There be dragons! Explicit threading was hacked into the code but
never finished as Tuomas (the author) moved on to other projects
before it was debugged, and no-one else has felt the need to hack on
it to get it working. I think it has been left so long because (A)
if anyone ever did get it working, it would be a nice thing to have,
and (B) it would take a certain amount of effort to disentangle and
remove the explicit threading calls.
Only implicit threading (via dimlist positional order) should be
considered useful.
On Feb 23, 2007, at 5:01 PM, Derek Lamb wrote:
I think I am not understanding something here:
perldl> $a = sequence(5,4,3);
perldl> $b = $a->thread(1);
perldl> ?$b
This variable is Double D [5,3] T1 [4] -C 0.00Kb
perldl> $c = $b->sumover; #the middle dimension
Segmentation fault
Normally I wouldn't care, and would just chalk it up to not
understanding something. But a very similar example is right out
of the thread() docs. As far as I can tell (I haven't mushed my
brain on the Indexing manpage too much), the example I just listed
should have been the same as $c=$a->mv(1,0)->sumover;
I'm trying to work on the PDL::Gaussian bug that Adam reported last
week, and this came up.
Any ideas about what I am doing wrong?
Derek
_______________________________________________
Perldl mailing list
[email protected]
http://mailman.jach.hawaii.edu/mailman/listinfo/perldl
_______________________________________________
Perldl mailing list
[email protected]
http://mailman.jach.hawaii.edu/mailman/listinfo/perldl
