Repository navigation
Matplotlib hatch color broken #8076
Description
This may have some similarities with #7992, but seems not identical to other bug reports I found.
Bug report
Bug summary
- It is no longer possible to set the color of hatching in patches. This still worked in 1.5.3-.
- In particular, it affects PathPatch for me, but seems a more general problem.
Code for reproduction
An example modified from http://matplotlib.org/users/path_tutorial.html is
import matplotlib.pyplot as plt
from matplotlib.path import Path
import matplotlib.patches as patches
verts = [
(0., 0.), # left, bottom
(0., 1.), # left, top
(1., 1.), # right, top
(1., 0.), # right, bottom
(0., 0.), # ignored
]
codes = [Path.MOVETO,
Path.LINETO,
Path.LINETO,
Path.LINETO,
Path.CLOSEPOLY,
]
path = Path(verts, codes)
fig = plt.figure()
ax = fig.add_subplot(111)
patch = patches.PathPatch(path, color='green', fill=False, hatch='xx', lw=2)
ax.add_patch(patch)
ax.set_xlim(-2,2)
ax.set_ylim(-2,2)
plt.show()Actual outcome
green rectangle filled with black cross hatching on white background.
Expected outcome
green rectangle filled with (more dense) green cross hatching on white background.
(Matplotlib 1.5.3-)
Matplotlib version
Python 3.6.0
IPython 5.2.2
Matplotlib 2.0.0
You should add a unit test for this.
Other examples
- An example of a similar problem can be found if you try to reproduce the plot at
http://stackoverflow.com/questions/38168948/how-to-decouple-hatch-and-edge-color-in-matplotlib
using their code.
Activity
dstansby commentedon Feb 15, 2017
Thanks for the easy to copy-and-paste example! Running it on our current 2.0.x branch, I get the image below; is that what you were expecting?
2sn commentedon Feb 15, 2017
Thank you for the very fast reply.
Yes, this is what it should look like. The version you were using not yet the current version on PyPi that I can install using pip, I assume; is this current version on the devel branch from git?
dstansby commentedon Feb 15, 2017
Yep, that's right. The branch I used is the current bugfix branch for 2.0, so the bug will be fixed when 2.0.1 is released.
dopplershift commentedon Feb 15, 2017
I think this was fixed by #7976 .
2sn commentedon Mar 26, 2017
Is there any ETA when version 2.0.1 may be released?
It seems there is many fixes that have run up ... you don't have to catch them all at once, it does not have to be perfect, it never will be, you can always release a version 2.0.2 to fix more issues, ...
For most of my collaborators wanting to use my packages and don't want to compile their own matplotlib version from git, the only really easy solution around for now is to revert back to 1.6.3, which is not really what we want ...
bastibl commentedon Apr 11, 2017
The bugfix branch also seems to have some problems. While it draws the patch correctly (as above), legends still use black hatches (i.e. don't match the patch).
dstansby commentedon Apr 11, 2017
Would it be possible to open a new issue with a minimal example that demonstrates broken hatches on legends?