I've tackled #1 on my original list, the subject of the most discussion here. Found a variety of bugs in the process - caching mistakes and the problem with matrices over QQbar discussed elsewhere.
My approach to eigenspaces and Galois conjugates was to introduce a "format" keyword with possible values of 'all' and 'galois'. The former attempts to locate eigenvalues in the base field, or extend to QQbar. When this fails, the exception message mentions the 'galois' option. This latter option preserves current behavior. The change in current behavior is that the 'all' option would become the default. So newbies could see QQbar elements that look familiar by default, pros can chose Galois conjugates when required (and will discover this possibility in an exception message). Two relevant comments from having my head in the code: (1) outside of the eigen-stuff code, the only place it was necessary to supply the 'galois' option was in the method for eigenspaces of a graph, (2) looking at doctests and code for the other eigen-routines, the more I felt this change would make for more consistency in the output across routines. http://trac.sagemath.org/sage_trac/ticket/11595 -- To post to this group, send an email to [email protected] To unsubscribe from this group, send an email to [email protected] For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org
