On Friday a colleague showed me an interesting chart, a map of maps. I believe the original was made in Tableau, but I decided to spin one up in R. I tweeted out the picture: A map of maps, showing the correlation between state house price growth rates You see pretty strong spatial correlation, with some interesting exceptions. Florida correlated with AZ, NV pic.twitter.com/9hzwZLkb41 — 📈 𝙻𝚎𝚗 𝙺𝚒𝚎𝚏𝚎𝚛 📊 (@lenkiefer) March 6, 2020 In this post I will supply the R code to make one.
I have been thinking about how the recent volatility could impact the economy. If travel and tourism contract due to fears of a pandemic, the impact will differ in markets around the United States. One way to think about this is to compute the Location Quotient, or the percentage of the employment in an area that is in the leisure and hospitality industry. Conside the graphic below: This map shows areas (states and core based statistical areas) color-coded by their location quotient.
Let us take a look at house price trends in the United States and across states and metro areas. Earlier this week I tweeted out a few charts on housing market trends. In most of the middle part of the country over the past 44 years there has been little growth in real (inflation-adjusted) house prices. In coastal states, a very different story. pic.twitter.com/PLbiNftha3 — 📈 𝙻𝚎𝚗 𝙺𝚒𝚎𝚏𝚎𝚛 📊 (@lenkiefer) July 10, 2019 In this post we’ll analyze real house prices since 1975, and per usual use R to wrangle data and make plots.
In this post I want review some trends in U.S. housing supply and demand. Specifically I want to look at county level trends in population, housing supply (the total number of housing units) and house prices. We’ll uncover some interesting trends. Per usual we will make our graphics with R. Preparing the data required several steps that I will outline in a follow up post. For now we’ll just proceed with the data I’ve put together.
Today was JOLTS Tuesday, when the U.S. Bureau of Labor Statistics releases updated data from the Job Openings and Labor Turnover Survey. I was talking about it earlier today, but before we get into that… If you care about dataviz check this out I saw this on Twitter today via Jon Schwabish. Link to a handy dataviz cheatsheet outlining Jon’s core dataviz principles. Prints out nicely on pdf. Back to the JOLTS.
IN THIS POST WE SHALL EXPLORE VALUE-SUPRESSING UNCERTAINTY PALETTES. One of my favorite new sites is xenographics that gives examples of and links to “weird, but (sometimes) useful charts”. The examples xenographics gives are undoubtedly interesting and might help inspire you if you’re looking for something new. One new (to me) graphic was something called Value-Suppressing Uncertainty Palettes (VSUP). See this research paper (pdf). VSUPs “allocate larger ranges of a visual channel when uncertainty is low, and smaller ranges when uncertainty is high”.
I HAVE BEEN EXPERIMENTING WITH A NEW WAY TO VISUALIZE DATA, a bivariate tilegridmap. When I get around to rolling out my tidyPowerPoint workflow we’re going to want something other than bars and lines to fill it up. A graph like this might be a fun option. We’ll build one, but first, just let me show you one I tweeted earlier today: bivariate #tilegridmap map anyone? pic.twitter.com/y3G5XExzoN — 📈 𝙻𝚎𝚗 𝙺𝚒𝚎𝚏𝚎𝚛 📊 (@lenkiefer) October 11, 2017 In this post, let’s go over how to make this plot with R.
TIME TO TRY OUT ANOTHER HOUSE PRICE VISUALIZATION. In this post we’ll try out a new way to visualize recent house price trends with R. Just this wekeend I saw a new package geofacet for organizing ggplot2 facets along a geographic grid. It allows use to construct a small multiple graph that roughly looks like the United States. (Thanks to [@yoniceedee](https://twitter.com/yoniceedee) for recommending geofacet). Let’s try it out using the same house price data we visualized recently.
IN THIS POST I WANT TO EXTEND ON yesterday’s post and build an animated bivariate choropleth. We’ll use the same data as yesterday and create a combined scatterplot with bivariate choropleth map and animate it with R. Let’s get right to it. Load data We’ll follow from yesterday and load our data and do some manipulations. In order for this to work we’ll need data from three sources: House Price Index Data in a .
NOTE: After I posted this (like within 5 minutes) I found this post which also constructs bivariate chropleths in R. IN THIS POST I WANT TO REVISIT SOME MAPS I MADE LAST YEAR. At that time, I was using Tableau to create choropleth maps, but in this post I want to reimagine the maps and make them in R. Last year in this post we looked at the relationship between population growth and the growth in housing units from 2010 to 2015.
LET’S PIXELATE AMERICA. This morning I happened across a fun blog post on how to generate Pixel maps with R via R weekly. The basic code is so easy, all you need is ggplot2 (which I get from the tidyverse). library(tidyverse) ## Warning: package 'tidyverse' was built under R version 3.5.1 ## -- Attaching packages -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- tidyverse 1.2.1 -- ## v ggplot2 3.0.0 v purrr 0.2.5 ## v tibble 1.
TODAY THE NATIONAL ASSOCIATION OF REALTORS (NAR) released (press release) data on metro area median sales prices of existing single-family homes (the U.S. Census and HUD report data on new home sales prices in a joint release). NAR makes the data available (Excel file). Let’s take a look at the data: ## Warning: Removed 1 rows containing missing values (position_stack). ## Warning: Removed 1 rows containing missing values (geom_text). ## Warning in grid.
I WANT TO SHARE WITH YOU a little bit of code to make this whimsical data visualization: Make a simple map First we can construct a map of the lower 48 U.S. states and add a marker for each city. These data are available in the us.cities data that come with the maps package. library(tidyverse) library(maps) data(us.cities) # get us city data from the package maps # drop AK and HI to get the lower 48 states: us.
IN THIS POST I’M JUST GOING TO share a few data visualizations I’ve been working on today. For most, no code, but these build off my previous posts here and here so you can check there for hints and I’ll post some more examples with code later. Population maps This one shows the evolution of population by county without the distribution plots I included last time. We discussed these data in our last post.
SOMETIMES YOU ACTUALLY LEARN SOMETHING from social media. Today on Twitter I happened across this Tweet via @kyle_e_walker: Anybody want population estimates for every US county & metro back to 1790? Well, here are mine. Have fun!https://t.co/QqdA6226kN — Jonathan Schroeder (@j_p_schroeder) August 15, 2016 Seems somebody posted estimates of the U.S. population by county (defined by 2010 county definitions) going back to 1790. This is a perfect dataset to practice my mapping with R.
IN THIS POST I WANT TO DOCUMENT some R code I’ve recently been working on combining maps and distribution plots. As I discussed earlier lots of interesting data will be released in the fall and I want to be ready for it. Some of these snippets can be recycled when the new data is available. Maps One area of data visualization with R I haven’t explored much is mapping. Part of this reason is because I’ve had other tools to use, but usually it’s because I’m in a hurry.
EARLIER THIS WEEK THE U.S. CENSUS BUREAU released data on population and housing units for counties across the U.S. in 2015. These data reveal important trends in population growth, and help shed light on recent house price trends. Housing unit growth One key factor driving housing market dynamics is the expansion of housing supply (or lack thereof). The updated estimates from Census allow us to see which areas have added the most housing units and how that relates to population and house price trends.