[Matplotlib-devel] step-by-step to installling from source using conda?
Eric Firing
efiring at hawaii.edu
Sun Aug 6 13:29:49 EDT 2017
That's what I do, also. For the installation from source I now use
either "pip install ." or "pip install -e .". Depending on how much I
am changing (e.g., switching from maintenance to master, or back), I do
"pip uninstall matplotlib" and "git clean -dfx" (only with all desired
changes committed, of course) to clean out the old version before
installing the new.
At least on a Mac, there can still be some bogus test failures.
Eric
On 2017/08/05 1:27 PM, Thomas Caswell wrote:
> I sometimes take a blunter approach, install Matplotlib via conda and
> then uninstall it which leaves behind the dependencies ;)
>
> Tom
>
> On Sat, Aug 5, 2017 at 5:33 PM Klymak Jody <jklymak at uvic.ca
> <mailto:jklymak at uvic.ca>> wrote:
>
> Hi Chris. Thanks so much. That'll be useful
>
> I think I was partially tricked by the fact that %matplotlib
> notebook wasn't working for master in a Jupiter notebook. but I
> see now that it doesn't wrk for anyone so I guess I was installing
> things adequately ;-)
>
> Thanks again. Jody
>
> Sent from my iPhone
>
> On Aug 5, 2017, at 14:22, Chris Barker <chris.barker at noaa.gov
> <mailto:chris.barker at noaa.gov>> wrote:
>
>> This should be pretty straightforward, but MPL's dependencie are
>> pretty complex. IN thoery, simplyL
>>
>> 1) create an environment
>> 2) install the dev dependencies
>> 3) build away.
>>
>> but in fact, step (2) there is not so straightforward, so here we
>> go...
>>
>> 1) I recommend using conda-forge as a source of packages -- it is
>> more up to date than defaults, and you are developing....
>>
>> conda config --add channels conda-forge
>>
>> 2) create an environment for doing your development:
>>
>> conda create -n mpl-dev python=3
>>
>> 3) activate that environment:
>>
>> source activate mpl-dev
>>
>> 4) install the requirements of mpl:
>>
>> OK -- this one is kinda messy for MPL. It has a lot of
>> requirements that depend someone on which sub-packages you are
>> going to build.
>>
>> but first you need pk-config, which MPL used to find/configure
>> stuff on *nix systems.
>>
>> (you may have this from Brew already)
>>
>> conda install pkg-config
>>
>> python setup.py --requires doesn't work, as mpl does a lot of
>> ative configuration.
>>
>> but:
>>
>> python ./setup.py dry-run
>>
>> will get you a good list:
>>
>> REQUIRED DEPENDENCIES AND EXTENSIONS
>> numpy: yes [not found. pip may install it below.]
>> six: yes [six was not found.pip will attempt to
>> install
>> it after matplotlib.]
>> dateutil: yes [dateutil was not found. It is
>> required for date
>> axis support. pip/easy_install may attempt to
>> install it after matplotlib.]
>> backports.functools_lru_cache: yes [Not required]
>> subprocess32: yes [Not required]
>> pytz: yes [pytz was not found. pip/easy_install may
>> attempt to install it after matplotlib.]
>> cycler: yes [cycler was not found.
>> pip/easy_install may
>> attempt to install it after matplotlib.]
>> tornado: yes [tornado was not found. It is required
>> for the
>> WebAgg backend. pip/easy_install may
>> attempt to
>> install it after matplotlib.]
>> pyparsing: yes [pyparsing was not found. It is
>> required for
>> mathtext support. pip/easy_install may
>> attempt to
>> install it after matplotlib.]
>> libagg: yes [pkg-config information for 'libagg'
>> could not
>> be found. Using local copy.]
>> freetype: no [The C/C++ header for freetype2
>> (ft2build.h)
>> could not be found. You may need to
>> install the
>> development package.]
>> png: yes [version 1.6.26]
>> qhull: yes [pkg-config information for 'libqhull'
>> could not
>> be found. Using local copy.]
>>
>> you don't want pip to try to install anything, so I created a
>> requirements file (enclosed)
>>
>> conda install --file conda_requirements_dev.txt
>>
>> And now you should be able to build!
>>
>> python setup.py build
>>
>> Note that you won't get all the back-ends -- other dependencies
>> need to be installed for that.
>>
>> BTW: it would be nice to have something like that
>> conda_requirements_dev.txt in the MPL repo -- the setup-tools
>> stuff is way to tied into the build/packaging/installing system.
>>
>> HTH,
>>
>> -CHB
>>
>>
>> On Tue, Jun 13, 2017 at 6:24 AM, Jody Klymak <jklymak at gmail.com
>> <mailto:jklymak at gmail.com>> wrote:
>>
>>
>> Hi all,
>>
>> I’m somewhat stymied trying to understand how to install
>> matplotlib in a development environment using conda.
>>
>> I have MacOS, and I have brew gcc etc. I know how to set up a
>> env in conda.
>>
>> I eventually want to be able to run the tests, so that I can
>> stop uploading things with PEP8 errors etc.
>>
>> Sorry if this is too vague. I can start over again and send
>> specific errors if that is easier, but I would bet someone has
>> a step-by-step.
>>
>> Thanks, Jody
>>
>>
>> _______________________________________________
>> Matplotlib-devel mailing list
>> Matplotlib-devel at python.org <mailto:Matplotlib-devel at python.org>
>> https://mail.python.org/mailman/listinfo/matplotlib-devel
>>
>>
>>
>>
>> --
>>
>> Christopher Barker, Ph.D.
>> Oceanographer
>>
>> Emergency Response Division
>> NOAA/NOS/OR&R (206) 526-6959 <tel:%28206%29%20526-6959> voice
>> 7600 Sand Point Way NE (206) 526-6329 <tel:%28206%29%20526-6329> fax
>> Seattle, WA 98115 (206) 526-6317 <tel:%28206%29%20526-6317>
>> main reception
>>
>> Chris.Barker at noaa.gov <mailto:Chris.Barker at noaa.gov>
>> <conda_requirements_dev.txt>
>> _______________________________________________
>> Matplotlib-devel mailing list
>> Matplotlib-devel at python.org <mailto:Matplotlib-devel at python.org>
>> https://mail.python.org/mailman/listinfo/matplotlib-devel
> _______________________________________________
> Matplotlib-devel mailing list
> Matplotlib-devel at python.org <mailto:Matplotlib-devel at python.org>
> https://mail.python.org/mailman/listinfo/matplotlib-devel
>
>
>
> _______________________________________________
> Matplotlib-devel mailing list
> Matplotlib-devel at python.org
> https://mail.python.org/mailman/listinfo/matplotlib-devel
>
More information about the Matplotlib-devel
mailing list