Thanks for the explanation. I would like to suggest either to revise the dictionary definition which is not as clear (at least it is not to some of us) or , even better, to avoid the single index constraint by allowing the selection of multiple and boxed agendas.
________________________________ From: Roger Hui <[email protected]> To: Programming forum <[email protected]> Sent: Thu, November 19, 2009 11:46:37 AM Subject: Re: [Jprogramming] Wrong agenda? [email protected] forms a train which then has nothing to do with agenda (unless the gerund m itself contains @.). [email protected] is a verb where [email protected] y computes vy to result in a single index which selects a single verb from the gerund m, which is then applied to y. Analogously for x [email protected] y . Historically, [email protected] was defined and implemented before [email protected] . ----- Original Message ----- From: Jose Mario Quintana <[email protected]> Date: Thursday, November 19, 2009 8:20 Subject: Re: [Jprogramming] Wrong agenda? To: Programming forum <[email protected]> > > different. In particular, [email protected] y is not the same as m...@.(v y) > y . > > I am still confused; the dictionary says: > > " > [email protected] is a verb defined by the gerundm with an agenda specified > byn ;that is, the verb represented by the train selected fromm > by the indicesn .Ifn is boxed, the train is parenthesized > accordingly. The [email protected] uses the result of the verbv to > perform the selection. > " > > How is [email protected] using the verb (0 1)"_ to (try to) perform the selection? > > > > > ________________________________ > From: Roger Hui <[email protected]> > To: Programming forum <[email protected]> > Sent: Thu, November 19, 2009 11:03:33 AM > Subject: Re: [Jprogramming] Wrong agenda? > > 3 : 'm...@.(v n) y' . Unfortunately for tacit fans, it's not tacit. > > More on the point: it is desirable but not mandatory for > the 8 variations of a conjunction to be related > m conj n y > x m conj n y > m conj v y > x m conj v y > u conj n y > x u conj n y > u conj v y > x u conj v y , > > just as it is the case that it is desirable but not mandatory > for the monad and dyad of a verb be related. > > I once told Ken that the 8 cases from each conjunction and > the 4 cases from each adverb provide such an embarrassment > of riches that EVEN HE will need some time to assign meanings > to all of them. > > > > ----- Original Message ----- > From: Thomas Costigliola <[email protected]> > Date: Thursday, November 19, 2009 7:49 > Subject: Re: [Jprogramming] Wrong agenda? > To: Programming forum <[email protected]> > > > On Thu, Nov 19, 2009 at 10:40 AM, Roger Hui > <[email protected]> > > wrote:> There is no bug. The forms [email protected] and [email protected] are fundamentally > > > different. In particular, [email protected] y is not the same as m...@.(v > y) > > y . > > > > Is it possible, using the form [email protected] y, that allows a train > from > > m be > > applied to y? > > > > > > > > > > > > > > ----- Original Message ----- > > > From: Dan Bron <[email protected]> > > > Date: Thursday, November 19, 2009 7:35 > > > Subject: Re: [Jprogramming] Wrong agenda? > > > To: Programming forum <[email protected]> > > > > > >> Pepe wrote: > > >> > What is wrong with the agenda in the last line? > > >> > (-`%)@.((0 1)"_) 1 2 3 4 5 > > >> > |rank error > > >> > > >> Nothing wrong with the expression; the interpreter has a > > >> bug. > > >> > > >> I guess few people use verbal agenda to construct trains, so > > >> this hasn't > > >> been caught before. Most often agenda is used to select a > > >> single verb > > >> from the agenda, so agenda expects the verb to return a scalar. > > >> > > >> I tried a workaround. I boxed the list so that the verb > > >> returns a scalar, > > >> which should have no effect on the train produced (the box would > > >> theoretically just put a pair of parens around the train): > > >> > > >> (- > > >> `%)@.((<0 1)"_) 1 2 3 4 5 > > >> |length error > > >> | (-`%)@.((<0 1)"_)1 2 3 > > >> 4 5 > > >> > > >> > > >> but no joy. > > >> > > >> -Dan > > >> > > >> > > >> ----- Original Message --------------- > > >> > > >> Subject: Re: [Jprogramming] Wrong agenda? > > >> From: Don Guinn <[email protected]> > > >> Date: Thu, 19 Nov 2009 07:47:05 -0700 > > >> To: Programming forum > > >> <[email protected]> > > >> Added parens to make the rank apply to the entire expression. > > >> > > >> ((-`%)@.((0 1))"_) 1 2 3 4 5 > > >> 0 1.5 2.66667 3.75 4.8 > > >> > > >> Is this what you wanted? > > >> > > >> On Thu, Nov 19, 2009 at 7:07 AM, Jose Mario Quintana < > > >> [email protected]> wrote: > > >> > > >> > What is wrong with the agenda in the last line? > > >> > > > >> > (- %) 1 2 3 4 5 > > >> > 0 1.5 2.66666667 3.75 4.8 > > >> > > > >> > (-`%)@.(0 1) 1 2 3 4 5 > > >> > 0 1.5 2.66666667 3.75 4.8 > > >> > > > >> > (-`%)@.((0 1)"_) 1 2 3 4 5 > > >> > |rank error > > >> > | (-`%)@.((0 1)"_)1 2 3 4 5 ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
