Lets pretend this is the data:
MusicGenre AgeGroup TimesPlay
Rock 10-15 100
Rock 16-20 200
Classical 10-15 5
Classical 16-20 4
If I do not group the data and I plot it
1. Then Rock 16-20 is the biggest bubble (biggest)
2. Followed by Rock 10-15 (2nd biggest)
3. Classical 10-15 (tiny)
4. Classical 16-20 (tiny)
But I get the same color for each bubble
If I group by Genre
I want to have the same color for Rock and a different one for classical and that happens, but then the bubble size
is determined by items in the same group, not compared with the rest
So Classical 10-15 and 16-20 now are big
Classical 10-15 = 5 out of 9 instead of 5 out of 309
Classical 16-20 = 4 out of 9 instead of 4 out of 309
Both are rendered big, because the grouping is taking in consideration only items inside the group and not comparing with other groups
I want to group to have different colors, but then the purpose of the bubble chart gets defeated, because having similar sizes doesn't make an useful visualization.
Is this a bug or it is intended to work this way?
Any help? Is there a way to have groups but the bubble sizes are based on all the items, not only on the ones inside the group?