Yeah, it always looks a bit scary to do X =: A I}X to modify only a small
part of X, but at the end of
https://www.jsoftware.com/help/jforc/modifying_an_array_m.htm
we are promessed that J does that in place.
I think this should not change when A depends on what was already there, in
I{X, as is very often the case. Of course it is important whether you agree
:-)
In your example you're moving things somewhere else inside the array. I
sort of understand it can be tricky to do that in-place when the source and
the destination overlap.
On Wed, 13 Nov 2019 at 14:43, Henry Rich <[email protected]> wrote:
> I should say, you can't count on the JE to make the best decision
> unaided. If you write
>
> big =: i. 1e4 1e4
> big =: (1 { big) 2} big
>
> it will not run inplace & will be very slow. Why? Because (1 { big)
> will produce a virtual result, i. e. a noun whose data is mapped to a
> subarray of another noun (viz big).
>
> Having the virtual result extant makes big ineligible for inplacing, so
> the 2} requires copying big. The JE can't look ahead to see that it
> should forego inplacing { for the greater good of inplacing big} .
>
> In this case, you must assign (1 { big) to a name, which will realize
> the value and remove the restraint on inplacing big} .
>
> Henry Rich
>
>
> On 11/12/2019 6:32 PM, Ben Gorte wrote:
> > Great, thanks!
> >
> > Greetings,
> > Ben
> >
> > On Wed, 13 Nov 2019 at 02:54, Henry Rich <[email protected]> wrote:
> >
> >> Fixed for next release. The problem was that (x{M) created a temporary
> >> in-place noun to refer to atoms of M; this raised the usecount of M
> >> rendering it uninplaceable for the amend (}). The in-place noun should
> >> have been released right after it was used, but erroneously it was not
> >> released until the end of the sentence.
> >>
> >> The workaround is to do the fetch from M in a different sentence. And do
> >> try those forms that avoid boxing.
> >>
> >> Henry Rich
> >>
> >> On 11/12/2019 2:17 AM, Ben Gorte wrote:
> >>> Dear J'ers
> >>>
> >>> I'm using large 3D arrays (voxels), such as
> >>> M =: ?400 400 400$100
> >>>
> >>> To add 1 to a bunch of voxels at positions
> >>>
> >>> I =: ?100 3$400
> >>>
> >>> I go like:
> >>>
> >>> timespacex 'M =: (1+(;/I){M) (;/I) } M'
> >>>
> >>> 1.3e_5 42048
> >>>
> >>> This was in version 8.07:
> >>>
> >>> JVERSION
> >>>
> >>> Engine: j807/j64/linux
> >>>
> >>> Release: commercial/2018-10-04T15:00:41
> >>>
> >>> Library: 8.07.26
> >>>
> >>>
> >>> In version 9.01, however: timespacex 'M =: (1+(;/I){M) (;/I) } M'
> >>>
> >>> 0.238664 5.36912e8
> >>>
> >>>
> >>> It became thousands of times slower! I'm doing this hundreds of
> thousands
> >>> of times and it takes weeks instead of minutes.
> >>>
> >>>
> >>> A multi-dimensional index for amend does not need to be boxed in j901,
> >> but:
> >>> timespacex 'M =: (1+(;/I){M) I } M'
> >>>
> >>> 0.238302 5.36894e8
> >>> gives pretty much the same result.
> >>> This is with:
> >>>
> >>> JVERSION
> >>>
> >>> Engine: j901/j64avx2/linux
> >>>
> >>> Beta-p: commercial/2019-11-06T09:48:25
> >>>
> >>> Library: 9.01.10
> >>>
> >>> I was wondering whether it happend in { or }, but to my great surprise:
> >>> timespacex 'new =: 1+(;/I) { M'
> >>>
> >>> 8e_6 21824
> >>> timespacex 'M =: new (;/I) } M'
> >>>
> >>> 6e_6 20992
> >>>
> >>> So the solution is to do it in two steps !?!?
> >>>
> >>> With unboxed I in step 2, I get even:
> >>>
> >>> timespacex 'M =: new I } M'
> >>>
> >>> 2e_6 2560
> >>>
> >>>
> >>> Am I missing a point? Could I have foreseen this?
> >>>
> >>> Greetings,
> >>> Ben
> >>> ----------------------------------------------------------------------
> >>> For information about J forums see http://www.jsoftware.com/forums.htm
> >> ----------------------------------------------------------------------
> >> For information about J forums see http://www.jsoftware.com/forums.htm
> >>
> > ----------------------------------------------------------------------
> > For information about J forums see http://www.jsoftware.com/forums.htm
>
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm
>
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
