On 7/31/13 9:39 AM, Robert Dodier wrote:
> On 2013-07-31, Barton Willis <willisb at unk.edu> wrote:
>
>> No, it's not possible. Both summation and integration ignore declared
>> dependencies (depends(A,i), for example). The user documentation for
>> integration mentions this; the documentation for sum doesn't.
> That's a bug, right? Probably relatively easy to fix too.
easy to change, probably wrong. Here's why...
Which i might that be?
Wrong scope perhaps if one has lexical scoping, anyway.
If one believes sum(a(i),i,1,n) means something like
sumfun ( lambda([i], a(i)), 1,n) then it is equivalent to
sumfun ( lambda([z],a(z)), 1,n).
but if you do depends(a,i), then
sumfun( lambda([i],a ) 1,n) is NOT equivalent to
sumfun( lambda([z],a ) 1,n).
So "fixing" the "bug" would in fact be introducing anomalous and
perhaps buggy
behavior.
There is also generally the issue of the extent to which such
functions as sum, product,plot, and even integrate should evaluate
their arguments.
RJf
>
> best
>
> Robert Dodier
>
> _______________________________________________
> Maxima mailing list
> Maxima at math.utexas.edu
> http://www.math.utexas.edu/mailman/listinfo/maxima