A photo of a crashed bike — Ciclista atropellado – CC-BY by Nicanor Arenas Bermejo

Day 20 & 21 of 30DayMapChallenge: « OpenStreetMap » and « Conflict » (previously).

Mapping the accidents between bicycles and cars in 2023 in France. We have had a few sad accidents recently showing a growing attention on cyclist security and the conflicts on the road.

We’ll use the Annual databases of road traffic injuries on an OSM background.

Config

library(dplyr)
library(tidyr)
library(readr)
library(janitor)
library(sf)
library(glue)
library(leaflet)

Data

The data guide is available (in french).

# vehicules-2023.csv
vehicles <- read_csv2(
  "https://www.data.gouv.fr/fr/datasets/r/146a42f5-19f0-4b3e-a887-5cd8fbef057b", 
  name_repair = make_clean_names) 

# caract-2023.csv
caract <- read_csv2(
  "https://www.data.gouv.fr/fr/datasets/r/104dbb32-704f-4e99-a71e-43563cb604f2", 
  name_repair = make_clean_names)

# usagers-2023.csv
user <- read_csv2(
  "https://www.data.gouv.fr/fr/datasets/r/68848e2a-28dd-4efc-9d5f-d512f7dbe66f", 
  name_repair = make_clean_names)

severity <- tribble(
    ~grav, ~severity,
        1, "Unharmed",
        2, "Killed",
        3, "Injured hospitalized",
        4, "Slightly injured") |> 
  mutate(severity = factor(
    severity, 
    labels = c("Killed",
               "Injured hospitalized",
               "Slightly injured",
               "Unharmed")))

Find out

# accidents where car and bikes are involved
bike_car_acc <- vehicles |> 
  filter(catv %in% c(1, 7)) |> # 1: bike ; 7: car
  count(num_acc, catv) |> 
  pivot_wider(names_from = catv, 
              values_from = n, 
              names_prefix = "catv_") |> 
  filter(catv_7 > 0 & catv_1 > 0) |> 
  pull(num_acc)

# bikers injuries
bikers <- vehicles |> 
  filter(num_acc %in% bike_car_acc,
         catv == 1) |>
  left_join(user, join_by(num_acc, id_vehicule)) |> 
  left_join(severity, join_by(grav)) |> 
  count(num_acc, severity) 

bikers_display <- bikers |> 
  mutate(outcome = glue("{severity} ({n})")) |> 
  arrange(severity) |> 
  summarise(.by = num_acc,
            outcome = glue_collapse(outcome, sep = "<br />"))

# accident locations
bike_accidents <- caract |> 
  filter(num_acc %in% bike_car_acc) |> 
  st_as_sf(coords = c("long", "lat"), crs = "EPSG:4326") |> 
  mutate(date = as.Date(glue("{an}-{mois}-{jour}"))) |> 
  left_join(bikers_display, join_by(num_acc))

n_accidents <- nrow(bikers_display)
n_killed <- bikers |> 
  filter(severity == "Killed") |> 
  summarise(n = sum(n)) |> 
  pull(n)

That’s 2858 accidents and 772 bikers killed (only with cars, not counting accidents with vans, lorries,…).

Map

bike_accidents |> 
  leaflet() |> 
  addTiles(attribution = r"(
           <a href="https://r.iresmi.net/">r.iresmi.net</a>.
           data: Ministère de l'intérieur 2023;
           map: <a href="https://www.openstreetmap.org/copyright/">OpenStreetMap</a>)") |> 
  addCircleMarkers(popup = ~ glue("<b>{date}</b><br /><br />
                                  biker status:<br />
                                  {outcome}"),
                   clusterOptions = markerClusterOptions())

