This might also be useful to new users (since there is no documentation yet).
The colors used in the visualisation indicate where that particular data point originates from in the original file/data. Things from the begining of the selection tend to be yellow and those from the end blue.
Don't bother looking for the histogram generation algorithms. There used to be one in C++ when the digram view was seperated but that is no more. The histograms aren't computed, we rely on the graphics for that. It's cleaner, nicer, and less work. We just draw all the points, if there happen to be a lot it will be more visible.