Create a good scale for color agents in TARGIT

This dashboard was created for our maps visualization contest and it contains a lot of tips and tricks that really can help you in your daily work creating dashboards with TARGIT.

This tip - based on this dashboard -  is focusing on color scales and maps.


Author of dashboard: Michael Overgaard from TARGIT partner KMD.

Notice the colors on the big map in the right hand side (in the red square).
The dots are colored based on a color scale like this:


If you look at the numbers behind the map (just the first rows in the table) you will be able to se a problem coming up:


Since Los Angeles is such an outlier, we can expect Los Angeles to become dark red and the rest of the datapoints to become impossible to distinguish – because the outlier will “ruin” the scale.

To get around that our KMD consultant did a calculation compressing the range of values in a clever way.

I guess that calculation could be done in many ways, but in this particular case it looks like this:

Sum(d-1, 0, m1)^0.15

(The ^  symbol means power)

Let’s skip the math lesson and just accept that when you have a decimal number as the exponent, you end up with a sort of root of the number – and the numbers will basically be much more compressed.

Like I said, there are lots of ways of achieving this, but this one works very fine – the result looks like this:


If you want to make absolutely sure that your color scale is utilized, you can "rig" the scale - min and max values - and this particular user did just that:


Now you have a scale that is much better for showing the nuances of each datapoint.

I recommend that you also see this video for further explanations on color scales with special attention to heat maps.

Also the author of this dashboard created another stunning dashboard on the subject of Covid-19 (with updated data). 

Notice - there are 2 more tips based on this dashboard:

Conditional labels to show outliers

Storytelling with enhanced focus points

Was this article helpful?
0 out of 0 found this helpful



Please sign in to leave a comment.