When I have a spare half hour or so I like to watch episodes of Fixer Upper. My favorite part – of course – is the unveil at the end where viewers get to see the results of whatever makeover has taken place. I find myself thinking about how these magical renovations would look in my house: a wall blown out here or there, a woman cave, a giant kitchen with granite counters, a freezer that dispenses crushed ice and a gourmet chef (can’t a girl dream?).
From watching these shows over time I’ve trained my brain to imagine what *could be* with every room in my house, a skill I never would have if Trading Spaces hadn’t arrived on the scene. I have some great ideas for the reno that will eventually happen…and I bet you do too.
Since the key to opening our brains to greater possibilities seems to be exposure to what could be, I thought I’d try it with data to get you thinking about different ways to use your data,
My inspiration for this post is a data visualization from Huffington Post’s great data team:
“Here’s Where Lawmakers Don’t Want Trans People to Use the Right Bathrooms,” by Alissa Scheller, July 22, 2016
What I love:
The use of squares to build a simple geography
You could use it to:
Show counties within a state, or census tracts, towns, or school districts within a city or county
This approach isn’t ideal if exact geographic placement of the pieces (states, counties, towns) is integral to understanding your data, or if there are so many “pieces” you can’t fit a readable label inside of them.
I thought I’d take this visualization and make it my own, showing you the steps along the way.
1. My first order of business was to settle on geographies. I decided to use my home state (Maine) and my native state (California), so I could compare the counties within those states (16 in Maine, 58 in California). The HuffPost USA-in-squares-example-graphic compares 50 entities; for readability sake you probably wouldn’t want to use this approach for more than 75 entities.
2. Then I set about drawing these states as a composition of their counties – in the shape of squares.
We wrote a blog post a while back called, “How to Rock Your Icon Drawings.” Check that out if you need a few pointers on drawing in your PowerPoint or Keynote programs.
Anyway, Maine was pretty easy to draw and label.
California took me a bit more time because of the variation in county size and the concentration of counties in Northern California.
Note: there is no “right” way to do this. Feel free to move things around a bit until they feel right to you.
3. My next task was to find a data set. A fun data set. And this is where the beer comes in. Because there is no question.
I found a fantastic infographic about 2015 craft beer sales and production from the Brewers Association, which led me to a state-by-state listing of breweries. The great state of Maine is home to 56 microbreweries and brewpubs – including Black Bear Brewery in Orono, which brews my favorite beer of all time, the magnificent Pail Ale (the featured image of this post) – while a whopping 557 lay claim to California addresses. Data? We’ve got it.
Since the breweries are listed by state like this:
and the data visualization is county-based, I created a spreadsheet listing the town of each microbrewery and brewpub in both Maine and California (on separate worksheets) in one column, and the corresponding county in the next.
4. Then I tallied the numbers to see how many microbreweries and brewpubs are in each county:
Yes, there are really 113 microbreweries and brewpubs in San Diego County, and there are 61 in the city of San Diego. Bottoms up!
5. Then I created a range for each state so I could eventually align the range with a corresponding color. If there is not enough color contrast it can be difficult to distinguish color in a data visualization, so I wanted a manageable range with no more than seven categories. In California the range looked like this:
With far fewer microbreweries and brewpubs than California – 56 to 557 – (even though Maine has a higher per capita ratio of microbrews and brewpubs per capita), Maine needed it’s own range.
6. Now it’s time for color. On Robin’s suggestion in the Beginner’s Guide to Displaying Data post, I looked to www.colorbrewer.org for some color inspiration and, while the color suggestions are very good and would work wonderfully for 95% of my work, they just didn’t speak to me for this data viz.
Enter this COLOURlovers blog post called, “Color Inspiration from Ales, Lagers, & Stouts: Beer!” Now we’re talking. I grabbed a selection of beer-inspired colors from several palettes and assigned each to a range within my data set, using the lightest colors to indicate fewest number of microbreweries and brewpubs, and the darkest to indicate highest concentrations.
7. Then I colored the squares in my Maine and California maps accordingly, added a legend and a title and voila!
Comparing a single school, sales, health, or indicators at the census, city, or county level? This approach may be for you.
I’ve got a few more beer-based data sets, so we’ll come back for another round of data “imagining” soon. My treat.
Thanks to my family members (kids included) for accompanying me on the Black Bear Brewery photo shoot. It’s a tough job but someone’s gotta do it.