Packaging/tranform into package

[jriddy]

Someone at my organization wants to be able to use this in a project, but it's not currently set up as a pip-installable python package, nor does it have a container package to refer to it in other code. After this it's easy to install this into a virtualenv with

$ pip install git+http://github.com/richzhang/PerceptualSimilarity.git

And then you could use it in your code as

from perceptualsimilarity.models import dist_models

This changes that in the simplest way possible: by moving it all into a perceptualsimilarity directory and adding a basic setup.py to allow it to be installed. With a bit more metadata in setup.py, you could upload it to PyPI as well.

If you like this, we could also transform the scripts to entry-points that get installed on the command-line PATH when you install from PyPI as well.

[jriddy]

We've found some issues with the import statements in the existing code...I'll get to fixing these things soon and update this PR

[richzhang]

Thanks a lot for making the package! Was the PR fixed with changes?

[jriddy]

No I'm sorry I've been busy due to release pressure at work. There's a few things that got broken when I changed this, but my boy Karl assures me that he has fixes. Once I get back from vacation in a few weeks I'll get that integrated, and I'll push my changes. Sorry that the vicissitudes of work life keep me from following up on my promise, but I'll get there 😉

…

On Mon, Sep 3, 2018, 21:28 Richard Zhang ***@***.***> wrote: Thanks a lot for making the package! Was the PR fixed with changes? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#13 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AA_aH_I_6nznXQnKwhR5pF2Sm4olW9ayks5uXgGJgaJpZM4VUkok> .

[richzhang]

Of course I understand that. Thanks so much!

[jriddy]

Hey just saw this, I can give this another go if you're interested, but this PR was so old it was stale.

Are you still interested in making this into a pip-distributable package? So it go on PyPI and such?

[richzhang]

Hi, I finally got around to this. pip install lpips :)