Julian, I can see the need to recognize both column and row vectors, but why not with np.matrix? I can see no need for a new operator and hope to be able to comment more fully on PEP 465 in a few days. Colin W. On 17-Mar-2014 7:19 PM, numpy-discussion-request@scipy.org wrote:

Send NumPy-Discussion mailing list submissions to numpy-discussion@scipy.org

To subscribe or unsubscribe via the World Wide Web, visit http://mail.scipy.org/mailman/listinfo/numpy-discussion or, via email, send a message with subject or body 'help' to numpy-discussion-request@scipy.org

You can reach the person managing the list at numpy-discussion-owner@scipy.org

When replying, please edit your Subject line so it is more specific than "Re: Contents of NumPy-Discussion digest..."

Today's Topics:

1. Re: [help needed] associativity and precedence of '@' (Nathaniel Smith) 2. Re: GSoC project: draft of proposal (Julian Taylor) 3. Re: [help needed] associativity and precedence of '@' (Christophe Bal) 4. Re: [help needed] associativity and precedence of '@' (Alexander Belopolsky) 5. Re: [help needed] associativity and precedence of '@' (Bago) 6. Re: [help needed] associativity and precedence of '@' (Christophe Bal) 7. Re: [help needed] associativity and precedence of '@' (Christophe Bal) 8. Re: [help needed] associativity and precedence of '@' (Nathaniel Smith)

----------------------------------------------------------------------

Message: 1 Date: Mon, 17 Mar 2014 22:02:33 +0000 From: Nathaniel Smith <njs@pobox.com> Subject: Re: [Numpy-discussion] [help needed] associativity and precedence of '@' To: Discussion of Numerical Python <numpy-discussion@scipy.org> Message-ID: <CAPJVwB=zBazN+fiYWJeiWOL=4a9Bf2XGxJGoTT8GFTt-kDUDZw@mail.gmail.com> Content-Type: text/plain; charset=UTF-8

On Mon, Mar 17, 2014 at 9:38 PM, Christophe Bal <projetmbc@gmail.com> wrote:

...

Here is the translation. ;-)

Hello, and what about something like that ?

a @ b @ c -> (a @ b) @ c a * b @ c -> (a * b) @ c a @ b * c -> a @ (b * c)

Easy to remember: the *-product has priority regarding to the @-product, and we just do @-product from left to right. In the terminology we've been using in this thread, this is "weak-left".

...

An advantage of this is that most parsers do analyze from left to right.

So I really think that it is a better choice than the weak-right one. We've mostly ignored this option because of assuming that if we want left-associativity, we should go with "same-left" instead of "weak-left". Same-left is:

a @ b @ c -> (a @ b) @ c a * b @ c -> (a * b) @ c a @ b * c -> (a @ b) * c

i.e., even more left-to-right than weak-left :-)

Do you think weak-left is better than same-left?