It’s inevitable that as soon as I write a post bemoaning a lack of perceived #MakeoverMonday inspiration this week, that someone steps up to the mark and creates something really nice. This week, it was the fun half of The Data Duo, Pooja Gandhi, who brought a smile to my face. Here’s her viz, and it looks like this:

Screen Shot 2017-04-11 at 19.01.50.png

I just love it. It isn’t the most complicated chart in the world, but once again Pooja’s design flair and use of a few dashboard tricks elevates her work to another level. Let’s explore a few key pointers, starting with the main chart:

Screen Shot 2017-04-11 at 19.04.18.png

It’s “just” a dual-axis chart (which I summarised in a post linked at the top of the page). The difference is that Pooja has whacked two dual-axis charts on Rows – one for Gold and one for Oil. This approach meant that two separate y-axes were drawn, and so separate Reference Lines could be added to indicate the average price of Gold and Oil with the dotted horizontal lines.

It also meant that each axis could be shaded in a way to tie the colour through to that used in the chart itself (it ties directly to the font colour and the colouring of the line element of each dual-axis chart):

Screen Shot 2017-04-11 at 19.14.34.png

It’s not a huge tip, but one of those incremental little tricks that add up over the course of time. I haven’t seen this particular Axis shading applied many times before, and I think it’s really effective in this viz.

On the left of the dashboard, there is a vertical column which dynamically updates when viewers hover over the charts:

Screen Shot 2017-04-11 at 19.18.59.png

That’s achieved by creating three additional worksheets, summarising the Gold Price vs. the 32 year average, the Oil Price vs. the 32 year average, and the Con(s)umer Price Index (typo Pooja!) The Gold and Oil sheets are the more complicated ones, so let’s look at the former:

Screen Shot 2017-04-11 at 19.22.37.png

And the relevant calculations:

Screen Shot 2017-04-11 at 19.23.16.png

What is [Gold Avg]? It’s this:

Screen Shot 2017-04-11 at 19.23.49

A cheeky little Level of Detail calculation. “The” main reference source to get your head round LoD is this. In this case, the absence of Dimension(s) in the curly brackets just indicates that the overall average of the Measure is being calculated. Nice and efficient.

Finally, a basic logic calculation to determine whether to stick “above” or “below” in the label text string:

Screen Shot 2017-04-11 at 19.27.01

These calcs combine to form the dynamic summary which forms part of the vertical “information pillar”.

Screen Shot 2017-04-11 at 19.29.27

To ensure dynamism, a dashboard filter action is applied:

Screen Shot 2017-04-11 at 19.34.30.png

So hovering over the charts updates the three worksheets which comprise the “information pillar”. That’s not all though – there is a second dashboard highlight action:

Screen Shot 2017-04-11 at 19.35.46.png

What’s that? It’s a new trick, for me at least. On the chart worksheet, I can unhide the MONTH(Date) Header to see what has been done with Reference Lines:

Screen Shot 2017-04-11 at 19.42.23.png

At a worksheet level, Pooja hadn’t set an action up, so I guess that’s why she instead configured it at the level of the dashboard. I set the action up on the worksheet itself, deleted the dashboard action for the same, and it all still appeared to work OK. The need for the action itself is that the vertical dotted Reference Line and its labels wouldn’t appear when you hover over the chart otherwise:

Screen Shot 2017-04-11 at 19.45.17

The label itself is on the Area Marks card of each dual-axis:

Screen Shot 2017-04-11 at 19.45.56

And an important thing to note is that the label is set to show when highlighted.

Screen Shot 2017-04-12 at 19.17.17.png

Pooja later confirmed that this specific tip was picked up from an article written by @tableautim. Tim took the time to share a video about the technique:

That’s it at a worksheet level. Time to focus on the dashboard, which is a Float-a-thon. I took two really neat tricks away from this, and they are tricks I have seen more and more of lately, so it’s well worth sharing them.

  1. Dotted line text boxes

One of these:

Screen Shot 2017-04-11 at 19.53.13

It’s purely aesthetic and it’s a little bit hacky. I do think Tableau could really kick on if basic formatting options available in products like Microsoft PowerPoint could be incorporated, to avoid the over-engineering required to achieve this look. First step is a basic blank text box, formatted to have a dotted border:

Screen Shot 2017-04-11 at 19.55.48.png

Over the top of that, you float another text box, this time populated with the required text and shaded. If it is sized correctly, then you cut through the dotted border of the “underneath” text box and break the border up:

Screen Shot 2017-04-11 at 19.59.43

2. “Connector lines”

These grey lines:

Screen Shot 2017-04-11 at 20.00.56

Another scenario where basic PowerPoint functionality would make life easier. The concept is that the lines tie elements of the viz together. To do this Pooja uses similar principles to the first bulleted tip. The lines themselves are simply shaded test boxes, made really thin:

Screen Shot 2017-04-11 at 20.03.54

When aligned at right-angles to create the “elbow”, a further shaded text box is floated over the line to complete the full effect:

Screen Shot 2017-04-11 at 20.05.02

These are simple and effective tips to implement. I find myself unearthing more of these little beauties the more I investigate dashboards, and it’s getting to the point where I really need to step up and start implementing them in my own work.

If you look back at everything in this dashboard, it’s actually quite straightforward. Definitely not “beginner” stuff but arguably at an “intermediate” level. Pooja has the benefit of an eye for design, knowledge of these little Tableau “hacks”, and experience of combining both those skillsets to create beautiful dashboards. However, hopefully this post serves to highlight that this polished finish is achievable once you have awareness of the little tricks needed to push your visualisations to the next level.

 

 

Advertisements