Like this?
({.@ni n0`n1`n2@.({:@ni) ]) ;:2}.ex1
┌─┬──┬──┬──┬──┬──┬──┐
│N│NW│NW│NE│NE│SW│SW│
└─┴──┴──┴──┴──┴──┴──┘
Or, if you do not like running ni twice:
(({.@] n0 [)`({.@] n1 [)`({.@] n2 [)@.({:@]) ni) ;:2}.ex1
┌─┬──┬──┬──┬──┬──┬──┐
│N│NW│NW│NE│NE│SW│SW│
└─┴──┴──┴──┴──┴──┴──┘
I hope this helps,
--
Raul
On Fri, Dec 15, 2017 at 5:57 PM, Brian Schott <[email protected]> wrote:
> I have been experimenting with code to implement the parsing of Jimmy's
> idea and example.
> However, I am stuck on an application of agenda and would like some help.
>
> ni =: [:(<./,]i.<./)(;:'S SE SW') i.~ ]
> n0 =: a:-.~(a:"_)`(0,{.@[)`]}
> n1 =: a:-.~(('NE';a:)"_)`(0,{.@[)`]}
> n2 =: a:-.~(('NW';a:)"_)`(0,{.@[)`]}
> na =: n0`n1`n2@.{:@[ NB. not used in the explicit verb that works!
>
> ex1 =: 'SE NE N NW NW NE NE S SW SW'
>
>
> The following verb named choose does what I want, but I would like to use
> agenda instead because agenda seems more tacit. I have only tested choose
> and
> my intended, but faulty, code with Jimmy's case 0 = N + S, but I hope to
> generalize and expand it.
>
> choose =: monad define
> ch =. ni y
> select. {: ch
> case. 0 do. return =. ({.ch)n0 y
> case. 1 do. return =. ({.ch)n1 y
> case. 2 do. return =. ({.ch)n2 y
> end.
> )
>
> choose 2}.;:ex1 NB. the desired result!
> ┌──┬──┬──┬──┬──┬──┐
> │NW│NW│NE│NE│SW│SW│
> └──┴──┴──┴──┴──┴──┘
>
>
> NB. I'm trying to operate on the 0=N+S
> NB. example
> ni 2}.;:ex1
> 5 0
> 5 0 n0 2}.;:ex1 NB. This produces my expected result
> ┌──┬──┬──┬──┬──┬──┐
> │NW│NW│NE│NE│SW│SW│
> └──┴──┴──┴──┴──┴──┘
> (ni na ]) 2}.;:ex1
> |rank error: n0
> | (ni na])2}.;:ex1
>
> My problem is not being able to use the agenda verb na
> to produce the gerund forms in amend verbs n0, n1 and n2.
>
>
> On Wed, Dec 13, 2017 at 4:22 PM, Jimmy Gauvin <[email protected]>
> wrote:
>
>> HI,
>>
>> I have been thinking about Brian's "cancellation" approach and maybe we can
>> sort of formalize it.
>>
>> Associative relations for cancelling paths:
>>
>> 0 = N + S = S + N
>> 0 = NE + SW = SW + NE
>> 0 = NW + SE = SE + NW
>>
>> Associative relations for simplifying paths:
>>
>> N = NE + NW = NW + NE
>> S = SE + SW = SE + SW
>> NE = N + SE = SE + N
>> NW = N + SW = SW + N
>> SE = S + NE = NE + S
>> SW = S + NW = NW + S
>>
>> Please note that destinations are equivalent but the paths are different.
>>
>> Now we can do symbolic algebraic simplification.
>>
>> For example :
>>
>> ┌──┬──┬─┬──┬──┬──┬──┬─┬──┬──┐
>> │SE│NE│N│NW│NW│NE│NE│S│SW│SW│
>> └──┴──┴─┴──┴──┴──┴──┴─┴──┴──┘
>>
>> N = NW + NE and SE = NE + S
>> ┌──┬──┬─┬──┬──┬──┬──┬─┬──┬──┐
>> │SE│NE│N│NW│N │ │SE│ │SW│SW│
>> └──┴──┴─┴──┴──┴──┴──┴─┴──┴──┘
>>
>> S = SE + SW
>> ┌──┬──┬─┬──┬──┬──┬──┬─┬──┬──┐
>> │SE│NE│N│NW│N │ │S │ │ │SW│
>> └──┴──┴─┴──┴──┴──┴──┴─┴──┴──┘
>>
>> 0 = N + S
>> ┌──┬──┬─┬──┬──┬──┬──┬─┬──┬──┐
>> │SE│NE│N│NW│ │ │ │ │ │SW│
>> └──┴──┴─┴──┴──┴──┴──┴─┴──┴──┘
>>
>> 0 = SE + NW
>> ┌──┬──┬─┬──┬──┬──┬──┬─┬──┬──┐
>> │ │NE│N│ │ │ │ │ │ │SW│
>> └──┴──┴─┴──┴──┴──┴──┴─┴──┴──┘
>>
>> 0 = NE + SW
>> ┌──┬──┬─┬──┬──┬──┬──┬─┬──┬──┐
>> │ │ │N│ │ │ │ │ │ │ │
>> └──┴──┴─┴──┴──┴──┴──┴─┴──┴──┘
>>
>> The count of remaining symbols is the distance in number of steps.
>>
>> Programming this is left as an exercise to the reader.
>>
>>
>> On Tue, Dec 12, 2017 at 3:08 PM, Brian Schott <[email protected]>
>> wrote:
>>
>> > No. That's not right. I need to repair my approach.
>> >
>> > On Tue, Dec 12, 2017 at 2:04 PM, Brian Schott <[email protected]>
>> > wrote:
>> >
>> > > Another way to approach this problem first part, which was refined by
>> > > Jimmy's
>> > > elegant solution, follows.
>> > >
>> > >
>> > >
>> > ----------------------------------------------------------------------
>> > For information about J forums see http://www.jsoftware.com/forums.htm
>> >
>> ----------------------------------------------------------------------
>> For information about J forums see http://www.jsoftware.com/forums.htm
>>
>
>
>
> --
> (B=) <-----my sig
> Brian Schott
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
