Mailman 3 October 2017 - Matplotlib-devel

CI: only run macosx on travis cron jobs / only branches
by Thomas Caswell Jan. 19, 2018

Jan. 19, 2018

Folks, https://blog.travis-ci.com/2017-09-22-macos-update Due to the travis OSX infrastructure not being able to keep up with demand they are going to limit concurrency of jobs that us osx. We have not seen many (any?) mac specific issues which I think means we are not getting much value out of running all of the PRs on OSX. We can probably get better linux concurrency (and help lighten the load on travis) if we do test on OSX less. Thoughts? Tom

4 3

Agenda for 2017-10-23 call
by Thomas Caswell Oct. 23, 2017

Oct. 23, 2017

Folks, Trying to plan ahead today! This is what I would like to talk about on the call (with the goal of getting 2.1.1 tagged this week). Sorry for leaking a bit of my internal state with the formatting (it looks nice with org-mode in emacs). Tom * call agenda ** REV [[mpl:9262]] Minor doc markup fixes. - State "WAIT" from "REV" [2017-10-22 Sun 20:28] ** REV [[mpl:9257]] FIX: segfault on truncated png *** REV [[mpl:9256]] reading truncated png can segfault python ** REV [[mpl:9534]] Fix webagg *** REV [[mpl:9380]] Cannot import pyplot. NameError: 'FigureManagerWebAgg' is not defined ** REV [[mpl:9477]] In LogTransform, clip after log, not before. *** REV [[mpl:9288]] Histograms disappear with logarithmic y-axis *** REV [[mpl:9457]] ax.fill_between broken for log scale and values below zero ** REV [[mpl:9324]] Allow kwarg handles and labels figure.legend and make doc for kwargs the same *** REV [[mpl:9320]] 2.1 figure.legend broken ** REV [[mpl:9304]] Fedora build patches *** WAIT [[mpl:6538]] On armv7hl, some get_cursor_data calls return 0 instead of None. - State "WAIT" from "REV" [2017-10-22 Sun 17:39] *** REV [[mpl:7158]] Arrays are not equal in 2.0.0b4 testsuite on Fedora rawhide/aarch64 (ARM v8 64bit) ** REV [[mpl:9359]] Keep track of axes in interactive navigation. *** REV [[mpl:9358]] zoom/pan stack bug in 2.1.0 ** REV [[mpl:9389]] Assign event to later Axes if zorders are tied. *** REV [[mpl:9388]] Mouse events have incorrect inaxes/data properties when axes overlap (matplotlib 2.1.0) ** [0/5] categoricals *** REV [[mpl:9336]] Integer Categorical Values Not Getting Mapped Correctly *** REV [[mpl:9340]] bugfix/test for #9336 integer overwrite in categorical *** REV [[mpl:9350]] Data types not preserved in categoricals *** REV [[mpl:9318]] Don't sort categorical keys. *** REV [[mpl:9494]] Categorical not hitting update path on fill_between ** REV [[mpl:8122]] keyword labelrotation is not recognized ** REV [[mpl:8947]] Different result, slower runtime of heatmap between 2.0.0 and 2.0.1 ** REV [[mpl:9429]] Undefined name `baseline`? ** REV [[mpl:9436]] Instance of 'TextBox' has no 'observers' member? ** REV [[mpl:9521]] fix xkcd context ** REV [[mpl:8069]] backend_driver.py overhaul *** REV [[mpl:9284]] Restore something like backend_driver

1 0

release schedule
by Thomas Caswell Oct. 14, 2017

Oct. 14, 2017

Folks, Re-addressing the long-term release schedule first discussed in May 2016 [1] (which we are now about a year behind on). 2018-01/02 : 2.2 LTS - this will be the LTS release and get patch level fixes until 2020 2018-07 : 3.0 - first release which is python 3 only. The level of backporting to 2.2 will depend on if someone champions it. I am personally only willing to do critical (failure to import and segfault fixes) backports. If functionality bug fixes are important to people, I am happy to delegate all responsibility of the 2.2.x series. For 3.0, I would like to start with python 3.6+ only. The release schedule for 3.7 is 2018-06 so that would return us to, what I understand used to be the standard, support for the last 2 versions of python. Further, both python 3.4 and 3.5 already have no planned future releases. Tom [1] https://mail.python.org/pipermail/matplotlib-devel/2016-May/000374.html

