Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
  • 02:53
    ojeda-e commented #21799
  • 02:47
    ojeda-e synchronize #21799
  • 02:07
    QuLogic commented #21796
  • 02:02
    QuLogic commented #21796
  • 01:19
    QuLogic commented #21796
  • 01:11
    QuLogic commented #21796
  • 00:50
    ojeda-e synchronize #21799
  • Dec 01 23:57
    QuLogic edited #21833
  • Dec 01 23:56
    QuLogic opened #21833
  • Dec 01 23:56
    QuLogic milestoned #21833
  • Dec 01 23:56
    QuLogic labeled #21833
  • Dec 01 22:25
    QuLogic milestoned #21783
  • Dec 01 22:20
    meeseeksmachine opened #21832
  • Dec 01 22:19
    meeseeksdev[bot] labeled #21832
  • Dec 01 22:19
    meeseeksdev[bot] milestoned #21832
  • Dec 01 22:19
    tacaswell closed #21820
  • Dec 01 22:19
    tacaswell closed #21783
  • Dec 01 22:19

    tacaswell on main

    Drop setuptools-scm requirement… Merge pull request #21820 from … (compare)

  • Dec 01 21:51
    tacaswell commented #21820
  • Dec 01 21:43
    timhoffm commented #21773
Thomas A Caswell
@tacaswell
Rohit Rawat
@xordux
is that test.png present somewhere or you created a blank image of size 800x600 to test it?
Thomas A Caswell
@tacaswell
I saved a mostly empty 800x600 image
In [1]: import matplotlib.pyplot as plt                                                                                                                                                                                                                                                                

In [2]: fig, ax = plt.subplots()                                                                                                                                                                                                                                                                       

In [3]: fig.set_size_inches(8, 6)                                                                                                                                                                                                                                                                      

In [4]: fig.savefig('/tmp/test.png')
Rohit Rawat
@xordux
ok, thank you.
@tacaswell getting same in windows as well:
(test_env) C:\PersonalProjects\MachineLearning\Github\18151>"C:\Program Files\ImageMagick-7.0.10-Q16-HDRI\convert" -size 800x600 -depth 8 -delay 20.0 -loop 0 tmp.png temp.m4v
convert: delegate failed `"ffmpeg.exe" -nostdin -v -1 -i "%M%%d.pam" -plays %I "%u.%m" 2> "%u"' @ error/delegate.c/InvokeDelegate/1898.
Thomas A Caswell
@tacaswell
can you please leave a comment on the imagemagick issue ?
Rohit Rawat
@xordux
yes, right away.
Thomas A Caswell
@tacaswell
thanks!
Rohit Rawat
@xordux
If I want to retest a PR without new commit, then should I do it like this ? Or is there some other way you do it?
Thomas A Caswell
@tacaswell
if the UI does not have a "re run the test" button in the check section (it is not always clear to me when things show up for me because I am an admin and when things are there for everyone), then close and re-open the PR
on the "checks" tab I have a "re-run failed" link next to the failed tests
Rohit Rawat
@xordux
There is no such button that I can find. I will close and re-open then.
Grigory Starkov
@Gregstrq
When a 3d surface is plotted, each of the surface patches has a constant color. Is it possible to have the colors change continously? (I think one can use linear interpolation inside the patch)
Bruno Beltran
@brunobeltran

looks like font_manager._rebuild is gone on master. What's the go-to advice for users that are installing many new fonts after installing matplotlib? Should I tell them to clear mpl.get_cachedir() directly each time?

The only way to do it from inside mpl itself now seems to be something like

font_manager.fontManager = font_manager._load_fontmanager(try_read_cache=False)?

(not saying that my go-to advice for people was to call a private method, but if someone who isn't me needed to update their recommendations 🐑)
Antony Lee
@anntzer
I think there should just be a public API for that (I believe this may be mentioned somewhere in one of the long font_manager issues...)
Benjamin Root
@WeatherGod
#18738 is some sort of spam. Have we decided on how to handle these, or just close without comment?
Dieter Werthmüller
@prisae

I have an issue with matplotlib v3.3 and interactive plots. Everything works fine for matplotlib<3.3, when updating there is no interactivity. The ipywidget buttons work, but nothing happens:

  • In a Jupyter Notebook
  • using %matplotlib notebook
  • using ipywidgets

I read through the v3.3 release notes, but I couldn't find anything in particular ...

