Currently, the number of bins to build a histogram is a static parameter (independent of the input dataset). But methods like the Freedman-Diaconis rule exist to allow setting the bin-width based on different datasets.

Of course, such automatic methods have to be used with great caution because usually they fail to distinguish between important differences in the data (for example see Anscombe's_quartet. But having this method within the Statistics program can be useful in general (at least to check).

This was suggested by Zahra Hosseini-Shahisavandi.