Download, unzip, read, clean the Facility Registry Service dataset

Usage

frs_get(
  only_essential_cols = TRUE,
  folder = NULL,
  downloaded_and_unzipped_already = FALSE,
  zfile = "national_single.zip",
  zipbaseurl = "https://ordsext.epa.gov/FLA/www3/state_files/",
  csvname = "NATIONAL_SINGLE.CSV",
  date = Sys.Date()
)

Arguments

only_essential_cols: TRUE by default. used in frs_read()
folder: NULL by default which means it downloads to and unzips in a temporary folder
downloaded_and_unzipped_already: If set to TRUE, looks in folder for csv file instead of trying to download/unzip. Looks in working directory if folder not specified.
zfile: filename, just use default unless EPA changes it
zipbaseurl: url, just use default unless EPA changes it
csvname: name of csv file. just use default unless EPA changes it
date: Retrieval/snapshot date, Sys.Date() by default. This is assigned to both the download_date and released attributes of the returned table. released here means the EJAM FRS snapshot date; it is not EPA's ZIP modification date or the GitHub publication date.

Details

Used by frs_update_datasets()

Uses frs_download(), frs_unzip(), frs_read(), frs_clean()

See examples for how package maintainer might use this.

See source code of this function for more notes. For a developer updating the frs datasets in this package, see frs_update_datasets()

frs_get() invisibly returns the table of data, as a table in data.table format It will download, unzip, read, clean, and set metadata for the data.

This function gets the whole thing in one file from

NATIONAL_SINGLE.CSV from https://ordsext.epa.gov/FLA/www3/state_files/national_single.zip

Other files and related information:

https://www.epa.gov/frs/frs-data-resources
https://www.epa.gov/frs/geospatial-data-download-service
https://www.epa.gov/frs/epa-frs-facilities-state-single-file-csv-download
Also could download individual files from ECHO for parts of the info: https://echo.epa.gov/tools/data-downloads/frs-download-summary for a description of other related files available from EPA's ECHO.

This function creates the following:



 > head(frs_by_programid)
           lat        lon  REGISTRY_ID   program   pgm_sys_id
   1: 44.13415 -104.12563 110012799846     STATE        #5005
   2: 41.16163  -80.07847 110057783590 PA-EFACTS         ++++
   3: 41.21463 -111.96224 110020117862       CIM            0
   4: 29.62889  -83.10833 110040716473 LUST-ARRA            0
   5: 40.71490  -74.00316 110019246163       FIS 0-0000-01097
   6: 40.76395  -73.97037 110019163359       FIS 0-0000-01103

   > frs_by_naics[1:2, ]
           lat        lon  REGISTRY_ID NAICS
   1: 30.33805  -87.15616 110002524055     0
   2: 48.77306 -104.56154 110007654038     0

   > names(frs)
   "lat"    "lon"   "REGISTRY_ID" "PRIMARY_NAME" "NAICS" "PGM_SYS_ACRNMS"

    > head(frs[,1:4]) # looks something like this:
           lat       lon  REGISTRY_ID                    PRIMARY_NAME
   1: 18.37269 -66.14207 110000307695      xyz CHEMICALS INCORPORATED
   x: 17.98615 -66.61845 110000307784                         ABC INC
   x: 17.94930 -66.23170 110000307800                   COMPANY QRSTU


  **WHICH SITES ARE ACTIVE VS INACTIVE SITES**

 See frs_active_ids() or frs_inactive_ids()

 Approx 4.6 million rows total 10/2022.

 table(is.na(frs$lat))
 table(is.na(frs$NAICS))

 It is not entirely clear how to simply identify
 which ones are active vs inactive sites.
 See inst folder for notes on that.
 This as of 2/10/23 is not exactly how ECHO/OECA defines "active"

  **WHICH SITES HAVE LAT LON INFO**

 As of 2022-01-31:  Among all including inactive sites,

  1/3 have no latitude or longitude.
  Even those with lat lon have some problems:
    Some are are not in the USA.
    Some have errors in country code.
    Some use alternate ways of specifying USA.

  **WHICH SITES HAVE NAICS OR SIC INDUSTRY CODES**

 Only 1/4 have both location and some industry code (27

 2/3 lack industry code (have no NAICS and no SIC).
    NAICS vs SIC codes:
 11 percent have both NAICS and SIC,
 9.5 percent have just NAICS =
    (21 percent have NAICS).
 12.5 percent have just SIC.
 2/3 have neither NAICS nor SIC.


  **WHICH COLUMNS TO IMPORT AND KEEP**

 approx 39 columns if all are imported, but most useful 10 is default.

 [1] "REGISTRY_ID"             "PRIMARY_NAME"        "PGM_SYS_ACRNMS"
 [4] "INTEREST_TYPES"    "NAICS_CODES"       "NAICS_CODE_DESCRIPTIONS"
 [7] "SIC_CODES"       "SIC_CODE_DESCRIPTIONS"  "LATITUDE83"
 [10] "LONGITUDE83"


  Some fields are csv lists actually, to be split into separate rows
   to enable queries on NAICS code or program system id:

  PGM_SYS_ACRNMS = 'c', # csv format like AIR:AK999, AIRS/AFS:123,
     NPDES:AK0020630, RCRAINFO:AK6690360312, RCRAINFO:AKR000206516"
  INTEREST_TYPES = 'c', # eg "AIR SYNTHETIC MINOR, ICIS-NPDES NON-MAJOR"
      NAICS_CODES = 'c',  # csv of NAICS

Download, unzip, read, clean the Facility Registry Service dataset

Usage

Arguments

Details

See also