I thank everyone for participating in the poll. It helped me understand which features are most important from the user perspective. I have been working on such visualizations for a while. I leave here the github link and the example:
I would be grateful if you file any bug report you come accross.
Thanks!
Jason Knight
The less code you use, the less there is to break
Adjustable bin size all the way, since depending on how you group things you can end up lying to yourself.
Generally I distrust visualizations on the grounds they are used so often to blatantly lie or to justify false premises. This is particularly common when the overall pool size of each bin or even the entire data point changes over time. The order in which multiple data sets can be 'stacked' in histograms is also often used to lie, exaggerating the change of whatever is on top.
Whilst it serves a use, too often it's just BS for BS' sake, taking advantage of the common person's predilection for what they see over what they would think of actual data -- hence why marketing scammers cream their panties whenever you even mention the idea.