Thanks, Tim, as usual it was just my terrible sense of direction.
On Wednesday, August 27, 2014 7:02:42 PM UTC-4, Tim Holy wrote:
>
> To be perfectly clear about the actual behavior: if A is of size (m,n,p),
> then
> sum(A, (1,3)) has size (1,n,1). So it's summing over dimensions 1 and 3.
>
> -
To be perfectly clear about the actual behavior: if A is of size (m,n,p), then
sum(A, (1,3)) has size (1,n,1). So it's summing over dimensions 1 and 3.
--Tim
On Wednesday, August 27, 2014 03:02:09 PM Andrew Dabrowski wrote:
> OK, thanks. I find the phrase "over the given dimensions" ambiguous -
Yeah, it's not a great phrase. You can pretty easily submit doc fix PR's on
GitHub if you've got better language in mind. Maybe call it summing "along" a
dimension?
-- John
On Aug 27, 2014, at 3:02 PM, Andrew Dabrowski wrote:
> OK, thanks. I find the phrase "over the given dimensions" ambig
OK, thanks. I find the phrase "over the given dimensions" ambiguous -
isn't it actually summing over the dimensions _not_ given?
On Wednesday, August 27, 2014 5:46:33 PM UTC-4, John Myles White wrote:
>
> sum(A, 1) works, as does sum(A, (1, )).
>
> More generally, sum(A, dims::Integer...) works
sum(A, 1) works, as does sum(A, (1, )).
More generally, sum(A, dims::Integer...) works, as does as sum(A,
(dims::Integer...), ).
-- John
On Aug 27, 2014, at 2:44 PM, Andrew Dabrowski wrote:
> In the doc for the Standard library I see:
>
> sum(A, dims)
> Sum elements of an array over the giv
In the doc for the Standard library I see:
sum(*A*, *dims*)
Sum elements of an array over the given dimensions.
In exactly what form should "dims" be given?
