On Dec 30, 2013, at 6:31 PM, Karl Glazebrook <[email protected]> wrote:
> I'd be interested to hear you expand on your comments Doug as to why PDL is 
> technically superior. I guess you are referring to the threading? I don't 
> know much about scipy though I gather it has something like that? Maybe we 
> should have a document on PDL's unique features.

I have only explored numpy/scipy a little, but I hold PDL to have an objective 
advantage in both threading and in ease of tying in outside codes.  Threading 
is orthogonal, elegant, and very, very powerful - although my experience is 
limited, I don't think scipy has equivalent tools - their vectorization is more 
like IDL's, if I understand right.  PP, Inline::Pdlpp, Inline::C, and 
Inline::F77 make it a snap to drop down to compiled code or link in external 
libraries.  (That's even before David's in-progress stuff on tcc reaches 
fruition.)  Again, I could be missing something but it appears Python still 
requires producing a several-file library to link in the simplest of C programs.

There are a number of smaller things that could just be subjective and that 
boil down to the religious war over the host language -- Perl makes it very 
easy to do a lot of things (like object manipulation) that I find difficult or 
extremely awkward in Python.  But enough people disagree on those issues in 
enough familiar ways that I can't really count them to Perl's or PDL's 
advantage over Python or Numpy.





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

Reply via email to