I've struggled recently with the colors of bubbles in bubble plots. I know that I can change it by using the Column Info in the data set and changing the Value Color there, but why can't I change the color directly in the Bubble Plot by using the 'Customize' menu?
When I use the Customize menu, the new color I select is not reflected in the bubble plot, but it does appear in the Script Window, and even when I relaunch the script the color does not change. Does the Fill Color in the Customize window refer to the color of the bubbles or to something else?
In general a bubble plot is displaying an average value rather than a single data point. Therefore the Value Color property does not get applied. If you assign a column to the Color role, then a blue-to-red colour gradient will be applied. To my knowledge, you are not able to change the color scheme of this gradient.
The gradient itself appears to be calibrated against the actual data range for the column, but the bubble plot is displaying average values. Therefore you may have a situation where the plot never shows the full potential colour range e.g. the raw data may go from 20 to 100 and the colour scale will reflect this, but the average values will only span 30 to 70, so you dont see the colour differentiation that you are looking for.
I think this would be a good feature request. At my company, we use a data visualization application called Tableau that permits you not only to choose the endpoints and the order of the color gradation, but also allows you to choose the specific colors in the range (and it can be more than two!).
This can be very useful if you're trying to visually denote a value judgment; for example, you may want your color range to go from red to yellow to green, and you may need to set which end of the continuum should start with red.
Tableau doesn't come close to JMP in terms of statistical power (and doesn't really try to, as it's not the intent of the application), but it has some graphical capabilities that would be very welcome in JMP. In this particular instance, I think the potential value to JMP is quite clear.