Repository navigation
Tutorial pyplot_scales.py crashes when used with plt.tight_layout() #6789
Description
Trying to replace plt.subplots_adjust with plt.tight_layout in the pyplot tutorial pyplot_scales.py (see #6779), seems to results in errors.
- Matplotlib version, Python version and Platform (Windows, OSX, Linux ...):
- Linux (at least Archlinux and CentOs 7)
- Tested on mpl 1.5.1 from Anaconda and mpl 2.0.0b3.post1799+ga4fdd60 from github
- MWE (kind of…):
Based on pyplot_scales.py:
import numpy as np
import matplotlib.pyplot as plt
plt.style.use('default')
# make up some data in the interval ]0, 1[
y = np.random.normal(loc=0.5, scale=0.4, size=1000)
y = y[(y > 0) & (y < 1)]
y.sort()
x = np.arange(len(y))
plt.figure(1)
plt.subplot(221)
plt.subplot(222)
plt.subplot(223)
# logit
plt.subplot(224)
plt.plot(x, y)
plt.yscale('logit')
#plt.subplots_adjust(top=0.92, bottom=0.08, left=0.10, right=0.95, hspace=0.25,
# wspace=0.35)
plt.tight_layout()
plt.show()When I run this example with plt.tight_layout instead of plt.subplots_adjust (or simply no subplot layout adjusting at all), I get these outputs (mpl 2.0.0b3.post1799+ga4fdd60):
- with Qt4Agg backend
In [1]: run pyplot_scales_mwe.py # that's the MWE posted above
/home/adrien/matplotlib/lib/matplotlib/transforms.py:380: RuntimeWarning: invalid value encountered in double_scalars
return points[1, 0] - points[0, 0]
/home/adrien/matplotlib/lib/matplotlib/tight_layout.py:199: RuntimeWarning: invalid value encountered in double_scalars
kwargs["wspace"] = hspace / h_axes
/home/adrien/matplotlib/lib/matplotlib/axis.py:1008: UserWarning: Unable to find pixel distance along axis for interval padding of ticks; assuming no interval padding needed.
warnings.warn("Unable to find pixel distance along axis "
/home/adrien/matplotlib/lib/matplotlib/axis.py:1018: UserWarning: Unable to find pixel distance along axis for interval padding of ticks; assuming no interval padding needed.
warnings.warn("Unable to find pixel distance along axis "
Traceback (most recent call last):
File "/home/adrien/matplotlib/lib/matplotlib/backends/backend_qt5agg.py", line 182, in __draw_idle_agg
FigureCanvasAgg.draw(self)
File "/home/adrien/matplotlib/lib/matplotlib/backends/backend_agg.py", line 464, in draw
self.figure.draw(self.renderer)
File "/home/adrien/matplotlib/lib/matplotlib/artist.py", line 68, in draw_wrapper
return draw(artist, renderer, *args, **kwargs)
File "/home/adrien/matplotlib/lib/matplotlib/figure.py", line 1262, in draw
renderer, self, dsu, self.suppressComposite)
File "/home/adrien/matplotlib/lib/matplotlib/image.py", line 139, in _draw_list_compositing_images
a.draw(renderer)
File "/home/adrien/matplotlib/lib/matplotlib/artist.py", line 68, in draw_wrapper
return draw(artist, renderer, *args, **kwargs)
File "/home/adrien/matplotlib/lib/matplotlib/axes/_base.py", line 2383, in draw
mimage._draw_list_compositing_images(renderer, self, dsu)
File "/home/adrien/matplotlib/lib/matplotlib/image.py", line 139, in _draw_list_compositing_images
a.draw(renderer)
File "/home/adrien/matplotlib/lib/matplotlib/artist.py", line 68, in draw_wrapper
return draw(artist, renderer, *args, **kwargs)
File "/home/adrien/matplotlib/lib/matplotlib/axis.py", line 1109, in draw
ticks_to_draw = self._update_ticks(renderer)
File "/home/adrien/matplotlib/lib/matplotlib/axis.py", line 942, in _update_ticks
tick_tups = [t for t in self.iter_ticks()]
File "/home/adrien/matplotlib/lib/matplotlib/axis.py", line 885, in iter_ticks
majorLocs = self.major.locator()
File "/home/adrien/matplotlib/lib/matplotlib/ticker.py", line 1721, in __call__
return self.tick_values(vmin, vmax)
File "/home/adrien/matplotlib/lib/matplotlib/ticker.py", line 1726, in tick_values
