#30DayMapChallenge: Restaurant heat map of Madison

Challenge 15: Food/drink

Show code

library(tidyverse)
library(sf)
library(tmap)
library(mapboxapi)
library(leaflet)
library(leaflet.extras)

Another quick #30DayMapChallenge post. The prompt for day 15: Food/drink. Public Health Madison & Dane County have a dataset on health inspections which includes all licensed restaurants in the county. The data are not geocoded, and so I use the mapboxapi package to geocode the locations:

Show code

establishments <- read_csv("data/Licensed_Establishment.csv")

# geocode and save geocoded data
# est_sf <- establishments %>% 
#   rowwise() %>% 
#   mutate(geometry = mb_geocode(AddressFull, output = "sf"),
#          geometry = geometry$geometry)
# 
# est_sf <- st_sf(est_sf)
# 
# write_rds(est_sf, file = "data/licensed_establishment_geocoded.RDS")
est_sf <- readRDS("data/licensed_establishment_geocoded.RDS")

The dataset includes “all operational licensed establishments receiving routine inspections,” that is, not just restaurants. A quick tmap (we have to filter out a food cart that’s located in Milwaukee):

Show code

tmap_mode("plot")
est_sf %>% 
  filter(AddrCity != "MILWAUKEE") %>% 
  tm_shape() +
  tm_dots("EstablishmentType") +
  tm_layout(legend.outside = T)

Now we filter to only establishments that are of the type “Primarily Restaurant” and create a heat map. Heat maps look cool, but I admittedly don’t fully understand how leaflet.extras generates them under the hood – certainly the map looks very different depending on the zoom level.

Show code

leaflet(est_sf %>% filter(EstablishmentType == "Primarily Restaurant")) %>% 
  addTiles() %>% 
  addHeatmap(radius = 6)