> Why not a middle ground? For some things it would be silly not to > support n dimensions, like dot product, but for others, like cross > product, the n-dimensional generalization is more complicated, and > (if I understand correctly), not even technically a vector. For those > cases, you could give an error if n > 3.
Yes, I implicitly excluded the wedge/cross product from my wish to have vectors in Rn. This is where we can make to border towards geometric/Grassmann/exterior/Clifford Algebra things. (But make sure one can enter that world from vector expressions) Actually, here is another cross product in R7: http://en.wikipedia.org/wiki/Seven-dimensional_cross_product And for R1, R2 we can trivially pad vectors to R3. Maybe one could even fill up R4,R5,R6 to R7? But that I never tried. > Could someone write up a list if everything that would be tricky to > do in n dimensions (rather than 3)? Someone mentioned creating a wiki > page. That would be a good place for this. Actually, I the abstract algebra code & paper I mentioned, all basic axioms of vector algebra are valid in Rn *except* the cross product. > By the way, you say Rn, but is there a reason to not use Cn instead? > Or maybe it won't actually matter for 99% of the code. Oh sure! The base field should not matter. (Probably char 0, maybe not even necessary to assume that.) Anyway, at least R and C should be supported, they should not be hard-coded though. -- 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 [email protected]. To post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/sympy?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
