What can you do in 60 mins? (i) Learn new Tableau feature(s). (ii) Perform data munging. (iii) Develop a Tableau dashboard. It might sound unbelievable but the answer is all of the above. At least that was what a group of approx. 40 people did at the Tableau User Group in Singapore yesterday!
For the session, we used Week 22's dataset on The world's most expensive prime property. It contains just two fields: City and Square meters. It's a small but informative dataset, and resonates so well with me, since we complained about how small our houses are as Singaporeans all the time, but well, this allows us to see if we're pot calling the kettle black.
I had previously seen a similar visualization (below) and really liked it (source). Yea honestly, it's exactly the same, because the visualizations are both developed by Statista. I thought the treemap chart was totally/ most appropriate for this data since we're talking about space and the rectangles reflect the floor sizing so well, it's like allocating land space in rectangular blocks.
But of course, there is always room to explore different ways of story-telling. And that is the spirit of Makeover Monday.
Given the one-hour, I was glad to find a new Tableau feature! Something I had always wanted to try: Cross-join. I wanted to compute the ratio of square meter in one country is to the others i.e. the purchasing power parity concept in terms of square meters instead of $. To be able to achieve that, I needed to "square" the data. For each city, I needed other cities to form a pair with it and get their corresponding information on square meters.
I found the easiest workaround after combing through several sites returned on my Google search "cross join tableau with itself". I created a dummy column with all "1" in the dataset. Somehow I'd conveniently skipped the advice in orange at the top. The text in orange reads: From release 10.2 it is possible to use dummy values 1 = 1 in Tableau's join dialog and thus not necessary to create dummy columns in data itself. Nonetheless, I was then able to do a cross-join and get the many-to-many relationship result I wanted! (We can now see two sheets under Data. I renamed the City in each sheet.) I could compute the ratio subsequently using the "Created a calculated field" under Analysis function.
I dragged Reference City to Filters and selected Singapore.
Before dragging SUM(Ratio) to Columns, I created a color tag and dragged it to Color:
IF([Reference City]=[All Cities]) THEN "Reference City" ELSE "All Cities" END
Here is my interactive visualization:
While this is not an intended objective of the Makeover Monday, we also got to see how Tableau allows us to develop decent, if not awesome, visualizations within a short span of time! It has been an inspiring experience to see what others can do in one hour.
Last but not least, this is my maiden participation in Makeover Monday. The one-week timeline format for a visualization has been intimidating me but who knows I could actually do the above in one hour? I forgot, it's all in the mind ;)
Believe you can (viz), and you can.
Datasets for other weeks of Makeover Monday can be found here.