Custom Parallel Coordinate

I saw this cool parallel coordinate chart in the Guardian on the heptathlon in the Rio Olympics. Displaying multi-variate data is tricky. A traditional parallel coordinate chart is some times difficult to read with the high number of overlapping lines. This new approach, which I have re--created with 2017 Madden ratings data in Tableau, allows you to see every point on parallel axes and then view the relative position of each point for a selected player. The correlation between variables is not as immediately apparent but it is easier to see a specific data point across variables.

Dear Data at MICA

I teach a Visual Analytics course at MICA as part of their Information Visualization (MPS) program. It's a great course and this is my second year teaching it. Last year I gave the students an assignment to recreate a Dear Data postcard and it was very successful. So this year I did it again and I thought would share my favorite post cards from two of my students.

The first one was created by Alex Schroeder and replicated the Week One post card from Dear Data. I think it's a very creative approach where the number of concentric circles represent the number of time's he checked a clock. The layout by hour of the day and day of the week makes it easy to see patterns. He even adds a reference line to indicate when he woke up each day (6:00 am). See Alex's post card below.

Starry Night Color Composition

This visualization looks at the color choices by Van Gogh in his famous paintings "The Starry Night" and "The Night Cafe". Using the EBImage package in R I converted every pixel of the painting into an RGB value. Then I imported the X & Y vectors of every point of the painting and the point's RGB values into Tableau. Then using the RGB color palette created by Merlijn Buit I visualized every vector in a scatter plot and mapped each RGB value to a specific color.

Beatles Sentiment Analysis

This analysis looks at the sentiment of all Beatles songs by songwriter similar to the Beatles analysis I did a few months ago. For this analysis I analyzed the lyrics of every Beatles song using the sentiment package in R which classified the Bayesian polarity of every song.

How To: Game of Thrones Analysis Viz

I don't usually write "how to" posts because me no write so good and I would rather spend 20 hours making a viz than 1 hour writing about it. But after reading Chris Love's great post on Tableau "edge-cases" I decided that it might be helpful to explain how I created the Game of Thrones viz which is somewhat of an "edge-case". So hopefully this is helpful.

why even make this? get a hobby
Usually the reason I create a viz is because I want to tell a cool story with an interesting dataset. For me a viz is usually a bit of a chicken or the egg type of situation as in which comes first the data or the story? Do you have a cool data set and find the good story in there; like in this Trump tweets viz? Or do you have an interesting hypothesis and find the data; like in this White House gender pay gap viz? 

But sometimes I just want to make some stupid thing that looks cool. This is why I made the GOT viz. Actually, I drew this image below one day on the metro and then thought how could I make this into a viz:

how did you make this? seriously try golf
I have recently noticed that designers that I admire tend to work with simple shapes like circles and triangles. And I had done some custom circle stuff in Tableau before. But these were all pretty straightforward. The trick of this viz is the little doohickey, flipped up circle thing at the top left quadrant. It actually kind of looks like the profile of Wilson the volleyball, Tom Hanks' love interest in Cast Away.