Hi,
I need to create one chart which is both stacked and clustered. For each year I need to have 2 clustered bars (target and projection) and each bar should be divided by media. Does anyone know how to do it?
(http://images75.fotosik.pl/591/27d831b4cb3f1311med.png)
This isn't a direct answer to your question, but is there a better way to display your information? It looks like you're trying to cram as much as possible into the graph without necessarily working out what you're trying to portray. So for example, in the middle group, I can't easily compare the Projection for Media 1 to the Target for Media 1. if I was going to be using that graph I'd find it very frustrating...but your audience might be different.
I'm with Chris on this one. A low of the arguments in this article (https://www.perceptualedge.com/blog/?p=2239) can be made for stacked bars in general.
If I were you I'd consider either using small multiples or a treemap.
Thanks for your pieces of advice. Since I didn't find the solution I'm going to prepare 2 separate stacked column charts - one for targets and one projections.
Here is another question. Is it possible to have the same axis range for both charts? In this way the user will be able to compare the height of the columns. Of course I can't set the minimum and maximum because this is highly dynamic and depends on the data. So sometimes it might be 0-20 but for another set of data it might be 0-1500.
The point is to have 2 charts with the same y-axis range.