In ropensci/stats19: Work with Open Road Traffic Casualty Data from Great Britain

To regenerate the file names use the following script:

library(rvest)
library(stringr)
u = paste0(
  "https://data.gov.uk/dataset/",
  "cb7ae6f0-4be6-4935-9277-47e5ce24a11f/road-safety-data")
u = "https://www.data.gov.uk/dataset/cb7ae6f0-4be6-4935-9277-47e5ce24a11f/road-safety-data"
page = read_html(u)
all_links = page %>%
  html_nodes("a") %>%       # find all links
  html_attr("href")

r = all_links %>% str_subset("\\.csv")

dr = c()
for(i in 1:length(r)) {
  dr[i] = sub("https://data.dft.gov.uk/road-accidents-safety-data/",
              "", URLdecode(r[i]))
  dr[i] = sub("https://data.dft.gov.uk/road-accidents-safety-data/",
              "", dr[i])
}

file_names = setNames(as.list(dr), dr)
file_name_df = tibble::tibble(
  file_name = unlist(file_names),
  url = r
  )
usethis::use_data(file_names, overwrite = TRUE)

# get original stats19 data filenames
download.file("https://github.com/ropensci/stats19/raw/master/data/file_names.rda", "file_names_old.rda")
load("file_names_old.rda")
file_names_old = file_names
usethis::use_data(file_names_old, overwrite = TRUE)
file_names = setNames(as.list(dr), dr)
# compare the objects
waldo::compare(file_names_old, file_names)
class(file_names_old)
file_names_char = unlist(file_names)
writeLines(file_names_char, "data-raw/file_names.txt")
readr::write_csv(file_name_df, "data-raw/file_name_df.csv")
file.edit("data-raw/file_names.txt")
file.remove("file_names_old.rda")
# All file names with 1979 in the name
file_names_1979 = file_names[grepl("1979", names(file_names))]
# $`dft-road-casualty-statistics-casualty-1979-latest-published-year.csv`
# [1] "dft-road-casualty-statistics-casualty-1979-latest-published-year.csv"

# $`dft-road-casualty-statistics-vehicle-1979-latest-published-year.csv`
# [1] "dft-road-casualty-statistics-vehicle-1979-latest-published-year.csv"

# $`dft-road-casualty-statistics-collision-1979-latest-published-year.csv`
# [1] "dft-road-casualty-statistics-collision-1979-latest-published-year.csv"

# 2023 data:
file_names_2023 = file_names[grepl("2023", names(file_names))]

usethis::use_data(file_names, overwrite = TRUE)

The accidents_sample_raw can be (re)generated using:

devtools::load_all()
# Obtained with:
dl_stats19(year = 2023, type = "collision")
accidents_2023_raw = read_collisions(year = 2023)
accidents_2023_raw = get_stats19(year = 2023, type = "collision", data_dir = tempdir(), format = FALSE)
set.seed(350)
sel = sample(nrow(accidents_2023_raw), 3)
accidents_sample_raw = accidents_2023_raw[sel, ]
# accidents_sample = format_collisions(accidents_sample_raw)
accidents_sample = accidents_sample_raw
accidents_sample_formatted = format_collisions(accidents_sample)
waldo::compare(accidents_sample_raw, accidents_sample_formatted)
usethis::use_data(accidents_sample_raw, overwrite = TRUE)
usethis::use_data(accidents_sample, overwrite = TRUE)

Similarly for casualites, use:

# Obtained with:
casualties_2023_raw = get_stats19(year = 2023, type = "casualty", data_dir = tempdir(), format = FALSE)
set.seed(350)
sel = sample(nrow(casualties_2023_raw), 3)
casualties_sample_raw = casualties_2023_raw[sel, ]
# casualties_sample = format_casualties(casualties_sample_raw)
casualties_sample = casualties_sample_raw
casualties_sample_formatted = format_casualties(casualties_sample)
waldo::compare(casualties_sample_raw, casualties_sample_formatted)
usethis::use_data(casualties_sample, overwrite = TRUE)

and for vehicles, use:

vehicles_2023_raw = get_stats19(year = 2023, type = "vehicle", data_dir = tempdir(), format = FALSE)
set.seed(350)
sel = sample(nrow(vehicles_2023_raw), 3)
vehicles_sample_raw = vehicles_2023_raw[sel, ]
vehicles_sample_formatted = format_vehicles(vehicles_sample_raw)
vehicles_sample = vehicles_2023_raw[sel,]
waldo::compare(vehicles_sample_raw, vehicles_sample_formatted)
usethis::use_data(vehicles_sample, overwrite = TRUE)

Also, to (re)generate the police_boundaries data, use:

# Obtained with:
library(sf)
u = "https://opendata.arcgis.com/datasets/3e5a096a8c7c456fb6d3164a3f44b005_3.geojson"
police_boundaries_wgs = sf::st_read(u)
names(police_boundaries_wgs)
police_boundaries = st_transform(police_boundaries_wgs, 27700)
names(police_boundaries)
police_boundaries = police_boundaries[c("pfa16cd", "pfa16nm")]