locs = self._raw_ticks(vmin, vmax)
File "/home/adrien/matplotlib/lib/matplotlib/ticker.py", line 1674, in _raw_ticks
nbins = max(min(self.axis.get_tick_space(), 9),
File "/home/adrien/matplotlib/lib/matplotlib/axis.py", line 2008, in get_tick_space
return int(np.floor(length / size))
ValueError: cannot convert float NaN to integer
which results in
(NB: it's a screenshot because savefig complained about the “ValueError: cannot convert float NaN to integer” stuff)
- with TkAgg backend
In [1]: run pyplot_scales_mwe.py # that's again the MWE posted above
/home/adrien/matplotlib/lib/matplotlib/transforms.py:380: RuntimeWarning: invalid value encountered in double_scalars
return points[1, 0] - points[0, 0]
/home/adrien/matplotlib/lib/matplotlib/tight_layout.py:199: RuntimeWarning: invalid value encountered in double_scalars
kwargs["wspace"] = hspace / h_axes
/home/adrien/matplotlib/lib/matplotlib/axis.py:1008: UserWarning: Unable to find pixel distance along axis for interval padding of ticks; assuming no interval padding needed.
warnings.warn("Unable to find pixel distance along axis "
/home/adrien/matplotlib/lib/matplotlib/axis.py:1018: UserWarning: Unable to find pixel distance along axis for interval padding of ticks; assuming no interval padding needed.
warnings.warn("Unable to find pixel distance along axis "
Exception in Tkinter callback
Traceback (most recent call last):
File "/home/adrien/anaconda/envs/Python2.7_matplotlib-git/lib/python2.7/lib-tk/Tkinter.py", line 1536, in __call__
return self.func(*args)
File "/home/adrien/matplotlib/lib/matplotlib/backends/backend_tkagg.py", line 283, in resize
self.show()
File "/home/adrien/matplotlib/lib/matplotlib/backends/backend_tkagg.py", line 354, in draw
FigureCanvasAgg.draw(self)
File "/home/adrien/matplotlib/lib/matplotlib/backends/backend_agg.py", line 464, in draw
self.figure.draw(self.renderer)
File "/home/adrien/matplotlib/lib/matplotlib/artist.py", line 68, in draw_wrapper
return draw(artist, renderer, *args, **kwargs)
File "/home/adrien/matplotlib/lib/matplotlib/figure.py", line 1262, in draw
renderer, self, dsu, self.suppressComposite)
File "/home/adrien/matplotlib/lib/matplotlib/image.py", line 139, in _draw_list_compositing_images
a.draw(renderer)
File "/home/adrien/matplotlib/lib/matplotlib/artist.py", line 68, in draw_wrapper
return draw(artist, renderer, *args, **kwargs)
File "/home/adrien/matplotlib/lib/matplotlib/axes/_base.py", line 2383, in draw
mimage._draw_list_compositing_images(renderer, self, dsu)
File "/home/adrien/matplotlib/lib/matplotlib/image.py", line 139, in _draw_list_compositing_images
a.draw(renderer)
File "/home/adrien/matplotlib/lib/matplotlib/artist.py", line 68, in draw_wrapper
return draw(artist, renderer, *args, **kwargs)
File "/home/adrien/matplotlib/lib/matplotlib/axis.py", line 1109, in draw
ticks_to_draw = self._update_ticks(renderer)
File "/home/adrien/matplotlib/lib/matplotlib/axis.py", line 942, in _update_ticks
tick_tups = [t for t in self.iter_ticks()]
File "/home/adrien/matplotlib/lib/matplotlib/axis.py", line 885, in iter_ticks
majorLocs = self.major.locator()
File "/home/adrien/matplotlib/lib/matplotlib/ticker.py", line 1721, in __call__
return self.tick_values(vmin, vmax)
File "/home/adrien/matplotlib/lib/matplotlib/ticker.py", line 1726, in tick_values
locs = self._raw_ticks(vmin, vmax)
File "/home/adrien/matplotlib/lib/matplotlib/ticker.py", line 1674, in _raw_ticks
nbins = max(min(self.axis.get_tick_space(), 9),
File "/home/adrien/matplotlib/lib/matplotlib/axis.py", line 2008, in get_tick_space
return int(np.floor(length / size))
ValueError: cannot convert float NaN to integer
which produces an empty figure.
