[Numpy-discussion] (Late) summary of PEP-225 discussion at Scipy

[Christopher Barker]

Eric Firing wrote:
> I think a stronger general numpy case might be made for the logical 
> operators than for matrix multiplication.  An alternative approach, and 
> I think preferable to introducing new logical operators, is PEP 335: 
> http://www.python.org/dev/peps/pep-0335/.

I like that!

However, it is orthogonal to the additional operators question.

Some thoughts:

I do like the idea of having additional operators, but it seems that the 
strongest use case is matrix operations. While the current matrix 
implementation is clearly inadequate, we had a good discussion a while 
back about how it could be improved, and maybe if we could find a way to 
implement that well, we'd be OK. I always did find Matlab's "dot' 
operators to be kind of annoying (though mostly because what I most 
often need was the "dot" version)


I guess what I'm saying is that adding a bunch of operators just for a 
few matrix operations may not be worth it.

If we can see substantial other uses, then it may be worth it -- are 
there any?

An example is the PEP referenced above -- it would not be nearly as 
strong if numpy were the only motivator -- I think the database example 
is a very strong addition -- is there something similar for additional 
operators?

-Chris







-- 
Christopher Barker, Ph.D.
Oceanographer

Emergency Response Division
NOAA/NOS/OR&R            (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov