Characteristic classes on manifolds

[YoungMath]

My dear developers,

right now, I'm working on my master thesis and my task is to implement characteristic classes (of the tangent bdl. of a manifold), such as the A-genus, into Sage. The implementation shall be based

upon this piece of work by my supervisor. Briefly, one can compute the classes out of the curvature matrix and the corresponding power series.

So far, I've implemented the graded algebra of mixed differential forms and the next step will be the matrix framework for the desired classes.

However, it might also be convenient to implement some methods into manifolds/differentiable/manifold.py directly.

Is this idea worth a ticket? You wanna see the code I've done so far?

Cheers,
YoungMath

[Eric Gourgoulhon]

Hi,

Yes sure! Please open a ticket.

(you might find some hints at https://sagemanifolds.obspm.fr/contrib.html ).

Best wishes,

Eric.

[MJ]

Great! So, the next step is cloning sage to my local repo and connect git to the trac server, right?

Are there some important things I have to take care of?

[Eric Gourgoulhon]

Le dimanche 31 mars 2019 12:58:26 UTC+2, Michael Jung a écrit :

Great! So, the next step is cloning sage to my local repo and connect git to the trac server, right?

Yep!

Are there some important things I have to take care of?

Above all, make sure that your code is fully python3 compatible

cf. https://wiki.sagemath.org/Python3-compatible%20code

since SageMath migration to python3 is getting close:

https://trac.sagemath.org/ticket/26212

[Travis Scrimshaw]

Dear MJ,

   Also if you plan on including your code in Sage (which it sounds like you are, +1!), it would be a good idea to break the code into smaller pieces to help facilitate and easier review. For example, the algebra itself can be one ticket. Feel free to also cc me (by using tscrim) on the tickets.

Best,

Travis

[Michael Jung]

Thanks for your interest! :) I started the ticket right away (to be honest, it took some effort as newbie).

I already finished the code, yet the first push will take some time. I need to create the doctests, first. Unless you say it's not necessary for now.

I'm looking forward to this project - hopefully I do not start banging my head on the desk during this process, like you guys do. :D

Regards,

Michael

P.S. Yeah, I changed my name a couple of times, But I'm a devel virgin. However, this one is fixed now.

[Michael Jung]

I started another ticket, just for my own (and maybe other's?) convenience. It's about the latex name of multiplied scalar fields.

Moreover, I've encountered a problem regarding the multiplication of my algebra. Suppose A and B are mixed differential forms, a is some differential form and f is a differentiable scalar field on M. Then, I get the following results:

sage: A.__mul__(B)

Mixed differential form A/\B on the 2-dimensional differentiable manifold M
sage: f.__mul__(A)
Mixed differential form f/\A on the 2-dimensional differentiable manifold M
sage: A.__mul__(a)
Mixed differential form A/\a on the 2-dimensional differentiable manifold M
sage: a.__mul__(A)
Mixed differential form A/\a on the 2-dimensional differentiable manifold M

All fine, except the very last line. The coercions are working quite well so far - without any errors. Yet, what might explain the very different results? Can you give me a first hint, though not seeing the code?

Cheers,

Michael

[Eric Gourgoulhon]

Le mardi 2 avril 2019 22:54:51 UTC+2, Michael Jung a écrit :

All fine, except the very last line. The coercions are working quite well so far - without any errors. Yet, what might explain the very different results? Can you give me a first hint, though not seeing the code?

Well it's difficult to answer without seeing the code... Can you share it by adding a branch to the ticket?

Best wishes,

Eric.

[michi...@kabelmail.de]

Yes, I could. But it's not well documented yet. Should I add it anyway?

Best regards,
Michael

-------- Originalnachricht --------
Betreff: [sage-devel] Re: Characteristic classes on manifolds
Von: Eric Gourgoulhon
An: sage-devel
Cc:

--
You received this message because you are subscribed to a topic in the Google Groups "sage-devel" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/sage-devel/hwsQWuLN0nc/unsubscribe.
To unsubscribe from this group and all its topics, send an email to sage-devel+...@googlegroups.com.
To post to this group, send email to sage-...@googlegroups.com.
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.

[Dima Pasechnik]

On Wed, Apr 3, 2019 at 10:48 AM michi...@kabelmail.de
<michi...@kabelmail.de> wrote:
>
> Yes, I could. But it's not well documented yet. Should I add it anyway?

as long as the ticket is not marked as "needs review", your branch
there can be WIP, why not?

> 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.

[Michael Jung]

Okay, the code has been uploaded. :)

Am 03.04.19 um 13:12 schrieb Dima Pasechnik:

[Eric Gourgoulhon]

Le mercredi 3 avril 2019 13:38:17 UTC+2, Michael Jung a écrit :

Okay, the code has been uploaded. :)

Thanks!

I found the reason. I am answering on the ticket.

Eric.

[michi...@kabelmail.de]

Thank you! I fixed it right away and commited a sage compilable version. Moreover, I added a jupyter example file (in german).

Best regards
Michael

-------- Originalnachricht --------
Betreff: Re: [sage-devel] Re: Characteristic classes on manifolds

Von: Eric Gourgoulhon
An: sage-devel
Cc:

--

[Michael Jung]

I added a link to a demo notebook in the description of the ticket.

Furthermore, I've encountered another unnice issue (see ticket comments). I'd appreciate when someone knows how to fix it.

Best regards

Michael

[Michael Jung]

Hello,

I changed the status to "needs_review" since the doctests are now complete. :)

What are the next steps? Lean back and sip tea?

Regards,
Michael

Am 06.04.19 um 18:54 schrieb Michael Jung:

--

[Travis Scrimshaw]

Hi Michael,

   Yep, basically, and then answer anything the reviewer(s) come up with.

Best,

Travis

To unsubscribe from this group and all its topics, send an email to sage-...@googlegroups.com.