Week 1: TIL Data School – You’re Hired

by Hashu Shenkar

After a tough and gruelling application process to get the job I had imagined the first week to be pretty relaxed. That’s how it usually works right? Quite the opposite!

Tuesday

We had our first day off so The Data School officially launched on Tuesday – you can see where I was going with my theory. In the morning I met the other seven amazing students, we were briefed on the School’s vision and had our IT inductions. Upon completion, we were asked to close our laptop lids for the day as we were going to spend the rest of the day working by hand. A moment of shock went by. By hand?! I had known nothing other than working on a computer everyday of my working life so this took me by surprise. It was reminiscent of an Indian drama scene where the camera points to everyone’s face reaction when something dramatic happens from every angle possible. Huh? By hand? You what?! What did he just say?

Be careful for what you wish for, it might come true. I recently spoke with a friend about how I wished I could go back to being a child, low and behold, we were handed jumbo crayons and A2 sheets of paper for our first class on Data Visualisation best practices, and, it actually felt like I was back at school! We worked in pairs to makeover existing visualisations. Throughout and after the exercises Andy talked us through a series of data visualisation best practices. Though challenging, I soon realised the benefit of the class. My mind was freed from the boundaries it was confined to when using a computer and was exposed to a previously unknown landscape of creativity. An excellent brain-stimulating day and I foresee myself putting pen to paper more often in the future.

IMG_5354

The evening provided us students with the chance to mingle with the consultants and few of The Information Lab’s clients. Pleased to have met my mentor Andrew Pick who encouraged me and provided me with some insight into life as a consultant. Thanks Andy! Whoa this is going to be confusing, both my mentor and coach are named Andy!

Wednesday

From gear 1 straight to gear 5. We joined the experienced consultants to attend a Tableau Zen class delivered by Tableau Zen Masters Andy Kriebel and Craig Bloodworth. The dynamic duo gave us some neat tips and tricks on Tableau having pushed it to its limits. The class was mind-boggling at times but I was proud to have replicated a few of the challenges, with some help from Robin Kennedy. A few of the features that were new to me included: Pareto charts, LOD calculations and using the INDEX() function for conditional formatting.

A fascinating bit of advice Craig gave us was to be careful with maps as they can be used to tell a particular story. He told us a story about how someone had created a visualisation showing the population of the United States on its map if it was made up of people who have died from HIV/AIDs in Africa. The author had filled it from left to right and roughly two thirds of the map was covered. Interestingly if filled from right to left, roughly only one third of the map will be covered so another author could easily tell a different story.

Chris Love wrapped up the day by introducing us to Alteryx and running through its powerful capabilities. From a Tableau user standpoint, one of the key uses of the software is for data preparation, getting data in the right format for Tableau to consume, which can be a very tedious process. I speak first hand after spending about four hours restructuring a poorly structured dataset for the first part of the application process for the job, which was to create some interesting content in Tableau public using data openly available on the Internet.

Thursday

Back to basics, Andy taught us the official terminology of objects on the Tableau interface then went through basics SQL queries, groups, hierarchies, folders and references lines. Next we had “Tableau Qualification Exam Prep” in our diary. Surprise surprise, gear 1 straight to gear 5 again! A pattern was emerging here. Instead we were given a Tableau Qualified Associate Practice Exam to take – certainly wouldn’t want to be a passenger in Andy’s car! The exam was challenging yet enjoyable and took the whole two hours allotted for me to complete. On a positive note the exam helped highlight areas of weakness, useful for the real exam at the Tableau London Conference, which is just seven days away!

In the afternoon it was time to continue working on our weekly projects. This week’s was to makeover a visualisation created by The Information Lab team. I was drawn a visualisation on Tea Party Nationalism Andy created in 2010. I had a number of challenges to overcome to make the deadline of, you guessed it, the next day! Fortunately Tom Brown rescued the day for me by effortlessly assisting me with the most difficult challenge I was bogged down with and couldn’t find a solution for online after extensive research. Dynamic sorting of the results of a parameter and displaying the marked labels in the correct format: one was an integer and one was a percentage.

Friday

As the roller-coaster week drew to an end, I expected one more oscillation as it was our project presentation day (wasn’t looking forward to it at this point) and two guest speakers were lined up – woohoo. Andy pulled off a massive coup in getting Tableau CEO Christian Chabot to come to talk to us. His speaking skills were just phenomenal; I was engaged for every moment he was with us. Everything from his charisma to the way he constructed and articulated his thoughts to tell stories made him an exceptional speaker, by far the best I have seen in person. Subsequently we were given a final chance to work on our projects for the looming presentation in the afternoon. In between Stu Wilson, Vice President for Alteryx EMEA kindly gave us a talk on how Alteryx fits in to the whole data analysis process. I learned a lot from him including how much time and money businesses can save by streamlining typical data analysis processes; an indispensable product for data analysts!

Presentation time finally arrived and the pressure was on. Stu stayed to watch along with Tom and of course Andy was there too. A number of the consultants also joined in via a web broadcast. We picked our numbers out of the hat for the order of the presentations and I got 7, my lucky number!

Tea Party Nationalism – Click here to visit my Tableau Public Profile and see my Viz!

Here is the original Viz Andy created. It shows the population of Tea Party members in 2010.

original

I liked that it was pretty straightforward to navigate through, had a clean design, the size of the map instantly grabs the reader’s attention and the red filter titles stood out.

It was good, however when looking at it I thought to myself okay this is great but what insight is this providing? It was evident that ResistNet was the most prevalent faction and I could navigate though the filters to see how this changed but I wanted to see some numbers. How many more members did ResistNet have compared to the Tea Party Patriots? Which cities had the highest number of members per state and how did these compare? You would have to hover over each city on the map to see the numbers on the tool tip but then you couldn’t compare it with the other cities.

Here is my makeover.

makeover

Since there were only five factions, I thought a tree map would be the best chart type to show the numbers and proportions to make comparisons. I made use of horizontal bar charts for state and city sorted by number of members to allow for comparison. I also used actions so that any change to the dashboard would filter through to every view i.e. if you click on ResisNet, you would only see the corresponding colours on the map and number of ResisNet members in the respective states and cities.

Once done I still thought more insight could be driven from this Viz. After some deep thinking the light bulb lit. I thought of downloading population data for 2010 so I could represent the number of Tea Party members in each state and city as a percentage of the total population, this would allow for a fair comparison between smaller and larger states and cities. I managed to do this for states but ran into some data blending issues for cities due to the way cities were named in the dataset so couldn’t sort it out in time. Nonetheless the presentation went well, I was proud of my work and really enjoyed the makeover!

1 down another 17 to go! An amazing week, lovely people, fantastic opportunities and though it’s been challenging, that’s where the growth takes place!