set_scale and set_xscale

[gsever]

These two method calls should perform the same log scaling action. However, the output from xaxis.set_scale('log') behaves differently. See the attached figures for the difference. I am guessing yaxis counterparts behave the same way.

fig = plt.figure(figsize=(5, 5))
ax1 = fig.add_subplot(1,1,1)
ax1.plot(np.random.randn(100))

ax1.xaxis.set_scale('log')
ax1.set_xscale('log')

plt.matplotlib.__version__ 
'1.3.x'

[pelson]

These two method calls should perform the same log scaling action.

I'm not so sure. If they're supposed to do the same thing, why bother having two of them?

For my own reference, the two relevant bits in the documentation: http://matplotlib.org/api/axes_api.html#matplotlib.axes.Axes.set_xscale and http://matplotlib.org/api/axis_api.html#matplotlib.axis.Axis.set_scale (set_scale is not documented)

Any comments @mdboom?

[tacaswell]

The axis.Axis.set_scale is used by Axes.set_xscale. It looks like set_scale should really be an internal function.

set_xscale calls self.xaxis.set_scale and then self._update_transscale which then gets the scale back from the x and y axis objects and uses the results to update the transform the objects in self.lines use to draw them selves.

[mdboom]

Yeah -- I think Axis.set_scale really should be private. Of course, some may be using it, so we'll have to retain a public wrapper around it that throws a deprecation warning for a while.