Today, in the Webinar, I got aware of the Density plot, which works great for classification problems. For regression problems with many points, the appropriate density function in many cases would actually be 'number of data points' (in a prediction vs. label plot at least...) - I could see this as the default behavior if 'Density Color = none'.
Additionally - and specifically due to the autoscale done on the axis - it would help if a diagnal (y=x) could be shown on scatter and related plots.
Stefan
thank you for your suggestions. The plotter's are on my list anyway, so you might add this request as a feature request to the bugtracker. There's a field that let's you specify if you are submitting a bug or request.
Greetings,
Sebastian