In article <4CA0549D.3060408@...202...>,
>> Did you try autoscale_view method?
>> Please post a sample script that reproduces the problem.
> I'm wondering if I'm doing something wrong then. Just now I was
> writing some notes about this for a tutorial, and tried this code:
> line, = plt.plot([1,2,3], label='my data')
> plt.title('My title')
> x = np.linspace(0, 1)
> y = x**2
> line.set_data(x, y)
> ax = gca()
> but I get the result shown in the screenshot. Am I misusing
> autoscale_view? As best I can tell from the docstring, I'm making
> correct use of it, but perhaps I'm missing something...
autoscale_view somewhat defeats the purpose of line.set_data, which is
intentionally minimalist. If you want autoscaling after update, but
don't want to simply clear and plot, then you need to explicitly update
the Axes.dataLim. You can do this using Axes.relim()
prior to calling autoscale_view().
Thank you. That is exactly what I needed.
It looks like Axes.get_ybound() will return the min and max displayed y
so whenever I add a point I can test if it is within those bounds;
if it is not then I can call relim(). That should preserve the
efficiency most of the time while still allowing autoscaling
and avoiding my having to keep accurate track of what's going on.
Eric Firing <efiring@...202...> wrote:
On 09/26/2010 09:43 PM, Fernando Perez wrote:
> On Sun, Sep 26, 2010 at 10:56 PM, Jae-Joon > > Lee<lee.j.joon@...287...> wrote: