Multiple types of Coordinate Systems for Vectors

142 views
Skip to first unread message

Arihant Parsoya

unread,
Sep 30, 2016, 12:06:57 PM9/30/16
to sympy
Hi All,

    I wanted to work on the idea of Multiple Coordinate Systems. Previously I submitted my PR for the same here. The old PRs related to this idea uses Lame` coeffecients whereas the GSOC ideas page says that there needs to be multiple classes for each coordinate system. Can anyone guide me on whats the right approach implement this idea?

Thanks,
Arihant Parsoya

Alan Bromborsky

unread,
Sep 30, 2016, 12:49:39 PM9/30/16
to sy...@googlegroups.com
In 3 dimensions there are 13 separable (orthogonal) coordinate systems.  See link -
How many do you want to implement?  I would have a coordinate_system class and use it to instantiate a particular coordinate system then when you  instantiate a vector space one of the parameters of  the vector space  instantiation would be the coordinate system.  This way you could have different coordinate systems in the same program. 

For my geometric algebra modules I started with a metric and derived the basis vectors and their derivatives.  If you are interested in this method go to -

github.com/brombo/galgebra

and look at galgebra.pdf in the doc directory.

--
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+unsubscribe@googlegroups.com.
To post to this group, send email to sy...@googlegroups.com.
Visit this group at https://groups.google.com/group/sympy.
To view this discussion on the web visit https://groups.google.com/d/msgid/sympy/33e8d677-92c1-4ec8-b956-54690bd65ba9%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Arihant Parsoya

unread,
Oct 1, 2016, 1:10:16 AM10/1/16
to sympy
Hi Brombo,

I went through galgebra.pdf file and I believe this approach would be good. Initially I want to implement Cartesian and Spherical coordinate systems and get them right before implementing other systems.


On Friday, September 30, 2016 at 10:19:39 PM UTC+5:30, brombo wrote:
In 3 dimensions there are 13 separable (orthogonal) coordinate systems.  See link -
How many do you want to implement?  I would have a coordinate_system class and use it to instantiate a particular coordinate system then when you  instantiate a vector space one of the parameters of  the vector space  instantiation would be the coordinate system.  This way you could have different coordinate systems in the same program. 

For my geometric algebra modules I started with a metric and derived the basis vectors and their derivatives.  If you are interested in this method go to -

github.com/brombo/galgebra

and look at galgebra.pdf in the doc directory.
On Fri, Sep 30, 2016 at 12:06 PM, Arihant Parsoya <parsoya...@gmail.com> wrote:
Hi All,

    I wanted to work on the idea of Multiple Coordinate Systems. Previously I submitted my PR for the same here. The old PRs related to this idea uses Lame` coeffecients whereas the GSOC ideas page says that there needs to be multiple classes for each coordinate system. Can anyone guide me on whats the right approach implement this idea?

Thanks,
Arihant Parsoya

--
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.

Alan Bromborsky

unread,
Oct 1, 2016, 10:19:09 AM10/1/16
to sy...@googlegroups.com
If you use the methods in galgebra.pdf remember (this is documented) that for all but rectangular coordinates the basis vectors derived from the metric are not normalized. Especially go over section 2.3.2 is galgebra.pdf.

To unsubscribe from this group and stop receiving emails from it, send an email to sympy+unsubscribe@googlegroups.com.

To post to this group, send email to sy...@googlegroups.com.
Visit this group at https://groups.google.com/group/sympy.

Arihant Parsoya

unread,
Dec 11, 2016, 8:53:06 AM12/11/16
to sympy
Hi,

I am trying to get myself acquainted with geometric algebra by following the book `Geometric Algebra for Physicists` (by Chris Doran). It would be very helpful if you can suggest some other readings which can be helpful in implementing the class `coordinate_system`.

Thanks

Alan Bromborsky

unread,
Dec 11, 2016, 11:09:33 AM12/11/16
to sy...@googlegroups.com
My notes base on "Geometric Algebra for Physicists" are at -

https://github.com/brombo/galgebra

Look at bookGA.pdf in the doc directory.

To unsubscribe from this group and stop receiving emails from it, send an email to sympy+unsubscribe@googlegroups.com.

To post to this group, send email to sy...@googlegroups.com.
Visit this group at https://groups.google.com/group/sympy.

Sourabh Garg

unread,
Dec 12, 2016, 12:32:44 AM12/12/16
to sympy
Hey,

I also want to work on it. if you have any link to book's pdf version ,please share with me.
Thanks

Sourabh Garg

Alan Bromborsky

unread,
Dec 12, 2016, 11:55:18 AM12/12/16
to sy...@googlegroups.com
Look at my notes.  They may be sufficient.

--
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+unsubscribe@googlegroups.com.
To post to this group, send email to sy...@googlegroups.com.
Visit this group at https://groups.google.com/group/sympy.

Arihant Parsoya

unread,
Jan 11, 2017, 2:07:35 PM1/11/17
to sympy
Hi, 

I have submitted a PR which implementes classes CoordinateSystem (Super class for remaining coordinate systems), CartesianCoordinateSystem and SphericalCoordinateSystem  using Metric and Lamè coeffecients here. Can you review it?

Thanks.

On Monday, December 12, 2016 at 10:25:18 PM UTC+5:30, brombo wrote:
Look at my notes.  They may be sufficient.
On Mon, Dec 12, 2016 at 12:32 AM, Sourabh Garg <sourabh.8...@gmail.com> wrote:
Hey,

I also want to work on it. if you have any link to book's pdf version ,please share with me.
Thanks

Sourabh Garg

--
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 post to this group, send email to sy...@googlegroups.com.
Visit this group at https://groups.google.com/group/sympy.

Arihant Parsoya

unread,
Jan 11, 2017, 2:19:02 PM1/11/17
to sympy

Arihant Parsoya

unread,
Feb 12, 2017, 2:56:18 PM2/12/17
to sympy
Hi Brombo,

Can you suggest some API specification for the new coordinate system?

Thanks

Francesco Bonazzi

unread,
Feb 13, 2017, 4:26:52 AM2/13/17
to sympy


On Sunday, 12 February 2017 20:56:18 UTC+1, Arihant Parsoya wrote:
Hi Brombo,

Can you suggest some API specification for the new coordinate system?

Have you had a look at sympy.diffgeom, there is a CoordSystem class and some tools to connect different kinds of coordinate systems. Unfortunately it's a bit different than the vector module.
Reply all
Reply to author
Forward
0 new messages