Which version of matplotlib are you using?

On Sat, Nov 15, 2014 at 2:25 PM, Ryan notifications@github.com wrote:

Pyplot.imsave does not seem to work.

import numpy as npimport matplotlib.pyplot as plt
f = np.random.rand(0, 255, shape=(92,92))
plt.imsave("test.png", f)

results in an AssertionError

---

AssertionError Traceback (most recent call last)
in ()
----> 1 plt.imsave("test.pdf", f)

/usr/lib/python2.7/site-packages/matplotlib/pyplot.pyc in imsave(_args, *_kwargs)
2218 @docstring.copy_dedent(_imsave)
2219 def imsave(_args, *_kwargs):
-> 2220 return _imsave(_args, *_kwargs)
2221
2222

/usr/lib/python2.7/site-packages/matplotlib/image.pyc in imsave(fname, arr, vmin, vmax, cmap, format, origin, dpi)
1312 canvas = FigureCanvas(fig)
1313 im = fig.figimage(arr, cmap=cmap, vmin=vmin, vmax=vmax, origin=origin)
-> 1314 fig.savefig(fname, dpi=dpi, format=format, transparent=True)
1315
1316

/usr/lib/python2.7/site-packages/matplotlib/figure.pyc in savefig(self, _args, *_kwargs)
1468 self.set_frameon(frameon)
1469
-> 1470 self.canvas.print_figure(_args, *_kwargs)
1471
1472 if frameon:

/usr/lib/python2.7/site-packages/matplotlib/backend_bases.pyc in print_figure(self, filename, dpi, facecolor, edgecolor, orientation, format, *_kwargs)
2141 *_kwargs)
2142 renderer = self.figure._cachedRenderer
-> 2143 bbox_inches = self.figure.get_tightbbox(renderer)
2144
2145 bbox_artists = kwargs.pop("bbox_extra_artists", None)

/usr/lib/python2.7/site-packages/matplotlib/figure.pyc in get_tightbbox(self, renderer)
1611 bb.append(ax.get_tightbbox(renderer))
1612
-> 1613 _bbox = Bbox.union([b for b in bb if b.width != 0 or b.height != 0])
1614
1615 bbox_inches = TransformedBbox(_bbox,

/usr/lib/python2.7/site-packages/matplotlib/transforms.pyc in union(bboxes)
720 Return a :class:Bbox that contains all of the given bboxes.
721 """
--> 722 assert(len(bboxes))
723
724 if len(bboxes) == 1:

AssertionError:

It appears that the Bbox.union() results in an empty list.
The matplotlibrc file that I'm using:
https://raw.githubusercontent.com/byuimpact/numerical_computing/develop/matplotlibrc

—
Reply to this email directly or view it on GitHub
#3802.

My apologies. I'm using Matplotlib 1.4.2 on archlinux

When I tried this example I had issues with generating the data (exceptions from numpy)

import numpy as np
import matplotlib.pyplot as plt
f = (np.random.rand(92,92)*255).astype('int')
plt.imsave("test.png", f)

works as expected. I suspect that you were passing an empty array to imsave which is what was blowing up.

In [15]: np.random.rand(0, 5)
Out[15]: array([], shape=(0, 5), dtype=float64)

This should probably fail more gracefully, but I don't think this is a bug.

Sorry, my bad. I gave the wrong sample code. See the original post.

Again, I think there's an error in the example. randint has no argument shape -- I think you meant size. Once fixing that, I couldn't reproduce the crash. Which version of matplotlib are you running? Sorry -- I see above you're running 1.4.2.

PR attached.