Update x,y.z values for an existing Line3D object #1629
Conversation
spirit as set_xdata and set_ydata for Line2D, but requires _all_ of the data to be set in one go (to not have to write the caching layer of Line2D again yet).
x,y,z can be specified.
| @@ -112,6 +117,21 @@ def set_3d_properties(self, zs=0, zdir='z'): | |||
| pass | |||
| self._verts3d = juggle_axes(xs, ys, zs, zdir) | |||
|
|
|||
| def set_3d_data(self, xs=None, ys=None, zs=None, zdir='z'): | |||
| x_old,y_old,z_old = self._verts3d | |||
There was a problem hiding this comment.
For PEP8 compliance, could you put spaces after the commas in this tuple?
|
Those Travis failures are false negatives. |
|
I raised an issue related to this in #1483. This is certainly a step in the right direction. |
|
Sorry about my poor style, pep8 is not ingrained in my brain yet. |
No worries. That's what we're here for. |
|
Question for the masses: This is debatable, but do we think this is enough like a bugfix to backport to the v1.2.x branch? |
|
Actually, I think there are enough problems with this code that I want to hold off on this. Plus, I have uncovered some other mplot3d bugs recently which solving may conflict with this one (it has to do with juggling axes). I would like to close this PR for now, but flag it for re-examination later. If others disagree, we can always re-open. |
|
@WeatherGod by this code do you mean mplot3d or this particular patch? If it's the patch, what should I have done better? |
|
The patch. You should never do bare exceptions (they hide RuntimeExceptions, among others). Also, the list comprehension to build the zs could be done better (maybe [z] * len(xs)), and some other smaller issues. But, really, the thing that I see as blocking this is another bug that I am currently investigating on my own. My current thoughts on how to deal with it may conflict with your patch. So, if we can hold off for a week while I figure this other thing out, we can re-examine this one then. |
|
ok. I copied both of those things (exception, list comprehension) from the existing [I am trying really hard to to sound defensive and spent far too long writing this note] |
|
Yes, the current mplot3d code is not PEP8-compliant at all, and really shouldn't be considered an example for others to follow. That is another upcoming project. We are currently trying to make sure that all new code going in is PEP8 compliant. Also, a slight issue I see. If one were to directly create the Line3D object, wouldn't this break because the self._verts3D wouldn't be created yet? |
|
You are correct about breaking |
There does not seem to be a clean way to update the x,y,z data for a
Line3Dobject.Added set_3d_data(x,y,z,zdir) which will update the data for an existing line.
added support for 'zdir` to init