In my blog I’ve talked a lot about the importance of communicating data using visualization and storytelling. In fact I have an entire “Portfolio” section on my blog with visualizations in addition to my GitHub “DataVizMisc” folder. This summer I took three intensive courses to fill in existing gaps in my data science graduate education at Indiana University – Tableau, mySQL and R. Today I’m going to delve into what I learned in the first part of my Tableau intensive.
You might be thinking, “Gee, you’re an aspiring data scientist with one machine learning conference speaking engagement behind you and you are taking a Tableau short course? Fair question. To give you a brief background, the first time I heard about Tableau was at a large conference in DC in 2015. The audience was mostly government employees, but what happened next will be forever ingrained in my first impressions of Tableau.
I thought two of the slides their marketing group put together were quite inappropriate and borderline offensive and gave this feedback to conference organizers. Even though I knew Tableau was a useful business intelligence and data analysis tool, I put off learning it intentionally for a bit. If you’re dying to see the slides from the November 2015 conference, here they are and here.
Well, back to the summer of 2017 and one of my first assignments was exploring sales data set. The first assignment was trying to see which product sub-category had the highest sales in a year. As you can see from the simple chart below, ‘Office Machines’ sold almost 2,220,000 and well exceeded the other Product Subcategories.
Another practice visualization looked at Income based on Census data from 1990. I wanted to see how many hours per week yielded the highest salary in England using Tableau. So I created the following visualization and it turns out that the highest salary was ($US 312,897) at 46 hours per week and the lowest salaries was $19,302 at 56 hours per week.
Next week we’ll look at answering questions using more sophisticated data visualizations in Tableau.