Help with matroids (and more generally?), PR #36492

[John H Palmieri]

This post concerns https://github.com/sagemath/sage/pull/36492. The main topic of that PR is matroids about which I know almost nothing, so I am not the right person to review it. The structure of the PR perhaps opens up broader questions. The author has some new code, along with a PDF documenting it, and it is currently not designed to be incorporated into the existing sage.matroids module, but as more of a standalone piece of code.

- Should we have a "contrib" directory where we can easily include efforts like this?

- Should this particular code be instead included in our thematic tutorials? (If so, it needs someone to shepherd it through the process.)

- Or should this code be folded into the existing `sage.matroids` stuff?

--

John

[Travis Scrimshaw]

Hi John,

   None of the above. I think maintaining good programming practices is important for the codebase and it is a good thing to teach contributors who might not understand. Hence, we should spend the time having them get it into a format that is acceptable for Sage. In the end, it should play nice with the current matroid code, but I don't think that is a requirement for a new feature.

Best,
Travis

[Dima Pasechnik]

This is about a  recent (2018?) generalisation of oriented matroids. Do we have anything about the usual oriented matroids?

[Aram Dermenjian]

There is currently a project that is in progress for getting (usual) oriented matroids into sage. It's still a work in progress, but an initial version is close to done. (See https://github.com/thecaligarmo/oriented_matroids for the project) Due to external time commitments, it hasn't finished as fast as I would have liked. (It's currently set-up as a package, but eventually it should nicely/easily be able to be integrated into sagemath itself)

Kindly,

Aram Dermenjian

--
You received this message because you are subscribed to the Google Groups "sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/6818C672-81A0-4224-B0F6-F3919884D555%40gmail.com.

[Dima Pasechnik]

On Wed, Nov 22, 2023 at 8:29 AM Aram Dermenjian
<aram.derme...@gmail.com> wrote:
>
> There is currently a project that is in progress for getting (usual) oriented matroids into sage. It's still a work in progress, but an initial version is close to done. (See https://github.com/thecaligarmo/oriented_matroids for the project) Due to external time commitments, it hasn't finished as fast as I would have liked. (It's currently set-up as a package, but eventually it should nicely/easily be able to be integrated into sagemath itself)

Just in case, I tried to install your package with current Sage
(10.2.rc4), and it installs fine and all the tests run
as "sage -t oriented_matroids" pass, too.
So that's a bit strange to read in Readme that "This package currently
does not work." - what is this meant to mean?

Dima

>
> Kindly,
> Aram Dermenjian
>
> On Wed, 22 Nov 2023 at 07:58, Dima Pasechnik <dim...@gmail.com> wrote:
>>
>> This is about a recent (2018?) generalisation of oriented matroids. Do we have anything about the usual oriented matroids?
>>
>>
>> On 22 November 2023 04:25:34 GMT, 'Travis Scrimshaw' via sage-devel <sage-...@googlegroups.com> wrote:
>>>
>>> Hi John,
>>> None of the above. I think maintaining good programming practices is important for the codebase and it is a good thing to teach contributors who might not understand. Hence, we should spend the time having them get it into a format that is acceptable for Sage. In the end, it should play nice with the current matroid code, but I don't think that is a requirement for a new feature.
>>>
>>> Best,
>>> Travis
>>>
>>> On Wednesday, November 22, 2023 at 8:27:31 AM UTC+9 John H Palmieri wrote:
>>>>
>>>> This post concerns https://github.com/sagemath/sage/pull/36492. The main topic of that PR is matroids about which I know almost nothing, so I am not the right person to review it. The structure of the PR perhaps opens up broader questions. The author has some new code, along with a PDF documenting it, and it is currently not designed to be incorporated into the existing sage.matroids module, but as more of a standalone piece of code.
>>>>
>>>> - Should we have a "contrib" directory where we can easily include efforts like this?
>>>> - Should this particular code be instead included in our thematic tutorials? (If so, it needs someone to shepherd it through the process.)
>>>> - Or should this code be folded into the existing `sage.matroids` stuff?
>>>>
>>>> --
>>>> John
>>>>
>> --
>> You received this message because you are subscribed to the Google Groups "sage-devel" group.
>> To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
>> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/6818C672-81A0-4224-B0F6-F3919884D555%40gmail.com.
>
> --
> You received this message because you are subscribed to the Google Groups "sage-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.

> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/CAKQMtipOAK77sQurUCLHpETV9sii5t_RPCuAJTQoxb9C6c%2BCkA%40mail.gmail.com.

[Aram Dermenjian]

The master branch has bugs in it and don't produce "correct" oriented matroids. These have been fixed in the current development branch, but the development branch needs documentation to be updated and also to be double checked to ensure there are no bugs/errors. (There's also additional things I'd like to add into it, but these can be saved for after it gets integrated into sage)

Aram

To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/CAAWYfq3uwJiMsMXpsx17fhk-L4Htvde4OUs833b3Cck9oSEHTw%40mail.gmail.com.

[Aram Dermenjian]

As a follow up to this email thread: Everything should be working now. If people want to test it out and/or see if there are any bugs/issues, feel free to have at it and add any issues into the github.

Question for humans: I'm planning on merging this into sage so that other developers can help grow it and it's not just me working on it. Would it be better to first wait for any bugs that people might notice and then add it into sage? Or should I just go ahead and start merging it in and any bugs can be fixed while it's in sage itself?

-Aram

[Matthias Koeppe]

The next step of integration could be to add this package as an optional package to Sage, see https://github.com/sagemath/sage/issues/31164

[Travis Scrimshaw]

Question for humans: I'm planning on merging this into sage so that other developers can help grow it and it's not just me working on it. Would it be better to first wait for any bugs that people might notice and then add it into sage? Or should I just go ahead and start merging it in and any bugs can be fixed while it's in sage itself?

 The review process is designed to catch bugs. Specifically, the reviewer should try to make sure the code is bug-free. Part of this is done by writing doctests. Of course, we cannot always be perfect about this, and we can fix things later. I would just create the PR for adding the code into Sage (assuming it is in the appropriate stage for this).

Best,

Travis