5 replies
Jody Klymak
@jklymak
Best open an issue.
7 replies
Jody Klymak
@jklymak
@tacaswell #18735 image interpolation again. I actually agree with the poster, and feel that we are doing this incorrectly. Yes over/under will bleed into good data, but if that happens, and you can see it, you need more resolution. The solution is not all the hoops we are jumping through.
Thomas A Caswell
@tacaswell
I am in the process of writing a response and do not agree with either of you
Jody Klymak
@jklymak
Great!
I mean so long as it becomes clearer, disagreement is healthy...
Thomas A Caswell
@tacaswell
if you use a non-finite kernel that bad must bleed in (but over/under should be handled ed correctly) as if you try to interpolate between [1, nan], the answer at ever point needs to be nan
the core problem is that the OP is looking at categorical data that should not be interpolated using continuous methods
Jody Klymak
@jklymak
You are taking the view that the rendered image is data, not that the rendered image is a visual representation of the data
is our resampling a data operation or an image operation?
I'd argue its an image operation, and should look good. If more precision is required, then more precision can be asked for
Thomas A Caswell
@tacaswell
doing the re-sampling in RGBA space also leads to colors not in the color map
Jody Klymak
@jklymak
Sure, which is fine
Thomas A Caswell
@tacaswell
in all cases
Jody Klymak
@jklymak
if the pixels are small enough Thats what anti-aliasing does
Thomas A Caswell
@tacaswell
how is it fine!
Jody Klymak
@jklymak
Because its meant to look good from a "distance".
Individual pixels are not data
Thomas A Caswell
@tacaswell
I am taking the position on this that @anntzer takes on symlog
Jody Klymak
@jklymak
If you need individual pixels to be data, then over-sample
Thomas A Caswell
@tacaswell
with the higher-order kernels resampling in RGBA space generated large regions of not-in-gamut color
Benjamin Root
@WeatherGod
totally agree with @tacaswell here. It was considered a horrendous bug when it was pointed out that we were creating out-of-gamut colors that we were willing to break baseline tests to fix it.
Jody Klymak
@jklymak
What do you mean by "out of gamut"? R>1 and/or R<0? Or just not in the colormap? If it was just out of the colormap, then I disagree - if you are subsampling an image, then you will get new colors at the color boundaries and they will be perceptually smoothing the adjacent colors.
Thomas A Caswell
@tacaswell
I mean color not in the colormap
it makes the image completely non-interpret-able
how to you show "red" in a viridis colorbar?
Thomas A Caswell
@tacaswell
left a long comment with pictures on the issue
in very short: core of the problem is using continuous resampling on categorical data and despite the current order being hard, doing it in the other order is un-equivocally wrong
Antony Lee
@anntzer
I have a vague feeling that the correct approach may be 1) to resample in data space when smoothing (i.e. large pixels, few data points, and using e.g. sinc) because in that case out-of-gamut colors are just ridiculous (this is the case in your reply) and we can see the operation as interpolating undersampled data points first, but 2) to resample in color space in the opposite case of multiple data points going into a single pixel, because I tend to imagine this (as suggested by the OP) as "viewing the image from afar" (although really averaging should be done in some visual space such as Lab instead of RGB, but heh...)
Jody Klymak
@jklymak
Right - I was just writing something like this. There are two things the user wants: to smooth their data so its not so "blocky" and to have a good looking image. They are different operations, one should be done in data space and one in color space.
Antony Lee
@anntzer
1) is about smoothing data to look non-blocky, which I think may or may not be something that we really need to provide (at least in theory we could always say, we're not going to interpolate the data for you), but 2) is really about displaying something sensible when there are more points than pixels (and in that case straight decimation, which was the previous default, can easily give very misleading results (thanks for implementing interpolation="antialiased" and fixing that :)))
Ian Hunt-Isaak
@ianhi
Are %matplotlib widget and %matplotlib ipympl the same thing? I really thought they were but https://github.com/matplotlib/matplotlib/issues/18741#issuecomment-709455563 has be confused