Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Any interest in appveyor wheels? #3786

Closed
thequackdaddy opened this issue Jun 26, 2017 · 12 comments
Closed

Any interest in appveyor wheels? #3786

thequackdaddy opened this issue Jun 26, 2017 · 12 comments
Labels
Milestone

Comments

@thequackdaddy
Copy link
Member

@thequackdaddy thequackdaddy commented Jun 26, 2017

We live in a Windows world, and building statsmodels on windows can be a little confusing, especially because the compiler is python version-specific. Additionally, we don't have wheels on pypi for windows and statsmodels 0.8.0. (Kind of some self-interest here because at work my personal desktop I can install the windows build tools. On some other machines, I can't but I can install wheels. I love statsmodels, but lacking the ability to pip install statsmodels always makes me do some extra steps.)

I played with having appveyor build wheels and it seems to add about 1 minute per build... so a total of 3 minutes for the current setup. That would get us windows wheels for python 2.7 and 3.6 (most of the time... when the whole statespace bug doesn't crash).

Appveyor could also theoretically upload the wheels to pypi only when master is compiled (and/or on a tag).

@josef-pkt
Copy link
Member

@josef-pkt josef-pkt commented Jun 26, 2017

If you or someone can set this up, then this would be great.

@josef-pkt
Copy link
Member

@josef-pkt josef-pkt commented Jun 26, 2017

Maybe we can reuse the cloud storage that other related packages are using.
@matthew-brett ?

@matthew-brett
Copy link
Contributor

@matthew-brett matthew-brett commented Jun 26, 2017

Yes, you're certainly welcome to use the Rackspace stuff for wheels. The killer problem is that we don't have Scipy wheels for Windows...

@waynenilsen
Copy link
Contributor

@waynenilsen waynenilsen commented Sep 6, 2017

@josef-pkt josef-pkt added this to the 0.9 milestone Oct 7, 2017
@bashtage
Copy link
Member

@bashtage bashtage commented Oct 31, 2017

@josef-pkt Should be deferred until after 0.9.

@waynenilsen
Copy link
Contributor

@waynenilsen waynenilsen commented Oct 31, 2017

Scipy windows wheels have officially landed in 1.0

Highlights of this release

Some of the highlights of this release are:

  • Major build improvements. Windows wheels are available on PyPI for the
    first time, and continuous integration has been set up on Windows and OS X
    in addition to Linux.
@jbrockmendel
Copy link
Contributor

@jbrockmendel jbrockmendel commented Sep 4, 2018

Is this still an issue?

@matthew-brett
Copy link
Contributor

@matthew-brett matthew-brett commented Sep 5, 2018

Er - no, in that we now have the https://github.com/MacPython/statsmodels-wheels repo, that includes an Appveyor wheel build. Yes, a little bit, in that that repo still hasn't generated wheels that pass the tests:

https://ci.appveyor.com/project/matthew-brett/statsmodels-wheels/build/1.0.33

Let's close this one when that does work...

@bashtage
Copy link
Member

@bashtage bashtage commented Sep 5, 2018

Master should be passing now. Is it worth checking before a release is nearly ready?

@matthew-brett
Copy link
Contributor

@matthew-brett matthew-brett commented Sep 5, 2018

Sadly master is not passing at the moment on the wheel builds - see #5108

@matthew-brett
Copy link
Contributor

@matthew-brett matthew-brett commented Sep 5, 2018

OK - it looks like master is passing now : https://travis-ci.org/MacPython/statsmodels-wheels/builds/424746178

So - I think this can be closed.

@thequackdaddy
Copy link
Member Author

@thequackdaddy thequackdaddy commented Sep 6, 2018

Thank you @matthew-brett

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Linked pull requests

Successfully merging a pull request may close this issue.

6 participants