"New, better alternative to Pie Charts: Treemap" this is a title I meet a lot in many blog posts and articles about data visualization.
Is it indeed true? Let's find out!
As a starting point, let's look at the "bad" guy - the pie chart.
What stories are being told by this graph?
It is commonly used to rank parts of a whole according to their relative dominance.
What makes it the first choice for most of us?
The circular shape makes us perceive it as complete. It is also easy to view the segments as constituent parts.
Circles are also aesthetically pleasing, so those who strive for visually appealing representations often utilize these graphs.
Now let's try to read a pie chart and understand its data relationships.
What if you had to answer the following questions based on the provided chart:
Which race had the most deaths? Which of the following is second to last? What percentage of the total killed were white people?
That is indeed a challenging task, isn't it?
The main challenge arises from the difficulty in comparing the areas of the segments within the chart.
Consequently, pie charts and similar graphs become challenging to interpret.
Colors for categories don't make it easy either since the brain tries to figure out how the color relates to the category, which in the example here just happens to be a random choice.
Let's see if the Treemap helps and make it easier. Here is a graph obtained from 'Statista' presenting the same data.
The challenge of comparing areas persists, and in this case, the only "advantage" is that the areas are represented as squares.
Colors issue is the same, as with the pie chart
So, what is the alternative?
We can compare the magnitude of individuals killed by the police in relation to the total using a straightforward percentage-based ranking graph, such as a bar chart. By comparing the lengths of the bars and arranging them in descending order from largest to smallest (or ascending order, depending on the desired narrative), we can effectively assess the relative proportions.
In an interactive report, such as the one created in Power BI, tooltips can display the exact values, offering additional supporting information.
Getting back to TreeMap, what are the scenarios in which it can be used?
If you want to understand the hierarchical relationship between groups and subgroups, particularly when there are multiple subgroups within various groups, presenting the overall picture becomes important without requiring precise data size comparisons. For example, consider a graph illustrating the population by continent in 2012, further segmented by country. Asia stands out as the largest continent in terms of population, with China and India leading the countries. However, if you wish to assess the magnitude of the gap between Europe and the Middle East, you would need to refer to additional graphs.
Link to download my PBIX with Examples: https://github.com/MadeiraData/MadeiraToolbox/blob/master/Power%20bi/Treemap_Alternative.pbix
Sources of other examples: https://www.statista.com/chart/5211/us-citizens-killed-by-police-2016/