On Aug 19, 12:26 pm, Jason Grout <jason-s...@creativetrax.com> wrote: > On 8/19/11 1:45 PM, Robert Bradshaw wrote: > > > [a, b; c, d].change_ring(QQ)
You'd have to take care that [1.000000000000000000000001, 1.0; 2.0, 3.0].change_ring(RealField(100)) doesn't lose precision on the way, which suggests that "[;]" syntax should result in a "matrix literal". > What if I want a 1-row matrix. Will this work? > > [a,b,c,d] In analogy to (a,) being a singleton, this should probably be [a,b,c,d;] > But then what about: > > [1, 2, > 3, 4] > > From the previous messages (where newlines are treated like > semicolons), that should be a square 2x2 matrix, but again, it is valid > python syntax. Hence, I don't think a preparser based solution should support this. Saving the "..." comes at the cost of having to type the ;s Also note that once we're doing this, we should probably also support matrices over matrices, so [ [1,2;3,4] , [5,6;7,8]; [9,10;11,12], [13,14;15,16]] should also be supported (just because your grammar specification will be a horribly ugly mess if you want to disallow this construct). This basically means that the preparser actually has to parse its input (http://en.wikipedia.org/wiki/ Pumping_lemma_for_regular_languages can be used that parenthesis matching cannot be done using regular languages, although some modern "regex" implementation might have features for it). Matrices over matrices are weakly supported presently: sage: M=matrix(2,2,[1,2,3,4]) sage: A=matrix(2,2,[M,M,M,M]) sage: B=A^2 sage: B[0,0] [14 20] [30 44] sage: M*M+M*M #that's what the entry should be [14 20] [30 44] but pretty much anything else fails (including printing these objects, because printing apparently needs hash(M)) Independent of the example above, nested matrix literals could happen anyway: [ 1, 2; 3, det( [4,1;0,1]) ] -- 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