1 0

PR review / new contributors
by Thomas Caswell Oct. 8, 2017

Oct. 8, 2017

Folks, Between 2.1.0 be released and my interview with NumFOCUS [1] we will hopefully be getting an influx of new contributors. We should always be kind and respectful in pull requests (even if the person you are talking to 'can take it', potential contributors will read it and may be turned away), but please be extra patient with new contributors. I think everyone with commit rights can see "new contributor" next to the users name in the GH UI. Remember that first contribution to a big project can be intimidating for many reasons, they may be simultaneously learning git, github, python, and the conventions of Matplotlib in addition to asking their work to be publicly judged by strangers! At least for my self, my first PR to Matplotlib was to change a 4 to a 7 [2] and it still took me most of an afternoon to work up the nerve to open the PR. For documentation only changes please use the threshold "Is it better than in was?" and avoid copy-editing-via-github comments. Tom [1] https://www.numfocus.org/blog/matplotlib-lead-developer-explains-why-he-can… [2] https://github.com/matplotlib/matplotlib/pull/1062

1 0

[ANN] Matplotlib 2.1 released
by Thomas Caswell Oct. 7, 2017

Oct. 7, 2017

We are happy to announce the release of Matplotlib 2.1. This is the second minor release in the Matplotlib 2.x series and the first release with major new features since 1.5. This release contains approximately 2 years worth of work by 275 contributors across over 950 pull requests. Highlights from this release include: - support for string categorical values - export of animations to interactive javascript widgets - major overhaul of polar plots - reproducible output for ps/eps, pdf, and svg backends - performance improvements in drawing lines and images - GUIs show a busy cursor while rendering the plot along with many other enhancements and bug fixes. The gallery, examples, and tutorials have been overhauled and consolidated: Examples: http://matplotlib.org/gallery/index.html Tutorials: http://matplotlib.org/tutorials/index.html A big thank you to everyone who contributed to this release! Wheels are available on pypi for win/mac/manylinux and for conda via conda-forge. Full whats new: http://matplotlib.org/users/whats_new.html#new-in-matplotlib-2-1 Full API changes: http://matplotlib.org/api/api_changes.html#api-changes-in-2-1-0 github stats: http://matplotlib.org/users/github_stats.html Tom

1 0

LIGO nobel prize
by Thomas Caswell Oct. 4, 2017

Oct. 4, 2017

Folks, This years Nobel prize in physics was awarded for the work done by LIGO/VIRGO [1] project. We know that Matplotlib is used for some of the LIGO analysis (see the notebooks published at https://losc.ligo.org/tutorials/ ). Just a reminder that our work on Matplotlib does have an impact! Tom [1] https://www.nobelprize.org/nobel_prizes/physics/laureates/2017/press.html

2 1

Re: [Matplotlib-devel] [Matplotlib-users] How to make \mathdefault{} the default?
by Daniele Nicolodi Oct. 3, 2017

Oct. 3, 2017

On 10/3/17 2:59 PM, Daniele Nicolodi wrote: > Hello, > > when adding LaTeX symbols to labels on my figures, most often I use > \mathdefault{} to make Matplotlib use the same fond for the regular text > and for the LaTeX part. Is there a way to configure Matplotlib to use > the regular text by default? Looking at the configuration file I > haven't found anything hinting to it. After staring a bit more at the example configuration file, I can answer my own question. This does the trick: matplotlib.rcParams['mathtext.fontset'] = 'custom' I'm a bit worried by the note next to this setting in the example configuration file, though: # The following settings allow you to select the fonts in math mode. # They map from a TeX font name to a fontconfig font pattern. # These settings are only used if mathtext.fontset is 'custom'. # Note that this "custom" mode is unsupported and may go away in the # future. Why is it unsupported? I find this functionality very neat. Thanks. Cheers, Daniele

1 0

v2.1 tagged
by Thomas Caswell Oct. 3, 2017

Oct. 3, 2017

Folks, Just pushed the v2.1.0 tag. There is still a bit of work to be done on docs. Will do wider announcement once that is done and binaries are available. Moved all of the un-closed issues to the 2.2 milestone. Tom

2 1

feature request
by Rogier Delporte Oct. 2, 2017

Oct. 2, 2017

Dear maintainer(s) Dear developpers I would like to request a new matplotlib feature. The current version of matplotlib shows the coordinates of the cursor in the bottom right corner. It would be great if some option or mode could be added which snaps the cursor to data points. This would take some guesswork out of reading a plot with a lot of points. I know this feature is similar to the mpldatacursor toolkit, but mpldatacursor does not provide a smooth and fast way to pinpoint a data point. I hope someone would like to implement this feature, because I don't have sufficient programming skills to implement this myself. Thanks for the awesomeness that is matplotlib! Rogier

4 4