PS. Also, please advertise this project on gap-forum and gap-support mailing lists. I am sure there will be more people interested in helping out if you explain that potentially one would be able to seamlessly call GAP functions from Python...
Dima On Mar 9, 7:58 am, Mike OS <mosul...@math.sdsu.edu> wrote: > I have some funding from my university to develop materials in SAGE > for use in my classes. The focus of the project is developing > educational materials but we'd also like to contribute to SAGE > development. I posted to sage-edu about educational issues. This > post concerns development, specifically groups in SAGE and the > relationship with GAP. I've hired two students, Rhea Morton and > Ryan Rosenbaum to work on the project, and we would like some > guidance on how we can contribute. > > Some observations: > > A. William started a libgap, analogous to libsing, but it has lain > fallow for a while. As we understand things, there is a somewhat > different challenge with GAP. Much of Singular's functionality is > written in C so that SAGE can access it via Cython, whereas much of > GAP is written in GAP's own language, so pexpect is required to access > it. How does this affect goals for libgap? Does this mean that > libgap will be inherently limited in efficiency or functionality? > > B. The class Group seems to only loosely combine the different > types of groups: permutation groups, abelian groups and matrix groups. > For example, subgroup is not implemented for matrix groups. > Also, the unit group of a ring (like U_n= Z_n*) and finitely > presented > groups aren't implemented (that we coud find). > Should these be implemented in SAGE via GAP or have their own > existence in SAGE? What are the goals for improvement of the Group > functionality? > > C. Rhea is preparing a homomorphism package that exposes GAP's group > homomorphism functions. In order to work it requires SAGE groups with > GAP compatible generators. It doesn't work with the Integer ring, for > example, because the Integer class doesn't extend from sage.groups. In > addition, it works closely with the FreeGroup class so both would have > to be submitted for it to be of much utility. Rhea has tested it with > the FreeGroup class, SAGE DihedralGroup, and SAGE SymmetricGroup. > The documentation and test cases are still being prepared. > > Comments? -- To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org