Raw FFI data to tibble

Transform raw FFI plot data into tidy data for easier use

Usage

ffi_to_tibble(
  departments,
  years,
  filter_list = NULL,
  folder,
  clean_empty = NULL,
  as_sf = FALSE,
  ...,
  .parallel_options = furrr::furrr_options(scheduling = 1L, stdout = TRUE),
  .verbose = TRUE
)

Arguments

departments: A character vector with the code for the departments, i.e. "01" or c("01", "10"). See examples for details.
years: A numeric vector with the years to extract de data from.
filter_list: A list of departments and plots to extract the data from. If NULL all plots for the department for all years will be extracted, which can use a big amount of memory. See details.
folder: The path to the folder containing the FFI csv files, as character.
clean_empty: Vector with column names from where to remove empty results. Can be one or more of "tree", "shrub", "herbs" and "regen". If more than one, only plots with data in all columns selected will be retained. Default to NULL, no cleaning is done.
as_sf: Logical indicating if the data must be returned as an spatial object. Default to FALSE.
...: Not used at the moment
.parallel_options: An object of class furrr_options. See furrr_options.
.verbose: Logical controlling if progress messages are shown.

Value

A nested tibble. This tibble contains a row per plot/year combination, with the plot metadata included, as well as columns containing tibbles with tree, shrub, and herbs information. See vignette("inventory_data_tibble", package = "forestables")

Details

This function will take every year specified and will retrieve and transform the plot data for the departments and plots provided. For that, csv files from FFI must reside in the folder indicated in the folder argument.

If no filter_list argument is provided, ffi_to_tibble will attempt to process all plots for the departments and years provided. This will result in sometimes hundred of thousands plots to be extracted, processed and returned, which in turn will cause a big use of memory (specially when running in parallel processes) and long times of calculation. Is better to provide a list of departments with the counties and plots to look for to narrow the process. This filter_list should have the following structure:


   list(
   "01" = 1404119,
   "10" = 900863,
   "11" = c(1436508, 1410492))
   )

forestables package offers workflows to create this automatically, see vignette("selecting_plots", package = "forestables") for more details.

Parallel

Processing the plots from within a year can be done in parallel (forestables uses internally the furrr package for this). This means that, if parallelization is active, several processes are launched to retrieve the plots data for that year. This is repeated for all years provided.

.parallel_options controls the finer details of how parallelization is performed (see furrr_options). But no parallelization can occur without setting first a plan. By default, the chosen plan is sequential, so no parellelization is done. Changing the plan, i.e. to multisession will allow ffi_to_tibble to use parallelization when retrieving the data.

Examples