(8&-) b._1 NB. normal to think this is weird

8&-

(-&5) b._1 NB. normal to think this is not weird, and normal to think both 
should be consistent.

5&+


What I think is the most natural would be the "radical": 8&+ result for the 
first expression.

(8&(-~)) b._1

8&+


Note that */+ gets it right, and %/- "wrong"

   (8&%) b._1

8&%

(8&%) b._1

(*&8) b._1

0.125&*

(8&*) b._1

0.125&*




On Sunday, October 18, 2020, 06:40:35 p.m. EDT, Jose Mario Quintana 
<[email protected]> wrote: 





I fail to see what the big issue regarding

    8 -inv 5
3

is.

Since  x u^:n y  is equivalent to  (x&u)^:n y  then  8 (-^:_1) 5  is
equivalent to  (8&-^:_1) 5, and the inverse of  8&- is itself, one
can do the math or just ask J,

  (8&-) b._1
8&-

Thus,

  (8&-)5
3

Am I confused?

On Sun, Oct 18, 2020 at 6:03 PM Henry Rich <[email protected]> wrote:
>
> I agree that - : (-~) might be a better obverse.  If there is only one
> inverse, it is for the monad.
>
> I like the idea of more documentation - care to add it to NuVoc?
>
> Henry Rich
>
> On 10/18/2020 6:00 PM, Raul Miller wrote:
> > Yes, that represents the expectation and the implementation. But I am
> > not seeing how there's any inverse here, for the dyadic case.
> >
> >    - b. _1
> > -
> >
> > Personally I would expect either
> >
> > - : +
> >
> > or
> >
> > - : (-~)
> >
> > or some writeup explaining the thinking that makes the current
> > implementation the right approach.
> >
> > Thanks,
> >
> > --
> > Raul
> >
> >
> > On Sun, Oct 18, 2020 at 5:49 PM Henry Rich <[email protected]> wrote:
> >>      (-&5 inv) 8
> >> 13
> >>      (8&- inv) 5
> >> 3
> >>
> >> You are thinking it does the first, but it does the second.
> >>
> >> Henry Rich
> >>
> >> On 10/18/2020 3:27 PM, David Lambert wrote:
> >>> JVERSION
> >>> Engine: j902/j64avx2/linux
> >>> Beta-h: commercial/2020-10-03T19:37:53
> >>> Library: 9.02.05
> >>> Platform: Linux 64
> >>> Installer: J902 install
> >>> InstallPath: /home/lambertdw/downloads/installs/j902
> >>> Contact: www.jsoftware.com
> >>>
> >>>
> >>>    NB. huh?  13 looks right to me.
> >>>    8 -inv 5
> >>> 3
> >>>
> >>>    NB. expected
> >>>    8 *inv 5
> >>> 0.625
> >>>
> >>>    8 +inv 5
> >>> _3
> >>>
> >>> ----------------------------------------------------------------------
> >>> For information about J forums see http://www.jsoftware.com/forums.htm
> >>
> >> --
> >> This email has been checked for viruses by AVG.
> >> https://www.avg.com
> >>
> >> ----------------------------------------------------------------------
> >> For information about J forums see http://www.jsoftware.com/forums.htm

> > ----------------------------------------------------------------------
> > For information about J forums see http://www.jsoftware.com/forums.htm
>
>
> --
> This email has been checked for viruses by AVG.
> https://www.avg.com
>
> ----------------------------------------------------------------------
> 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

Reply via email to