On Feb 4, 11:30 am, Jason Grout <jason-s...@creativetrax.com> wrote:
> I was thinking of confusion for the user, because the user has to be > more careful about how they call the function because the arguments > change. Now, this can be alleviated by the user using keyword > arguments, or paying attention and being aware of integration removing a > variable. Again, I don't think this is insurmountable, and arguably > it's desirable to make the user think about how the function domain is > changing, but I feel obligated to bring up the issue as a point against > changing things in a backwards-incompatible way. > > Thanks, > > Jason I think it is much more confusing to supply an extra argument (whose value does not matter at all!) then to think about change in the domain. From a mathematical viewpoint it seems unambiguous to me that the current behaviour is wrong and should be changed, and so far there were no implementation reasons for keeping it. If dropping a variable and allowing unnamed evaluation is confusing, it may not be allowed as well, but I personally think that it is quite desirable. I hit it while preparing for my multivariate calculus class where I teach students that taking a function f(x,y) over some region R and integrating it in y from g(x) to h(x) gives the area of the cross- section A(x). If we integrate it in x from u(y) to v(y), we get a perpendicular cross-section B(y). If we now integrate one more time, we get a number V - the volume under the graph. Sage output suggest that we are still getting functions of (x,y) - still over R or what is its new domain?.. Why would any user try to evaluate, say, V(2,1) and what would be the expected result? Would this user expect V(3,42) return the same value? If yes, how about just using V with no arguments? If no - the user does not seem to understand what is going on. Raising an error for an attempt to evaluate V(2,1) can be a very useful cue in this case. So - I propose opening a ticket to fix this. Thank you, Andrey -- 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