- If this is an image generation bug attach a screenshot demonstrating the issue:
See above. - If this is a regression (Used to work in an earlier version of Matplotlib), please
note where it used to work:
The behavior seems to be the same with mpl 1.5.1 and (at least) both TkAgg and Qt4Agg backends.
Activity
dstansby commented on Jan 19, 2017
Here's a more minimalist example that throws the same error as above:
dstansby commented on Jan 28, 2017
Todo (see #7880):
janmedlock commented on Mar 1, 2017
I've hit this bug, too. A slightly different minimal example, that doesn't depend on tight_layout:
import matplotlib.pyplot as plt plt.plot([0.1, 0.9], [1, 2]) plt.xscale('logit') plt.xlabel('x') plt.show()Rendering with QT5Agg causes a similar error as in the original report:
Traceback (most recent call last): File "/usr/lib/python3/dist-packages/matplotlib/backends/backend_qt5agg.py", g FigureCanvasAgg.draw(self) File "/usr/lib/python3/dist-packages/matplotlib/backends/backend_agg.py", linw self.figure.draw(self.renderer) File "/usr/lib/python3/dist-packages/matplotlib/artist.py", line 63, in draw_r draw(artist, renderer, *args, **kwargs) File "/usr/lib/python3/dist-packages/matplotlib/figure.py", line 1143, in draw renderer, self, dsu, self.suppressComposite) File "/usr/lib/python3/dist-packages/matplotlib/image.py", line 139, in _draws a.draw(renderer) File "/usr/lib/python3/dist-packages/matplotlib/artist.py", line 63, in draw_r draw(artist, renderer, *args, **kwargs) File "/usr/lib/python3/dist-packages/matplotlib/axes/_base.py", line 2409, inw mimage._draw_list_compositing_images(renderer, self, dsu) File "/usr/lib/python3/dist-packages/matplotlib/image.py", line 139, in _draws a.draw(renderer) File "/usr/lib/python3/dist-packages/matplotlib/artist.py", line 63, in draw_r draw(artist, renderer, *args, **kwargs) File "/usr/lib/python3/dist-packages/matplotlib/axis.py", line 1150, in draw self.label.draw(renderer) File "/usr/lib/python3/dist-packages/matplotlib/artist.py", line 63, in draw_r draw(artist, renderer, *args, **kwargs) File "/usr/lib/python3/dist-packages/matplotlib/text.py", line 762, in draw raise ValueError("posx and posy should be finite values") ValueError: posx and posy should be finite valuesI tried all the backends on my machine, with the same result (but perhaps slightly different error messages).
The script works fine if plt.xlabel() is removed, so I investigated the labeling code. Before rendering the label's position (using plt.gca().xaxis.get_label().get_position()) is (0.5, 0) and after rendering it is (0.5, -inf).
Setting the y position to -inf happens in matplotlib.axis.XAxis._update_label_position(), which gets called during rendering. Both before and after (partial?) rendering, plt.gca().spines['bottom'].get_path() is Path(array([[13., 0.], [13., 0.]]), None). But the transformed path,
is Path(array([[7373.45454545, 52.8], [7373.45454545, 52.8]]), None) before and Path(array([[nan, nan], [nan, nan]]), None) after (partial?) rendering.
And I'm over my head at this point.
janmedlock commented on Mar 2, 2017
... and calling gca().spines['bottom']._adjust_location() ('bottom' for logit x scale, 'left' for logit y scale) before rendering prevents the error:
import matplotlib.pyplot as plt plt.plot([0.1, 0.9], [1, 2]) plt.xscale('logit') plt.xlabel('x') gca().spines['bottom']._adjust_location() plt.show()It also works with dstansby's tight_layout example:
import matplotlib.pyplot as plt fig, axs = plt.subplots(2, 2) axs[1, 1].set_yscale('logit') axs[1, 1].spines['left']._adjust_location() plt.tight_layout() plt.show()dstansby commented on Sep 24, 2018
Seems to be fixed on master, feel free to re-open if it turns out to still be a problem.