Identify transposed geographic coordinates

This function flags and corrects records when latitude and longitude appear to be transposed.

bdc_coordinates_transposed(
  data,
  id = "database_id",
  sci_names = "scientificName",
  lat = "decimalLatitude",
  lon = "decimalLongitude",
  country = "country",
  countryCode = "countryCode",
  border_buffer = 0.2,
  save_outputs = FALSE
)

Arguments

data

data.frame. Containing a unique identifier for each record, geographical coordinates, and country names. Coordinates must be expressed in decimal degrees and WGS84.

id

character string. The column name with a unique record identifier. Default = "database_id".

sci_names

character string. The column name with species scientific name. Default = "scientificName".

lat

character string. The column name with latitude. Coordinates must be expressed in decimal degrees and WGS84. Default = "decimalLatitude".

lon

character string. The column with longitude. Coordinates must be expressed in decimal degrees and WGS84. Default = "decimalLongitude".

country

character string. The column name with the country assignment of each record. Default = "country".

countryCode

character string. The column name with an ISO-2 country code.

border_buffer

numeric >= 0. A distance in decimal degrees used to created a buffer around the country. Records within a given country and at a specified distance from the border will be not be corrected. Default = 0.2 (~22 km at the equator).

save_outputs

logical. Should a table containing transposed coordinates saved for further inspection? Default = FALSE.

Value

The input data, as a tibble with the original column names, plus the column "coordinates_transposed" when at least one record was corrected. The column indicates if verbatim coordinates were not transposed (TRUE). Otherwise records are flagged as (FALSE) and, in this case, verbatim coordinates are replaced by corrected coordinates in the columns named by lat and lon. "coordinates_transposed" is informational: transposed records are corrected, not invalidated, so the column has no leading dot and is not considered by bdc_summary_col or bdc_filter_out_flags.

Details

This test identifies transposed coordinates resulted from mismatches between the country informed for a record and coordinates. Transposed coordinates often fall outside of the indicated country (i.e., in other countries or in the sea). Different coordinate transformations are performed to correct country/coordinates mismatches. Importantly, verbatim coordinates are replaced by the corrected ones in the returned database. A database containing verbatim and corrected coordinates is created in "Output/Check/01_coordinates_transposed.csv" if save_outputs == TRUE. The columns "country" and "countryCode" can be retrieved by using the function bdc_country_standardized.

See also

Other prefilter: bdc_basisOfRecords_notStandard(), bdc_coordinates_country_inconsistent(), bdc_coordinates_empty(), bdc_coordinates_from_locality(), bdc_coordinates_outOfRange(), bdc_country_standardized(), bdc_scientificName_empty()

Examples

if (FALSE) { # \dontrun{
id <- c(1, 2, 3, 4)
scientificName <- c(
  "Rhinella major", "Scinax ruber",
  "Siparuna guianensis", "Psychotria vellosiana"
)
decimalLatitude <- c(63.43333, -14.43333, -41.90000, -46.69778)
decimalLongitude <- c(-17.90000, -67.91667, -13.25000, -13.82444)
country <- c("BOLIVIA", "bolivia", "Brasil", "Brazil")

x <- data.frame(
  id, scientificName, decimalLatitude,
  decimalLongitude, country
)

# Get country code
x <- bdc_country_standardized(data = x, country = "country")

bdc_coordinates_transposed(
  data = x,
  id = "id",
  sci_names = "scientificName",
  lat = "decimalLatitude",
  lon = "decimalLongitude",
  country = "country_suggested",
  countryCode = "countryCode",
  border_buffer = 0.2,
  save_outputs = FALSE
)
} # }