Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
  • 19:15
    Kreijstal edited #24203
  • 19:15
    Kreijstal edited #24203
  • 19:14
    Kreijstal opened #24203
  • 19:11
    scottshambaugh labeled #24108
  • 19:11
    scottshambaugh labeled #24108
  • 19:05
    oscargus commented #24161
  • 19:04

    oscargus on v3.6.x

    Backport PR #24193: DOC: Explai… Merge pull request #24200 from … (compare)

  • 19:04
    oscargus closed #24200
  • 19:04

    oscargus on v3.6.x

    Backport PR #24194: DOC: Improv… Merge pull request #24201 from … (compare)

  • 19:04
    oscargus closed #24201
  • 19:03
    oscargus labeled #24202
  • 19:02
    dependabot[bot] labeled #24202
  • 19:02

    dependabot[bot] on github_actions

    Bump pypa/cibuildwheel from 2.1… (compare)

  • 19:02
    dependabot[bot] opened #24202
  • 18:55
    scottshambaugh commented #23485
  • 18:19
    jklymak commented #24145
  • 18:04
    jacoverster synchronize #24145
  • 18:02
    jacoverster commented #24145
  • 17:59
    jacoverster synchronize #24145
  • 17:56
    jacoverster synchronize #24145
Thomas A Caswell
@tacaswell
Yes, there is path simplification code that is run. I think the rcParams["path.simplify"] can be set to False to turn it off
and if that fails, there is a per-artsist level control as well
Franklin Wei
@built1n
tacaswell: yes, that appeared to solve that issue, but apparently some SVG renderers do some ugly things when rendering dense polylines... :(
my polylines are actually just discretized circular arcs, but in 3D - is there a way to draw an Arc object on a 3D axes object?
Thomas A Caswell
@tacaswell
short answer: maybe;
medium answer: at the renderer level we have a limited set of functions (which is how we can write to so many output formats) and "arc" is not one of them;
longer answer: at the Path level we support bezier codes so you may be able to do a bit better (but not perfect) with fewer points in the path
ah, "in 3D" I do not know if we support bezier there....
the way we support 3D is that at draw time we look at the view, do the projection to 2D and then draw the 2D artist.....even if we correctly transform the control points, I would be very surprised if the bezier paths drawn in 2D are actually correct for 3D
hannah
@story645
how do folks feel about hacktober fest? I'm particularly intrigued with the low/no-code options https://hacktoberfest.com/about/#low-or-non-code
Thomas A Caswell
@tacaswell
Jody Klymak
@jklymak
Yes very interesting. We have a data serving institution on campus, and I never understood what they were trying to do; apparently it was the query-centric approach. Nice to see someone delineate the dataset approach
Melissa Weber Mendonça
@melissawm
Hi folks - any maintainers available to join us tomorrow on the new contributor meeting? Just so I can plan ahead :)
5 replies
Subbalakshmi
@subba2048
Hello, I'm new here, request to be added to '#incubator'.
1 reply
Elliott Sales de Andrade
@QuLogic
done
1 reply
Jody Klymak
@jklymak
Hi all: Weekly Meeting. Note new HackMD note.
1 reply
Please add agenda items (PRs that need attention) probably special attention to 3.6.1 issues
Maybe an easy one we don't need meeting time: matplotlib/matplotlib#24088
hannah
@story645

Can someone put on call please:
Hello, Projects.
It's time to gather your project Updates.

Please reply with the most relevant news from your project, which may include things like:
recent releases
upcoming events
job or hiring announcements
calls for participation
Include links/URLs where possible so our readers can get more information.
Please send your responses by EOD Thursday, October 13th, 2022.

