On Thu, Feb 26, 2009 at 10:40 PM, Rob Beezer <goo...@beezer.cotse.net> wrote: > > Hi David, > > Thanks for your comments. I'm very interested in making Sage more > approachable for students, so I'm all for anything that improves that, > without hampering research applications. > > In my linear algebra textbook, I've been dogmatic about every vector > being a column vector, even going so far as constructing row spaces as > sets of column vectors. Definitely over the top, and arguably wrong. > But I've been trying to spare the students some initial confusion > about rows and columns and when to transpose, and when not to. > > Would it be going too far to suggest vectors could be tagged as "row" > or "column" (or None)? A right kernel would return a module/subspace > with basis vectors that were tagged as columns and a left kernel would > return a module/subspace with basis vectors that were tagged as rows. > A column space would have columns as basis vectors and a row space > would have basis vectors that are rows. I could then see sanity > checks other places, like the product of a matrix with a vector on the > right or left could check for the vector being a column or a row. An > inner product could require a row vector, then a column vector, in > that order as arguments. A transpose method on vectors would just > toggle the tag. And so on. These checks could be phased in, choosing > where to query the tag and where to respect the None tag. I could see > a payoff in preventing some incompatible operations in user-level code > with clear messages (eg "attempting to multiply a column vector on the > left of a matrix"). > > Would this be too disruptive in other places?
It was my impression that your proposal above is identical to the one that David Kohel just proposed. > I hope I've understood your comments correctly and haven't run off too > far in the wrong direction. I think you have. --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to sage-devel-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
