Björn Rabenstein

I'm playing with this now. It looks generally very nice, but there seem to be some quirks. Here my observations (really just as an outside observer, without any insight into...

> I never had the expectation that the area under each bar in the exponential view had to be adjusted to be proportional to the number of observations and was...

https://en.wikipedia.org/wiki/Histogram#/media/File:Travel_time_histogram_total_n_Stata.png is the proverbial histogram from Wikipedia . Note the unit on the y axis (people per time interval).

Sorry for the delay in my response. Let's clarify the scaling of the height of the buckets first: I think the linear case is pretty clear, and your implementation might...

Another thought: You could also do the multiplication thing in the exponential view "strictly correct" right now (I assume you have to change the current code just slightly to take...

Height looks good now. The only thing left for exponential view is to position the bars correctly. I try to summarize our discussion we had on Slack: I don't know...

Note that the above always puts the zero bucket directly adjacent to the highest negative bucket and the lowest positive bucket, no matter if there is a gap in between....

Things look quite good. So this is a perfect example, as far as I can see:  The zero bucket is unpopulated here, but rendered correctly as an empty bucket...

> In this last case, are they truly of different width or are they overlapping? They are truly different width. Buckets are never really overlapping. > Also looks like the...

Another idea for `findClosestToZero`: Let it just find the zero bucket (the one that goes from negative to positive), and return its lower and upper limit. If there is no...