imshow(..., alpha=0.5) produces different results in 2.x

I think I have whittled this down sufficiently from the Cartopy example that is failing. Running imshow(..., alpha=0.5) produces a different result in 2.0.0b1 vs. 1.5.1 (both installed via conda.)

import numpy as np
import matplotlib.pyplot as plt
import matplotlib as mpl

# mpl.style.use('classic')

im = np.zeros((100, 100, 3), dtype=np.uint8)
plt.imshow(im, extent=(1, 10, 1, 10), alpha=0.5)
plt.xlim(0, 11)
plt.ylim(0, 11)

plt.savefig('imshow_alpha.png')

Using 1.5.1:

Using 2.0.0b1:

Using 2.0.0b1 with the classic style:

GIMP tells me that the first is 50%, while the latter two are 75%.

@mdboom

@mdboom, is this a result of the image architecture overhaul?

On Mon, Jun 6, 2016 at 1:47 AM, Elliott Sales de Andrade <
notifications@github.com> wrote:

I think I have whittled this down sufficiently from the Cartopy example
that is failing. Running imshow(..., alpha=0.5) produces a different
result in 2.0.0b1 vs. 1.5.1 (both installed via conda.)

import numpy as npimport matplotlib.pyplot as pltimport matplotlib as mpl

mpl.style.use('classic')

im = np.zeros((100, 100, 3), dtype=np.uint8)
plt.imshow(im, extent=(1, 10, 1, 10), alpha=0.5)
plt.xlim(0, 11)
plt.ylim(0, 11)

plt.savefig('imshow_alpha.png')

Using 1.5.1:
[image: imshow_alpha_1 5 1]
https://cloud.githubusercontent.com/assets/302469/15812915/443b626e-2b88-11e6-89ca-14c0afbf4648.png
Using 2.0.0b1:
[image: imshow_alpha_2 0]
https://cloud.githubusercontent.com/assets/302469/15812917/4d8bc746-2b88-11e6-8e9f-68a37cd456d6.png
Using 2.0.0b1 with the classic style:
[image: imshow_alpha_2 0_classic]
https://cloud.githubusercontent.com/assets/302469/15812919/543d038e-2b88-11e6-9607-e80ae97955f6.png

GIMP tells me that the first is 50%, while the latter two are 75%.

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
#6540, or mute the thread
https://github.com/notifications/unsubscribe/AARy-NGkkXjrVwUIRPMnkrtSrR4Cnj66ks5qI7RogaJpZM4IulFm
.

@mdboom

I would be curious to know how the result compares with output from
gtkcairo.

On Tue, Jun 7, 2016 at 11:32 AM, Benjamin Root ben.v.root@gmail.com wrote:

@mdboom, is this a result of the image architecture overhaul?

On Mon, Jun 6, 2016 at 1:47 AM, Elliott Sales de Andrade <
notifications@github.com> wrote:

I think I have whittled this down sufficiently from the Cartopy example
that is failing. Running imshow(..., alpha=0.5) produces a different
result in 2.0.0b1 vs. 1.5.1 (both installed via conda.)

import numpy as npimport matplotlib.pyplot as pltimport matplotlib as mpl

mpl.style.use('classic')

im = np.zeros((100, 100, 3), dtype=np.uint8)
plt.imshow(im, extent=(1, 10, 1, 10), alpha=0.5)
plt.xlim(0, 11)
plt.ylim(0, 11)

plt.savefig('imshow_alpha.png')

Using 1.5.1:
[image: imshow_alpha_1 5 1]
https://cloud.githubusercontent.com/assets/302469/15812915/443b626e-2b88-11e6-89ca-14c0afbf4648.png
Using 2.0.0b1:
[image: imshow_alpha_2 0]
https://cloud.githubusercontent.com/assets/302469/15812917/4d8bc746-2b88-11e6-8e9f-68a37cd456d6.png
Using 2.0.0b1 with the classic style:
[image: imshow_alpha_2 0_classic]
https://cloud.githubusercontent.com/assets/302469/15812919/543d038e-2b88-11e6-9607-e80ae97955f6.png

GIMP tells me that the first is 50%, while the latter two are 75%.

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
#6540, or mute the
thread
https://github.com/notifications/unsubscribe/AARy-NGkkXjrVwUIRPMnkrtSrR4Cnj66ks5qI7RogaJpZM4IulFm
.

I tried GTK3Cairo, but it's current broken:

Traceback (most recent call last):
  File ".../lib/matplotlib/backends/backend_gtk3cairo.py", line 45, in on_draw_event
    self._render_figure(w, h)
  File ".../lib/matplotlib/backends/backend_gtk3cairo.py", line 34, in _render_figure
    self.figure.draw (self._renderer)
  File ".../lib/matplotlib/artist.py", line 63, in draw_wrapper
    draw(artist, renderer, *args, **kwargs)
  File ".../lib/matplotlib/figure.py", line 1129, in draw
    renderer, self, dsu, self.suppressComposite)
  File ".../lib/matplotlib/image.py", line 139, in _draw_list_compositing_images
    a.draw(renderer)
  File ".../lib/matplotlib/artist.py", line 63, in draw_wrapper
    draw(artist, renderer, *args, **kwargs)
  File ".../lib/matplotlib/axes/_base.py", line 2373, in draw
    mimage._draw_list_compositing_images(renderer, self, dsu)
  File ".../lib/matplotlib/image.py", line 163, in _draw_list_compositing_images
    flush_images()
  File ".../lib/matplotlib/image.py", line 147, in flush_images
    image_group[0].draw(renderer)
  File ".../lib/matplotlib/artist.py", line 63, in draw_wrapper
    draw(artist, renderer, *args, **kwargs)
  File ".../lib/matplotlib/image.py", line 472, in draw
    renderer.draw_image(gc, l, b, im)
  File ".../lib/matplotlib/backends/backend_cairo.py", line 180, in draw_image
    im.shape[1]*4)
TypeError: expected a writeable buffer object

It looks like alpha is getting double applied?

FIX: only apply alpha once for images

Navigation Menu

Repository navigation

imshow(..., alpha=0.5) produces different results in 2.x #6540

mpl.style.use('classic')

mpl.style.use('classic')

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

imshow(..., alpha=0.5) produces different results in 2.x #6540

Description

Activity

WeatherGod commented on Jun 7, 2016

mpl.style.use('classic')

WeatherGod commented on Jun 7, 2016

mpl.style.use('classic')

QuLogic commented on Jun 8, 2016

tacaswell commented on Jun 8, 2016

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions