The first part of this duo of posts focused on three of the viz’s I downloaded this week. Time to start tackling the other ones now.

Lesson Four: Matt Francis – Making data points stand out

I just liked Matt’s viz. It’s simple, well laid out, and nicely highlights DR Congo as the focus of the visualisation. The are no complications here or anything I need to really note apart from the effectiveness of the simplicity. The colouring is manually set, so again my preconceived expectations of complex calculations behind everything go unfounded!

Lesson Five: Josh Tapley – Making data points stand out

As with Matt’s example, I just really liked the simplicity of Josh’s design. In this instance, I just wanted to validate my approach to creating the sort of sparkline-esque effect of this worksheet within the viz:

Screen Shot 2016-08-24 at 20.06.34

Realistically, it’s nothing more than “nice”, but I really like crisp, clean and minimal charts, so I thought I’d knock it up myself.

To do so, I first dragged Country to the filter shelf to limit the selection to those depicted above. I know from Josh’s various banners and titles that he’s looking at the Top 5 countries throughout the duration of the dataset, so the easiest way to filter the countries is this:

Screen Shot 2016-08-24 at 20.12.27

OK. So now we have the foundation of the content of the worksheet. Now we just need to drag the relevant Measures and Dimensions to the relevant row / column shelves. First up, we can deduce from the chart that time is on the x-axis, and the SUM(Deaths) is the y-axis.

Let’s whack them in those places, and set the chart type to Line. We also know that each of the five countries has a dedicated section, and that it divides the view into five separated columns. So, obviously – we need to put Country on columns. This happens:

Screen Shot 2016-08-24 at 20.20.01

It’s a good start. Things to do are to sort the countries in descending order by SUM(Deaths), and then strip away superfluous formatting. Note that the 4 nulls just indicate sparsity in the Kenyan data which is reflected in the gap in the line for that country.

First, let’s sort the sort! It was just a case of right-clicking the Country Dimension on the filter shelf, and setting the correct sort criteria:

Screen Shot 2016-08-24 at 20.14.36

The formatting is a simple case of:

  • Remove Year header
  • Colour line to green
  • Hide field name for columns
  • Remove y-axis header
  • Clean up gridlines
  • Tidy up row and column dividers
  • Drag the Country column down a touch to create space for DR Congo to fit
  • Add labels to line ends and manually reposition them

I didn’t get it 100%, but it’s close enough:

Screen Shot 2016-08-24 at 20.33.55

Lesson Six: Mike Moore – Colouring an entire country with a the same colour

I’m easily pleased! Small things stand out for me, and I really liked how Mike’s viz had the map of Africa all with a uniform black. It looked simple, and is simple, but I wanted to check it out to confirm the methodology.

Screen Shot 2016-08-24 at 20.38.27

How is it done? Beforehand, I expected it to either be a Set of African countries to isolate the continent with a uniform colours applied. Or, I wondered if it would make use of the custom territories feature in Tableau 10. Before opening the sheets behind Mike’s viz, I attempted to use the first of my assumptions:

Screen Shot 2016-08-24 at 20.44.11

Why is it different? Well, after applying a 100% washout to the map, unticking all Map Layers and setting the colour to black, it’s pretty obvious that the we have data sparsity as the entire map isn’t filled. I also couldn’t remove country borders for some reason. So what has Mike done? It’s just a floating image!

Screen Shot 2016-08-24 at 20.47.53

Advertisements