Where y bins stop: edges step by size from start until they reach or pass end. Default: the largest sample.
size?
unknown
Width of the y bins: a number, milliseconds or 'M<n>' (months) on date axes, a number of categories on category axes. Default: a nice round size from nbinsy or the samples.
start?
unknown
First y bin edge. Default: the smallest sample, shifted down to a nice round value (integers get edges at half-integers). Dates as date strings; categories by their serial number (default -0.5).
