Ralf Hemmecke wrote:
>
> > We also have "sparse" series domains, you can try them.
>
> Oh... how can I be sure from the documentation that I don't run into the
> same problem?
Unfortunately, no warranty in documentation...
> >From my understanding of how a "sparse" series should be implemented, my
> problem will probably not occur. But should a user of the Algebra
> library be required to look into the code to find that out. Shouldn't
> such information be given in the documentation?
I do not think so. I spent some time studing program specification.
One nice observation is that programs and specifications are
essentially the same object: you can treat program as a logical
formula and you can treat arbitrary logical formula as a
(generalized) program. The difference is in detail: real
programs are quite specific while specifications leave many
things unspecified. In particular trying to execute
specifications frequently leads to horribly inefficient
programs. But once you specify various details like
using sparse representation practical differences between
specifications and programs tend to vanish. In fact there
is programing methodology which starts from specification
and then adds details (refines specification) to arrive
at practical program.
So AFAICS art of specification and documentation is largely
about omiting information. It makes sense to give some
intutions like "this domain uses dense representation"
or "this domain uses sparse representation", but trying
to give precise performance warranty would go too far.
--
Waldek Hebisch
--
You received this message because you are subscribed to the Google Groups
"FriCAS - computer algebra system" 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 https://groups.google.com/group/fricas-devel.
For more options, visit https://groups.google.com/d/optout.
