R/arrange-SpatVector.R

In dieghernan/tidyterra: 'tidyverse' Methods and 'ggplot2' Helpers for 'terra' Objects

#' Order a `SpatVector` using column values
#'
#' @description
#' `arrange.SpatVector()` orders the geometries of a `SpatVector` by the
#' values of selected columns.
#'
#' @export
#' @rdname arrange.SpatVector
#' @name arrange.SpatVector
#'
#' @seealso [dplyr::arrange()]
#'
#' @family single table verbs
#' @family dplyr.rows
#' @family dplyr.methods
#'
#' @importFrom dplyr arrange
#'
#' @param ... <[`data-masking`][rlang::args_data_masking]> Variables, or
#'   functions of variables. Use [`desc()`][dplyr::desc()] to sort a variable in
#'   descending order.
#'
#' @inheritParams distinct.SpatVector
#' @param .by_group If `TRUE`, will sort first by grouping variable. Applies to
#'   grouped `SpatVector` only (see [group_by.SpatVector()]).
#'
#' @return A `SpatVector` object.
#'
#'
#' @section \CRANpkg{terra} equivalent:
#'
#' [terra::sort()]
#'
#' @section Methods:
#'
#' Implementation of the **generic** [dplyr::arrange()] function for
#' `SpatVector` class.
#'
#'
#' @examples
#'
#' library(terra)
#' library(dplyr)
#'
#' v <- vect(system.file("extdata/cyl.gpkg", package = "tidyterra"))
#'
#'
#' # Single variable
#'
#' v %>%
#'   arrange(desc(iso2))
#'
#'
#' # Two variables
#' v %>%
#'   mutate(even = as.double(cpro) %% 2 == 0, ) %>%
#'   arrange(desc(even), desc(iso2))
#'
#'
#' # With new variables
#' v %>%
#'   mutate(area_geom = terra::expanse(v)) %>%
#'   arrange(area_geom)
arrange.SpatVector <- function(.data, ..., .by_group = FALSE) {
  # Use index
  indexvar <- make_safe_index("tterra_index", .data)

  # Use as_tibble conversion avoiding geom
  tbl <- as_tibble(.data)
  tbl[[indexvar]] <- seq_len(nrow(tbl))

  arranged <- dplyr::arrange(tbl, ..., .by_group = .by_group)

  # Regenerate
  vend <- .data
  vend <- vend[arranged[[indexvar]], ]

  vend <- group_prepare_spat(vend, arranged)

  return(vend)
}


#' @export
dplyr::arrange