Be more explicit about bins too narrow for float representation

[janpipek]

If the computed range for the binning divided by the number of bins is lower than the minimum float difference at the scale, we receive an error [ValueError: Bins not in rising order.] which is not very informative.

To reproduce:

data = [1, np.nextafter(1, 2)]
physt.h1(data)

It also happens when the range is 0, like in:

data = [1, 1]
physt.h1(data)

[janpipek]

There should be a way how to find bins even when there is just one unique value. Numpy does that.