std_coordinates

Fix and Flag Geographic Coordinates in Biodiversity Records

barRoso::std_coordinates()

Description

Checks the geographic coordinates of biodiversity records following the workflow of the bdc package. Coordinates with latitude and longitude transposed or with inverted signs are fixed, while all other problems are flagged, so the user knows what was fixed and which records need attention.

Details

The function applies, in this order:

Parsing of coordinates stored as text, e.g. with decimal commas ("-12,5"). Values that cannot be read as numbers (e.g. "Bloqueada" in speciesLink) are set to NA and flagged.
bdc::bdc_coordinates_empty() and bdc::bdc_coordinates_outOfRange().
bdc::bdc_country_standardized(), used internally to get the country codes (the original country column is not changed).
bdc::bdc_coordinates_transposed(), which FIXES coordinates with latitude and longitude swapped and/or with inverted signs, based on the country of each record.
bdc::bdc_coordinates_country_inconsistent(), for coordinates falling outside the country informed in the record.
bdc::bdc_coordinates_precision(), for coordinates with few decimals.
CoordinateCleaner::clean_coordinates(), with the tests used in the space module of bdc that do not need downloads: country capitals, country and province centroids, biodiversity institutions, zero and equal coordinates, and the GBIF headquarters.

The flag columns follow the bdc convention, where TRUE means the record passed the test and FALSE means it was flagged: .coordinates_empty, .coordinates_outOfRange, .coordinates_country_inconsistent, .rou, .equ, .zer, .cap, .cen, .gbf, .inst. As in bdc, the column coordinates_transposed has no leading dot because the problem was fixed (FALSE means the coordinates were corrected). So, the output can be used directly with bdc::bdc_summary_col() and bdc::bdc_filter_out_flags().

A column coordinateIssues describes in words what was fixed or flagged in each record, e.g. "transposed, fixed (latitude and longitude swapped); low precision".

Arguments

Argument	Description
df	A data frame containing biodiversity records.
colname_decimalLatitude	Column name for latitude in decimal degrees (default: `"decimalLatitude"`).
colname_decimalLongitude	Column name for longitude in decimal degrees (default: `"decimalLongitude"`).
colname_country	Column name for country (default: `"country"`). If the column is not found, the country-based tests are skipped.
colname_scientificName	Column name for the taxon name, used to report the transposed coordinates (default: `"scientificName"`).
fix_transposed	Logical; if `TRUE`, fix transposed coordinates (default: `TRUE`).
check_country	Logical; if `TRUE`, flag coordinates outside the informed country (default: `TRUE`).
check_space	Logical; if `TRUE`, flag coordinates at capitals, centroids, institutions, zero or equal coordinates and GBIF headquarters (default: `TRUE`).
min_decimals	Minimum number of decimals of both latitude and longitude, below which coordinates are flagged as low precision (default: `2`).
rm_original_column	Logical; if `TRUE`, the original coordinate columns are removed. If `FALSE`, they are retained with the *Original suffix (default: `FALSE`).

Value

The input data frame with numeric coordinates, with transposed coordinates fixed, plus the bdc-like flag columns and the column coordinateIssues.

Examples

df <- read.csv("gbif_download.csv")
df_coords <- std_coordinates(df)

# Summarize the flags and remove flagged records with bdc
df_coords <- bdc::bdc_summary_col(df_coords)
df_clean <- bdc::bdc_filter_out_flags(df_coords, col_to_remove = "all")

--- title: 'std_coordinates' description: 'Fix and Flag Geographic Coordinates in Biodiversity Records' toc: true toc-depth: 3 --- ```{r} #| eval: false barRoso::std_coordinates() ``` ### Description Checks the geographic coordinates of biodiversity records following the workflow of the [bdc](https://brunobrr.github.io/bdc/) package. Coordinates with latitude and longitude transposed or with inverted signs are fixed, while all other problems are flagged, so the user knows what was fixed and which records need attention. ### Details The function applies, in this order: 1. Parsing of coordinates stored as text, e.g. with decimal commas (`"-12,5"`). Values that cannot be read as numbers (e.g. `"Bloqueada"` in speciesLink) are set to `NA` and flagged. 2. `bdc::bdc_coordinates_empty()` and `bdc::bdc_coordinates_outOfRange()`. 3. `bdc::bdc_country_standardized()`, used internally to get the country codes (the original country column is not changed). 4. `bdc::bdc_coordinates_transposed()`, which FIXES coordinates with latitude and longitude swapped and/or with inverted signs, based on the country of each record. 5. `bdc::bdc_coordinates_country_inconsistent()`, for coordinates falling outside the country informed in the record. 6. `bdc::bdc_coordinates_precision()`, for coordinates with few decimals. 7. `CoordinateCleaner::clean_coordinates()`, with the tests used in the space module of bdc that do not need downloads: country capitals, country and province centroids, biodiversity institutions, zero and equal coordinates, and the GBIF headquarters. The flag columns follow the bdc convention, where `TRUE` means the record passed the test and `FALSE` means it was flagged: `.coordinates_empty`, `.coordinates_outOfRange`, `.coordinates_country_inconsistent`, `.rou`, `.equ`, `.zer`, `.cap`, `.cen`, `.gbf`, `.inst`. As in bdc, the column `coordinates_transposed` has no leading dot because the problem was fixed (`FALSE` means the coordinates were corrected). So, the output can be used directly with `bdc::bdc_summary_col()` and `bdc::bdc_filter_out_flags()`. A column `coordinateIssues` describes in words what was fixed or flagged in each record, e.g. `"transposed, fixed (latitude and longitude swapped); low precision"`. ### Arguments | Argument | Description | |---|---| | df | A data frame containing biodiversity records. | | colname_decimalLatitude | Column name for latitude in decimal degrees (default: `"decimalLatitude"`). | | colname_decimalLongitude | Column name for longitude in decimal degrees (default: `"decimalLongitude"`). | | colname_country | Column name for country (default: `"country"`). If the column is not found, the country-based tests are skipped. | | colname_scientificName | Column name for the taxon name, used to report the transposed coordinates (default: `"scientificName"`). | | fix_transposed | Logical; if `TRUE`, fix transposed coordinates (default: `TRUE`). | | check_country | Logical; if `TRUE`, flag coordinates outside the informed country (default: `TRUE`). | | check_space | Logical; if `TRUE`, flag coordinates at capitals, centroids, institutions, zero or equal coordinates and GBIF headquarters (default: `TRUE`). | | min_decimals | Minimum number of decimals of both latitude and longitude, below which coordinates are flagged as low precision (default: `2`). | | rm_original_column | Logical; if `TRUE`, the original coordinate columns are removed. If `FALSE`, they are retained with the *Original suffix (default: `FALSE`). | ### Value The input data frame with numeric coordinates, with transposed coordinates fixed, plus the bdc-like flag columns and the column `coordinateIssues`. ### Examples ```r df <- read.csv("gbif_download.csv") df_coords <- std_coordinates(df) # Summarize the flags and remove flagged records with bdc df_coords <- bdc::bdc_summary_col(df_coords) df_clean <- bdc::bdc_filter_out_flags(df_coords, col_to_remove = "all") ```