Skip to contents

Minimum convex polygon around records

Source: R/make_mcp.R
make_mcp.Rd

For use as predict boundary for a taxa, but also see make_predict_boundary()

Usage

make_mcp(
  presence,
  out_file,
  force_new = FALSE,
  pres_x = "long",
  pres_y = "lat",
  in_crs = 4326,
  out_crs = in_crs,
  buf = 0,
  clip = NULL
)

Arguments

presence

Cleaned and filtered dataframe of presences

out_file

Character. Path for the mcp to be saved. Will be saved by sfarrow::st_write_parquet(). Currently will not work very well with any full stop in the path. Other file types are changed to .parquet

force_new

Logical. If out_file exists, recreate it?

pres_x, pres_y

Character. Name of the columns in presence that have the x and y coordinates

in_crs

epsg code for coordinates in presence

out_crs

epsg code for coordinates in output mcp. Usually the same as predictors

buf

Distance to buffer the mcp. Passed to the dist argument of sf:st_buffer() so units should follow that guidance.

clip

sf to clip the mcp

Value

sf. out_file saved.

Examples


out_dir <- file.path(system.file(package = "envSDM"), "examples")

data <- file.path(system.file(package = "predicts"), "ex") |>
  fs::dir_ls(regexp = "\\.csv$") |>
  tibble::enframe(name = NULL, value = "path") |>
  dplyr::mutate(taxa = gsub("\\.csv", "", basename(path))
                , presence = purrr::map(path, rio::import, setclass = "tibble")
                , presence = purrr::map(presence
                                        , \(x) x |>
                                          dplyr::filter(!is.na(lat)
                                                        , !is.na(lon)
                                          )
                )
                , out_dir = fs::path(out_dir, taxa)
                , out_mcp = fs::path(out_dir, "mcp.parquet")
  )


# mcps --------

purrr::pwalk(list(data$presence
                  , data$out_mcp
)
, \(x, y) make_mcp(x, y, pres_x = "lon")
)