: External pointer to Grid<float> with Most Dissimilar Variable index (1-based)

Examples

if (FALSE) { # \dontrun{
result <- maxent_mess(list(g1, g2),
            list(temp_train_vals, precip_train_vals),
            c("temp", "precip"))
mess_mat <- maxent_grid_to_matrix(result$mess_grid)
mod_mat  <- maxent_grid_to_matrix(result$mod_grid)
} # }
if (FALSE) { # \dontrun{
result <- maxent_mess(list(g1, g2),
            list(temp_train_vals, precip_train_vals),
            c("temp", "precip"))
mess_mat <- maxent_grid_to_matrix(result$mess_grid)
mod_mat  <- maxent_grid_to_matrix(result$mod_grid)
} # }