>>>>> Chris Marshall <[email protected]> writes:

 >> Now there, my question is: does it seem feasible to add semantic
 >> information to the PDL dimensions?

[…]

 >> Also, there's a question on how should the behavior be altered in
 >> presence of the semantically-tagged variables?

[…]

 > This has been discussed somewhat in relationship to generalized
 > coordinate system transformations.  It is possible to subclass PDL to
 > implement the features you desire.  See the PDL::Objects for a start.

        ACK, thanks!  (Going to think on the thing for some time.)

        Thanks to the ability to overload ->copy (), I guess that it'll
        be quite simple to preserve the “meaning” of the indices when
        the data is passed through the simpler functions.

        However, if, e. g., I wish for an exception to be generated
        should “incompatible” arrays be involved into an arithmetic
        operation (like $a + $b, where the 0'th dimensions of $a and $b
        correspond to, e. g., Unix time and atmospheric pressure), will
        it be necessary to overload nearly all the non-unary operations?

        (And it doesn't seem anything like a sane task, since new
        methods just like to appear out of nowhere at the occasions
        unforseen.)

-- 
FSF associate member #7257

Attachment: pgp1i5tk8tcDU.pgp
Description: PGP signature

_______________________________________________
Perldl mailing list
[email protected]
http://mailman.jach.hawaii.edu/mailman/listinfo/perldl

Reply via email to