[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Efficient code for operating on pairwise vector diffs
From: |
Joseph Wakeling |
Subject: |
Re: Efficient code for operating on pairwise vector diffs |
Date: |
Sun, 08 Nov 2009 22:41:00 +0100 |
User-agent: |
Thunderbird 2.0.0.23 (X11/20090817) |
Jaroslav Hajek wrote:
> Hi,
> sorry for the late reply. For an OctaveForge file, I think you should
> write an Octave-like inline docstring. Error checks (number&type of
> arguments) would also be nice. The dim argument also sounds good.
> If you target just Octave, the exist("bsxfun") check is pretty unnecessary.
Sure, that sounds good.
Just on a personal level, are you happy with the author credit in the
code? It felt slightly cheeky saying '(c) me' and only thanking you --
it seemed OK in the original longer piece of code I wrote using bsxfun
on one line, but less so in a function where your suggestions are the
majority of the working lines :-)
Brief overview of what I'd like to do with the complete code this is
part of. We're preparing code to be both Octave- and MATLAB-compatible
(which is why the bsxfun check is in there, only the most recent MATLAB
has bsxfun). I'll probably put it onto Launchpad or GitHub as something
stand-alone to be maintained, with a permissive but GPL-compatible
license (BSD, I guess). Then octave-forge stuff can be derived from it.
Does this sound OK?
Again, sorry for not being more forthcoming with actual code examples
right now, but it will get there. :-)
Best wishes,
-- Joe