On 11/5/2010 3:59 PM, Edwin Woollett wrote:
> I am attaching aslaksen1.pdf and
> aslaksen2.pdf, which describe a
> minimal set of cas tests for behavior in the presence of
> complex numbers. They supplement
> the Michael Wester tests.
>
> Ted
>
>
> _______________________________________________
> Maxima mailing list
> Maxima at math.utexas.edu
> http://www.math.utexas.edu/mailman/listinfo/maxima
Unfortunately it seems that the author repeats the not-very-useful
assertion that the proper branch of some function F is determined
by its argument. e.g. F(z) for real z should be ..... .
whereas much analysis requires that the proper branch of F be
chosen by some context in which F appears.
These papers do tend to point out that some obvious identities
on the real line are not true in the complex plane, but all it
says of this situation is that your program shouldn't simplify
things. This is too is not so helpful, because certain
expressions DO simplify somewhat.
Overall, not very helpful if you are trying to write a program
to do what a clever human might do.
RJF