Vector Integration Proposal for GSoC feedback

79 views
Skip to first unread message

Faisal Riyaz

unread,
Mar 28, 2020, 3:00:44 AM3/28/20
to sy...@googlegroups.com
Dear Developers,

Please provide your feedback on my GSoC proposal for Vector Integration.

Here is the link to the wiki.

It is still not complete. I am looking forward to your feedback.

Best regards
Faisal Riyaz

Faisal Riyaz

unread,
Mar 29, 2020, 2:52:10 AM3/29/20
to sy...@googlegroups.com
To view my GSoC proposal in google docs, please use the following link

I am looking forward to your comments.

Best
Faisal Riyaz
 

Aaron Meurer

unread,
Mar 29, 2020, 3:59:42 AM3/29/20
to sympy
The Google Docs document seems to not be filled out all the way.

Aaron Meurer
> --
> You received this message because you are subscribed to the Google Groups "sympy" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to sympy+un...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sympy/CAEe-xqTx17gYCbkATb%3DyHK%3D3HSJf4EaFP%2BrmZEooqmi5NuZdPQ%40mail.gmail.com.

Faisal Riyaz

unread,
Mar 29, 2020, 4:23:27 AM3/29/20
to sy...@googlegroups.com
I am sorry. Pasted the wrong link. Here is the correct link.

I still have to complete some sections.

Thanks

Faisal Riyaz

unread,
Mar 29, 2020, 9:23:45 PM3/29/20
to sy...@googlegroups.com
Hello All,

Any suggestions? I will add functions for relating line integral and surface integral (Stoke's Theorem and Green's Theorem). I do not still know how difficult will it be. Also, In my proposal, I have tried to provide the API of the functions. Is it enough or should I also give a rough implementation? Feel free to comment here or the google docs.

Thanks
Faisal Riyaz

Aaron Meurer

unread,
Mar 29, 2020, 10:03:59 PM3/29/20
to sympy
Discussing the implementation would be more important. You don't need
to have actual code (other than possibly pseudocode). The API you
suggested would probably end up being something different, but the
implementation would be the same regardless of the API.

Aaron Meurer
> --
> You received this message because you are subscribed to the Google Groups "sympy" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to sympy+un...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/sympy/CAEe-xqQBKZooor5eiQ5CPbkctoVFTHD5xxLFVnDtxFm9Ye2D6A%40mail.gmail.com.

Faisal Riyaz

unread,
Mar 30, 2020, 12:54:28 AM3/30/20
to sy...@googlegroups.com
I am looking at the Maple functionalities for vector Integration. I think Maple has the best functionality for vector integration among other CAS.
They have many different classes for commonly used surfaces and curves. Some of them are
Circle
Ellipse
Sphere
Sector
Box
Triangle


For integration over Surface, I suppose I have to calculate the unit normal vector. Parametrix representation of the surface needs to be determined. And represent the differential area in terms of parameters. This differential element will be associated with every surface. Then represent the given vector/scalar field in terms of parameters of the surface.  Then the required integration can be easily performed. Also, the direction of normal vector needs to be determined.

Similarly, for integration over curves, we have to calculate the differential line element. Represent the line element and scalar/vector field in terms of parameters.

Finding a parametric representation of well-known surfaces is easy but it can be difficult for some surfaces.

I am going in the right direction. Please suggest. Also, who will be the potential mentor for this project?

Thanks
Faisal


Alan Bromborsky

unread,
Mar 30, 2020, 10:47:49 AM3/30/20
to sy...@googlegroups.com
I would suggest calculating for a geometric simplex


since simplexes are the building blocks for finite element method calculations.  I think one of the main applications of your package could be for calculating the boundary element method equations for various problems. 

Faisal Riyaz

unread,
Mar 30, 2020, 11:15:10 AM3/30/20
to sy...@googlegroups.com
Thank you for your suggestions, Alan. For the GSoC, I will try implementing all the functionality for 3D as the vector package now only supports 3D. I will define a class for the tetrahedron. 

Unfortunately, I am not very familiar with BEM equations but I am looking into them.

Thanks
Faisal Riyaz

Faisal Riyaz

unread,
Mar 30, 2020, 8:55:59 PM3/30/20
to sy...@googlegroups.com
Hello all,

If possible, please provide your feedback on this approach.

Thanks
Faisal Riyaz
Reply all
Reply to author
Forward
0 new messages