Open-Sourcing Our Analysis

blog.modeanalytics.com

24 points·Nonnormalizable··4 comments

I'm not a particular fan of pie charts or donut charts, but the visualization used in this tool is actually helpful at understanding and comparing nested amounts.

The reason I don't usually like pie charts or donut charts is because it is difficult to make comparisons between circular shapes. In most places people use a pie chart, I think a bar chart would be better. It's easier to compare different lengths of rectangles than guesstimate differences of pieces of pies or half circles.

Something about this type of visualization with nested depths seems to be an outlier. If this information was portrayed in rectangles I'm not sure it would be better.

Yeah, I generally agree about pie and donut charts. I thought about unrolling this one, so that the inner most circle would just be one long bar, and each successive step would just be an adjacent bar. I'm not sure that makes it better (and might make it worse, since the whole thing might end up being really tall).

There might be a clever way to do this with a tree map, but it wouldn't be particularly intuitive to read.

As a fairly new "data scientist," one of the ways I've been constantly jealous of my Real Programer coworkers is that there's a ton of code out there, open sourced, on how to build websites or whatever. But comparatively little on doing analysis (and that mostly data viz). Here are steps towards fixing the lack!

I'm a big fan of using this type of visualization to add an extra dimension in a digestible way. Adding the third dimension is always the toughest one.

Reply on news.ycombinator.com