library(stats19)

crashes_2017 = get_stats19(year = 2017, type = "Accidents", ask = FALSE)
crashes_sf = format_sf(crashes_2017)
west_yorkshire =
  police_boundaries[police_boundaries$pfa16nm == "West Yorkshire", ]
crashes_wy = crashes_sf[west_yorkshire, ]
casualties_2017 = get_stats19(year = 2017, type = "Casualties", ask = FALSE)
vehicles_2017 = get_stats19(year = 2017, type = "Vehicles", ask = FALSE)
library(tidyr)
library(dplyr)

sel = casualties_2017$accident_index %in% crashes_wy$accident_index
casualties_wy = casualties_2017[sel, ]
cas_types = casualties_wy %>%
  select(accident_index, casualty_type) %>%
  group_by(accident_index) %>%
  summarise(
    Total = n(),
    walking = sum(casualty_type == "Pedestrian"),
    cycling = sum(casualty_type == "Cyclist"),
    passenger = sum(casualty_type == "Car occupant")
  )
cj = left_join(crashes_wy, cas_types)

base::setdiff(names(cj), names(crashes_wy))

crashes_types = cj %>%
  filter(accident_severity != "Slight") %>%
  mutate(type = case_when(
    walking > 0 ~ "Walking",
    cycling > 0 ~ "Cycling",
    passenger > 0 ~ "Passenger",
    TRUE ~ "Other"
  ))
library(sf)
crashes_pedestrians = crashes_types %>%
  filter(walking > 0)
# convert to lon lat CRS
crashes_pedestrians_lonlat = st_transform(
  crashes_pedestrians, crs = 4326)
n <- nrow(crashes_pedestrians_lonlat)
m <- grDevices::colorRamp(c('red', 'green'))( (1:n)/n )
crashes_pedestrians_lonlat$col <-
  colourvalues::colour_values(
    crashes_pedestrians_lonlat$accident_severity,
    palette = m, include_alpha = F)
geojson <- geojsonsf::sf_geojson(
  crashes_pedestrians_lonlat[,c("col","accident_severity")],
  factors_as_string = FALSE)
# write(geojson, "R/crashes.geojson")