binary_blobs fails with io.imshow

[Mec-iS]

Description

Tested in Python 3.7 and 3.8 on Ubuntu 18. I am using pyQT5 as matplot backend by matplotlib.use('Qt5Agg')
This chunk of code fails:

import skimage.io as io
import skimage.data as data
img = data.binary_blobs(length=512, blob_size_fraction=0.1, seed=5)
io.imshow(img)
io.show()

with stacktrace:

Traceback (most recent call last):
  File "src/sk_img/00_basics.py", line 9, in <module>
    io.imshow(img)
  File "/.venv/lib/python3.8/site-packages/skimage/io/_io.py", line 159, in imshow
    return call_plugin('imshow', arr, plugin=plugin, **plugin_args)
  File "/.venv/lib/python3.8/site-packages/skimage/io/manage_plugins.py", line 207, in call_plugin
    return func(*args, **kwargs)
  File "/.venv/lib/python3.8/site-packages/skimage/io/_plugins/matplotlib_plugin.py", line 150, in imshow
    lo, hi, cmap = _get_display_range(image)
  File "/.venv/lib/python3.8/site-packages/skimage/io/_plugins/matplotlib_plugin.py", line 97, in _get_display_range
    ip = _get_image_properties(image)
  File "/.venv/lib/python3.8/site-packages/skimage/io/_plugins/matplotlib_plugin.py", line 55, in _get_image_properties
    is_low_contrast(image))
  File "/.venv/lib/python3.8/site-packages/skimage/exposure/exposure.py", line 651, in is_low_contrast
    limits = np.percentile(image, [lower_percentile, upper_percentile])
  File "<__array_function__ internals>", line 5, in percentile
  File "/.venv/lib/python3.8/site-packages/numpy/lib/function_base.py", line 3818, in percentile
    return _quantile_unchecked(
  File "/.venv/lib/python3.8/site-packages/numpy/lib/function_base.py", line 3937, in _quantile_unchecked
    r, k = _ureduce(a, func=_quantile_ureduce_func, q=q, axis=axis, out=out,
  File "/.venv/lib/python3.8/site-packages/numpy/lib/function_base.py", line 3515, in _ureduce
    r = func(a, **kwargs)
  File "/.venv/lib/python3.8/site-packages/numpy/lib/function_base.py", line 4064, in _quantile_ureduce_func
    r = _lerp(x_below, x_above, weights_above, out=out)
  File "/.venv/lib/python3.8/site-packages/numpy/lib/function_base.py", line 3961, in _lerp
    diff_b_a = subtract(b, a)
TypeError: numpy boolean subtract, the `-` operator, is not supported, use the bitwise_xor, the `^` operator, or the logical_xor function instead.

The same operation succeed using matplotlib:

import skimage.io as io
import skimage.data as data
import matplotlib.pyplot as plt

img = data.binary_blobs(length=512, blob_size_fraction=0.1, seed=5)
plt.imshow(img)
plt.show()

Way to reproduce

Run the chunk above

Version information

3.8.8 (default, Feb 22 2021, 15:43:28) 
[GCC 7.5.0]
Linux-5.4.0-66-generic-x86_64-with-glibc2.27
scikit-image version: 0.18.1
numpy version: 1.20.1

same with:

3.7.9 (default, Sep 29 2020, 15:13:16) 
[GCC 7.5.0]
Linux-5.4.0-66-generic-x86_64-with-debian-buster-sid
scikit-image version: 0.18.1
numpy version: 1.20.1

[mkcor]

Hello @Mec-iS,

Since you're mentioning PyQt5, it's not clear to me whether your usage is via non-interactive scripts...

scikit-image/skimage/io/_io.py

Lines 182 to 190 in 17db79a

	def show():
	'''Display pending images.
	Launch the event loop of the current gui plugin, and display all
	pending images, queued via `imshow`. This is required when using
	`imshow` from non-interactive scripts.
	A call to `show` will block execution of code until all windows
	have been closed.

Having said this, let me quote @jni: "in the medium term (~1y) we aim to pull out the matplotlib dependency and make sure the focus of scikit-image is on image processing, not display or IO. So you should probably start migrating your downstream code to start using plt.imshow directly."

[Mec-iS]

ok. thanks for the reference, so basically io.imshow() will be deprecated.

I am using the library in scripts that are supposed to visualize the results into subplots.