Kyle Walker walkerke

Choropleth example using Mapbox GL JS

Choropleth using an external geojson (DC Census Tracts by Population Change - 2000), data-driven styling, and queryRenderedFeatures to show hovered polygon's features in a sidebar tooltip.

This hit #rstats today:

Has anyone made a dumbbell dot plot in #rstats, or better yet exported to @plotlygraphs using the API? https://t.co/rWUSpH1rRl
— Ken Davis (@ken_mke) October 23, 2015

So, I figured it was worth a cpl mins to reproduce.

While the US gov did give the data behind the chart it was all the data and a pain to work with so I used WebPlotDigitizer to transcribe the points and then some data wrangling in R to clean it up and make it work well with ggplot2.

It is possible to make the top "dumbbell" legend in ggplot2 (but not by using a guide) and color the "All Metro A

Main differences are:

All the initial data acquisition & munging is kept in one pipeline and the data structure is kept as a tbl_df
I ended up having to use html_session since the site was rejecting the access (login req'd) w/o it
The for loop is now an apply iteration and pbapply gives you a progress bar for free which is A Good Thing given how long that operation took :-)
The move to pbapply makes it possible to do the row-binding and left-joining in a single pipeline, which keeps everything in a tbl_df.
Your for solution can be made almost as efficient if you do a img_list <- vector("list", 150) so the list size is pre-allocated.

BTW: the popup code is brilliant and tigris is equally as brilliant!

	library(tidyverse)
	library(sf)
	library(tigris)

	# start by picking a state from https://github.com/Microsoft/USBuildingFootprints
	# WARNING: these files can be pretty big. using arizona for its copious subdivisions and reasoanable 83MB.
	url_footprint <- "https://usbuildingdata.blob.core.windows.net/usbuildings-v1-1/Arizona.zip"
	download.file(url_footprint, "Arizona.zip")
	unzip("Arizona.zip")

	<!DOCTYPE html>
	<html>
	<head>
	<meta charset='utf-8' />
	<title></title>
	<link href='https://api.tiles.mapbox.com/mapbox-gl-js/v0.26.0/mapbox-gl.css' rel='stylesheet' />
	<link href='https://www.mapbox.com/base/latest/base.css' rel='stylesheet' />
	<link href='site.css' rel='stylesheet' />
	</head>
	<body class='clip loading'>

	library(magrittr)
	library(dplyr)
	library(ggplot2)

	population <- read.csv("https://raw.githubusercontent.com/andriy-gazin/datasets/master/ageSexDistribution.csv")

	population %<>%
	tidyr::gather(sex, number, -year, - ageGroup) %>%
	mutate(ageGroup = gsub("100 і старше", "≥100", ageGroup),
	ageGroup = factor(ageGroup,

	purrr::map_df(2009:2014, function(i) {
	api_info <- jsonlite::fromJSON(sprintf("http://api.census.gov/data/%s/acs5/", i))
	api_info$dataset[, c("title", "c_unavailableMessage")]
	})

	#' @export
	coord_proj <- function(proj="+proj=robin +lon_0=0 +x_0=0 +y_0=0 +ellps=WGS84 +datum=WGS84 +units=m +no_defs",
	inverse = FALSE, degrees = TRUE,
	ellps.default="sphere", xlim = NULL, ylim = NULL) {
	try_require("proj4")
	coord(
	proj = proj,
	inverse = inverse,
	ellps.default = ellps.default,
	degrees = degrees,

	{
	"cells": [
	{
	"cell_type": "code",
	"execution_count": 1,
	"metadata": {
	"collapsed": true
	},
	"outputs": [],
	"source": [

	# TODO: Add a Shiny dropdown to select demographic variable
	library(leaflet)
	library(noncensus)
	library(dplyr)

	data("counties", package="noncensus")
	data("county_polygons", package="noncensus")
	data("quick_facts", package="noncensus")

	counties <- counties %>%