Emmanuel Bauma Murairi
@Emmastro
Hi everyone! Can I be added to the #incubator channel? Just joined, want to explore how I can contribute on the project
Jody Klymak
@jklymak
A bit late
Emmanuel Bauma Murairi
@Emmastro
As of the #incubator channel doesn't exist anymore?
1 reply
hannah
@story645
For the arvix post poked coiled and Matt wants to know if there's anything we want him to direct readers to on the Matplotlib side
hannah
@story645
Like I'm thinking maybe the third party page if they made libraries based on their science or showcase to link their papers?
Thomas A Caswell
@tacaswell
That seems reasonable
Elliott Sales de Andrade
@QuLogic
looks like the greetings action was fixed today
hannah
@story645
🥳
Elliott Sales de Andrade
@QuLogic
looking for the second review on #24111 so I can do 3.6.1 today
story645
@story645:matrix.org
[m]
Looks like Tom requested @jklymak review it?
And also @timhoffm so I thought it was waiting on 'em to approve?
Jody Klymak
@jklymak
I think @timhoffm should approve. However its 01:30 in Germany?
Elliott Sales de Andrade
@QuLogic
should be fine; the things he requested were changed
Elliott Sales de Andrade
@QuLogic
Tim Hoffmann
@timhoffm
Dubious. Either it is typosquatting or an honest rebuild (it seems like we only started to provide arm64 wheels since 3.5). Anyway, this should be clearly marked as third party and use its own description and not link to our github/docs. Is there a way to contact the author (or alternatively PyPI)?
11 replies
IcyDuck
@NRaudseps
Hi, I'm new here. Can I be added to the #incubator channel?
1 reply
chromafunk
@chromafunk
This message was deleted
why is the amplitude wave ( axs[0] ) being plotted on top of axs[1] ?
if I plot something else on axs[0] it works fine
chromafunk
@chromafunk

`# Set up plot
fig, axs = plt.subplots(2)

load file with librosa

x, sr = librosa.load(file)
white_check_mark
eyes
raised_hands

amplitude of audio over time

axs[0].set_title('Amplitude over Time')
axs[0].set_ylabel('Amplitude')
axs[0].set_xlabel('Time (seconds)')

Add the audio data to the plot

axs[0] = plt.plot(time_sf, soundwave_sf, alpha=0.5)

zero crossing rate

zcrs = librosa.feature.zero_crossing_rate(x)
axs[1].set_title("Zero Crossing")
axs[1] = plt.plot(zcrs[0])`

givent the fact that when I plot something else on axs0 it works, could this be a bug ?
Jody Klymak
@jklymak
@chromafunk please post at https://discourse.Matplotlib.org
Franklin Wei
@built1n
@tacaswell so amazingly, I managed to hack something up with matplotlib.patches.Arc and pathpatch_2d_to_3d() to get fully vectorized elliptic arcs embedded in 3D
1 reply
it actually works very well - I'm using it to generate some optical diagrams for publication, and it look so good that I can just drop the SVG right into the manuscript
Melissa Weber Mendonça
@melissawm

I think I am missing some dependency even though I am following the development setup guide. Matplotlib builds fine but the docs build fails with

ImportError: /home/melissa/projects/matplotlib/lib/matplotlib/ft2font.cpython-310-x86_64-linux-gnu.so: undefined symbol: FT_Load_Glyph

Any tips?

2 replies
Jamie
@inchinn1
Just want to mention that fig.subplot_mosaic is great and hope it moves from provisional to permanent. I already mostly use it to build multipanel figures. It is much more intuitive.
2 replies
hannah
@story645
Anyone have objections to matplotlib/matplotlib#24096 ?
24 replies
Jody Klymak
@jklymak
Hi all, weekly meeting agenda: https://hackmd.io/NIl7IVPLTTi-CILXLnNu7w. Relatively full, but some things could be moved around if you have additions.
story645
@story645:matrix.org
[m]
For the record, the issue was raised by Anthony, who is a very longstanding, very prolific contributor - and in the issue Tim suggested that it should actually be a tutorial. I think we can start w/ an example and maybe down the line flesh out into a tutorial?
I think other formats are buried as comments in the other examples (and I'd maybe vote for removing that in a followup if this PR goes in)
Noa Tamir
@noatamir
I can't join today's meeting. Have a good one!
Elliott Sales de Andrade
@QuLogic
going to merge #24158 soon to fix CI if no-one else does soon