Bo Jacoby & Marshall Lochbaum wrote: > 1&o.&.:+: NB. mathematical equivalence
Ok, you guys win. :) -Dan PS: Does d. work with &. or &.: ? Please excuse typos; composed on a handheld device. -----Original Message----- From: Marshall Lochbaum <[email protected]> Sender: [email protected] Date: Fri, 19 Nov 2010 19:27:58 To: 'Programming forum'<[email protected]> Reply-To: Programming forum <[email protected]> Subject: Re: [Jprogramming] Accessing a verb as the composition ofits"primitives" How about 1&o.&.:+: ? (Okay, I cheated. This is mathematical, not semantic, equivalence. But I like it anyway.) Marshall -----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of Dan Bron Sent: Friday, November 19, 2010 6:27 PM To: J Programming Subject: Re: [Jprogramming] Accessing a verb as the composition ofits"primitives" Bingo. Or, to put it another way, (1&o. * 2&o.) d. 1 and sincos d. 1 mean the same thing, because (1&o. * 2&o.) and sincos mean the same thing. As proof, try sincos d.1 f. instead of sincos f. d. 1 . -Dan PS: of course if you're treating sincos vs (1&o. * 2.&o.) textually or grammatically, they look different; I'm just talking about syntactic equivalence here. BTW, 1 2 */ .(o./) ] is a fun semantic equivalent to sincos (which is different from syntactic equivalence! eg d. won't work). Can anyone make that cuter? Please excuse typos; composed on a handheld device. -----Original Message----- From: "Sherlock, Ric" <[email protected]> Sender: [email protected] Date: Sat, 20 Nov 2010 11:56:26 To: Programming forum<[email protected]> Reply-To: Programming forum <[email protected]> Subject: Re: [Jprogramming] Accessing a verb as the composition of its"primitives" There may be there is some confusion here. What Dan is saying is that the following two lines give the same result (which they do for me): (1&o. * 2&o.) d. 1 ] 0.4 0.69670671 sincos d. 1 ] 0.4 0.69670671 In other words the derivative of sincos is correctly taken when sincos d. 1 is applied to arguments, it just doesn't "show" the algebraic solution when entered without arguments. If you want that then use f. as Raul suggests. > From: Alex Gian > Sent: Saturday, 20 November 2010 11:33 > > Not on any of the J systems I've tried! (Including Linux, > Win32(Wine), or Windows Mobile / PPC) > > I thought it was just a weird peculiarity of J, obviously d. "should" > work on a user defined verb if it can. > > I tried with other verbs, like p. too, just in case o. was causing the > problem. Nope, once you define a verb in terms of its primitives d. > don't work no more. Just FYI > > On Fri, 2010-11-19 at 21:58 +0000, Dan Bron wrote: > > Raul is right, but just to be clear, when applied to arguments, > > sincos d. 1 > and sincos f. d. 1 will have identical results*. Just type into the > IJX without arguments, the latter looks different from the former for > exactly the same reason sincos f. looks different from sincos . > > > > That is, f. explicitly requests its argument be exploded into its > > components > (but again, the argument to f. applied to its own arguments will havbe > the same results, exploded or not - that's the point of naming stuff - > subordinating detail / hiding complexity). > > > > -Dan > > > > * I haven't tested this, but if it isn't true, that's an interpreter bug. > > > > > > > > > > Please excuse typos; composed on a handheld device. > > > ---------------------------------------------------------------------- > 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