Figure 1: Bike accidents with cars in France – 2023

--- title: "Bike accidents" description: "Days 20 & 21 of 30DayMapChallenge. OSM & conflicts" author: "Michaël" date: "2024-11-21T21:00:00" date-modified: last-modified categories: - R - spatial - datavisualization - 30DayMapChallenge - OSM draft: false freeze: true editor_options: chunk_output_type: console chunk_output_type: console execute: eval: true --- ```{r} #| label: preview #| echo: false #| fig-cap: "Ciclista atropellado -- CC-BY by Nicanor Arenas Bermejo" #| fig-alt: "A photo of a crashed bike" #| fig-link: https://www.flickr.com/photos/palabrista/9178280891/ #| class-output: "preview-image" #| out-width: "100%" knitr::include_graphics("images/9178280891_801b751a38_c.jpg") ``` Day 20 & 21 of [30DayMapChallenge](https://30daymapchallenge.com/): « OpenStreetMap » and « Conflict » ([previously](/#category=30DayMapChallenge)). Mapping the accidents between bicycles and cars in 2023 in France. We have had a [few](https://www.lemonde.fr/en/france/article/2024/10/18/cyclist-killed-in-paris-this-case-is-stirring-up-a-great-deal-of-emotion-among-people-who-travel-by-bike_6729762_7.html) [sad](https://www.francebleu.fr/infos/faits-divers-justice/un-rassemblement-a-chatellerault-en-memoire-de-l-homme-en-bleu-6829937) accidents recently showing a growing attention on cyclist security and the conflicts on the road. We'll use the [*Annual databases of road traffic injuries*](https://www.data.gouv.fr/fr/datasets/bases-de-donnees-annuelles-des-accidents-corporels-de-la-circulation-routiere-annees-de-2005-a-2023/) on an OSM background. # Config ```{r} #| label: setup library(dplyr) library(tidyr) library(readr) library(janitor) library(sf) library(glue) library(leaflet) ``` # Data The [data guide](https://www.data.gouv.fr/fr/datasets/r/8ef4c2a3-91a0-4d98-ae3a-989bde87b62a) is available (in french). ```{r} #| label: data #| cache: true # vehicules-2023.csv vehicles <- read_csv2( "https://www.data.gouv.fr/fr/datasets/r/146a42f5-19f0-4b3e-a887-5cd8fbef057b", name_repair = make_clean_names) # caract-2023.csv caract <- read_csv2( "https://www.data.gouv.fr/fr/datasets/r/104dbb32-704f-4e99-a71e-43563cb604f2", name_repair = make_clean_names) # usagers-2023.csv user <- read_csv2( "https://www.data.gouv.fr/fr/datasets/r/68848e2a-28dd-4efc-9d5f-d512f7dbe66f", name_repair = make_clean_names) severity <- tribble( ~grav, ~severity, 1, "Unharmed", 2, "Killed", 3, "Injured hospitalized", 4, "Slightly injured") |> mutate(severity = factor( severity, labels = c("Killed", "Injured hospitalized", "Slightly injured", "Unharmed"))) ``` # Find out ```{r} #| label: processing # accidents where car and bikes are involved bike_car_acc <- vehicles |> filter(catv %in% c(1, 7)) |> # 1: bike ; 7: car count(num_acc, catv) |> pivot_wider(names_from = catv, values_from = n, names_prefix = "catv_") |> filter(catv_7 > 0 & catv_1 > 0) |> pull(num_acc) # bikers injuries bikers <- vehicles |> filter(num_acc %in% bike_car_acc, catv == 1) |> left_join(user, join_by(num_acc, id_vehicule)) |> left_join(severity, join_by(grav)) |> count(num_acc, severity) bikers_display <- bikers |> mutate(outcome = glue("{severity} ({n})")) |> arrange(severity) |> summarise(.by = num_acc, outcome = glue_collapse(outcome, sep = " ")) # accident locations bike_accidents <- caract |> filter(num_acc %in% bike_car_acc) |> st_as_sf(coords = c("long", "lat"), crs = "EPSG:4326") |> mutate(date = as.Date(glue("{an}-{mois}-{jour}"))) |> left_join(bikers_display, join_by(num_acc)) n_accidents <- nrow(bikers_display) n_killed <- bikers |> filter(severity == "Killed") |> summarise(n = sum(n)) |> pull(n) ``` That's `r n_accidents` accidents and `r n_killed` bikers killed (only with cars, not counting accidents with vans, lorries,...). # Map ```{r} #| label: fig-leaflet #| fig-cap: "Bike accidents with cars in France -- 2023" #| fig-alt: "Map of bike accidents with cars in France in 2023" #| out-width: "100%" bike_accidents |> leaflet() |> addTiles(attribution = r"( <a href="https://r.iresmi.net/">r.iresmi.net</a>. data: Ministère de l'intérieur 2023; map: <a href="https://www.openstreetmap.org/copyright/">OpenStreetMap</a>)") |> addCircleMarkers(popup = ~ glue("{date} biker status: {outcome}"), clusterOptions = markerClusterOptions()) ```