sum(p, dims) already has a meaning: reduce over multiple dimensions 
simultaneously. Consequently, adding brackets around the dimensions is not 
available for indicating "please keep this dimension."

Best,
--Tim

On Friday, May 27, 2016 12:42:42 PM CDT Gustavo Goretkin wrote:
> sum(p, 1:2) should be the same as sum(sum(p, [1]), [2]), right?
> 
> On Fri, May 27, 2016 at 11:54 AM, a. kramer <[email protected]> wrote:
> > If you are proposing that sum(p,1) should drop dimensions but sum(p,[1])
> > should not, I don't think there is an easy solution like that.  What
> > should
> > sum(p,1:2) do?
> > 
> > On Friday, May 27, 2016 at 6:52:58 AM UTC-4, Christoph Ortner wrote:
> >> On Thursday, 26 May 2016 03:03:03 UTC+1, Tim Holy wrote:
> >>> Since you asked...from my perspective, the easiest argument against it
> >>> is
> >>> 
> >>>     pnormalized = p ./ sum(p, 1)
> >>> 
> >>> If you drop the summed dimension, that won't work anymore. One can write
> >>> 
> >>>     pnormalized = p ./ reshape(sum(p, 1), 1, size(p, 2))
> >> 
> >> Would the point of the new notation be to write
> >> 
> >>      pnormalized = p ./ sum(p, [1])


Reply via email to