Roger, I think you are exactly right. Teaching x i. x is clearer and could easily be followed by instruction regarding the ~ (reflexive) adverb
thanks for pointing this out. cheers, bob On -Feb2-2010, at -Feb2-201011:17 AM, Roger Hui wrote: >> Thanks for the information that = (self classify) may not be the >> most efficient approach for the machine, but pedagogically there >> may be conceptual advantages to giving the learner a single step >> solution, compared to the i.~ which would require an explanation >> of adverbs as well. Once again I am finding the fastest way to >> learn is to teach (or attempt to teach ;) ) > > As far as the beginner is concerned, both = and i.~ are single step. > You don't necessarily have to explain the expression; instead, > you can explain the result: > > x=: 'mississippi' > = x > 1 0 0 0 0 0 0 0 0 0 0 > 0 1 0 0 1 0 0 1 0 0 1 > 0 0 1 1 0 1 1 0 0 0 0 > 0 0 0 0 0 0 0 0 1 1 0 > i.~ x > 0 1 2 2 1 2 2 1 8 8 1 > x i. x > 0 1 2 2 1 2 2 1 8 8 1 > > 0 1 2 3 4 5 6 7 8 9 10 > m i s s i s s i p p i > 0 1 2 2 1 2 2 1 8 8 1 > > I wasn't not necessarily talking about machine efficiency. > =x takes more room on the screen, and in your brain. > > > > ----- Original Message ----- > From: bob therriault <bobtherria...@mac.com> > Date: Tuesday, February 2, 2010 10:50 > Subject: Re: [Jprogramming] The Ambiguous Dictionary > To: Programming forum <programming@jsoftware.com> > >> Thanks Roger, >> >> I had been thinking about using + (plus) as well because it can >> generalize as a template to other arithmetic 0 rank dyadic >> verbs. Your examples provide a starting point for the concepts. >> I'll see what I can come up with. >> >> Thanks for the information that = (self classify) may not be the >> most efficient approach for the machine, but pedagogically there >> may be conceptual advantages to giving the learner a single step >> solution, compared to the i.~ which would require an explanation >> of adverbs as well. Once again I am finding the fastest way to >> learn is to teach (or attempt to teach ;) ) >> >> cheers, bob >> >> On -Feb2-2010, at -Feb2-20109:47 AM, Roger Hui wrote: >> >>>> Take a J primitive which is relatively simple but maybe hard >> for a >>>> beginner to intuit, like monadic "=" (self-classify). Not something >>>> with bells and whistles like ";:". Personally I'd like to see an >>>> animation of self-classify, because coming from APL I'm still not >>>> altogether comfy with it in my own mind. >>> >>> Use + (plus). It'd be so "obvious" that the viewer/reader >>> would think "How else would you have it?" >>> >>> 2 + 3 >>> 5 >>> >>> 2 + 3 4 5 >>> 5 6 7 >>> 20 30 10 + 7 >>> 27 37 17 >>> >>> 20 30 10 + 3 4 5 >>> 23 34 15 >>> 20 30 + 3 4 5 >>> |length error >>> | 20 30 +3 4 5 >>> >>> i. 3 4 >>> 0 1 2 3 >>> 4 5 6 7 >>> 8 9 10 11 >>> >>> 2 + i.3 4 >>> 2 3 4 5 >>> 6 7 8 9 >>> 10 11 12 13 >>> 20 40 60 + i.3 4 >>> 20 21 22 23 >>> 44 45 46 47 >>> 68 69 70 71 >>> (i.3 4) + 100 200 19 >>> 100 101 102 103 >>> 204 205 206 207 >>> 27 28 29 30 >>> 20 40 60 80 + i.3 4 >>> |length error >>> | 20 40 60 80 +i.3 4 >>> >>> (i.3 4) + (10 + i.3 4) >>> 10 12 14 16 >>> 18 20 22 24 >>> 26 28 30 32 >>> >>> p.s. I believe that the current monadic = is not one of >>> the better verbs in J. The representation is wasteful, >> being >>> approximately O(n^2), and if the representation is wasteful >>> then the implementation is necessarily wasteful. >>> For self-classification i.~ is much better. Moreover, >> the "key" adverb >>> (not available at the time the monad = was first defined) >>> has made a self-classification verb much less needed. >>> >>> >>> >>> ----- Original Message ----- >>> From: Ian Clark <earthspo...@googlemail.com> >>> Date: Tuesday, February 2, 2010 8:59 >>> Subject: Re: [Jprogramming] The Ambiguous Dictionary >>> To: Programming forum <programming@jsoftware.com> >>> >>>> URL embedding is something I must leave to others. I'd be starting >>>> from cold as regards Moinmoin. Maybe we should simply fall >> back on >>>> html? Remind me why we need anything fancier. >>>> >>>> Voice-over I have strong feelings against. Distracting, ineffective >>>> and culturally dependent. Traditional and reassuring maybe, >> but only >>>> in your own mother tongue, accent and mannerisms. Nothing >> going for >>>> it. >>>> >>>> And I say that as an ex-lecturer (and courseware developer). >>>> >>>> Use soothing music. It's only wallpaper. And then only in a YouTube >>>> showcase, not in a serious expository text. >>>> >>>> The same goes for moving text, or "titles" as I think movie makers >>>> call them. Only good for identifying the footage, like a >>>> clapper-board. The effort to read it can distract the viewer >>>> from the >>>> object you want him/her to attend to. There will always be >>>> accompanying text. I think the purpose of animation in technical >>>> exposition is to convey spatial concepts non-verbally. Therefore >>>> verbiage is nothing but a distraction. >>>> >>>> I like your use of coloured dots. It conveys succintly that the >>>> content of a given cell doesn't matter: just the interplay between >>>> cells. If you use numbers, or even letters, the viewer wastes >>>> brain-cycles processing these before grasping they are of no >>>> importance. If you need more modal dimensions: shape and >> size. As >>>> regards shape, chicks and tadpoles may be slightly better >> than squares >>>> and circles, but one must balance being boring against being >>>> distracting. I'd err on the "boring" side to start with, and >>>> spice-up >>>> judiciously. People assume animations have got to be Loony >>>> Toons. But >>>> someone prepared to consider programming in any language, let >>>> alone J, >>>> needs an attention-span longer than a 5 year old TV watcher. >>>> >>>> The viewer may have to study the animation for quite some >> time to >>>> absorb its principle. Eye candy soon gets irritating. See the >>>> "grasshopper escapement" animation, halfway down in >>>> http://en.wikipedia.org/wiki/John_Harrison -and work out how it >>>> manages to be friction-free. Jiminy Cricket would soon begin >> to grate. >>>> >>>> Further to that example, I think continuous animation loops >> are good. >>>> (Though it's nice to be able to stop 'em!) >>>> >>>> Take a J primitive which is relatively simple but maybe hard >> for a >>>> beginner to intuit, like monadic "=" (self-classify). Not something >>>> with bells and whistles like ";:". Personally I'd like to see an >>>> animation of self-classify, because coming from APL I'm still not >>>> altogether comfy with it in my own mind. >>>> >>>> And remember the better it conveys its message, the more >> trivial and >>>> "obvious" it looks and the less time the viewer spends >> looking >>>> at it. >>>> It's why writers of public notices are always too pompous and >> prolix:>> they don't want to appear dumb. It's really quite >> unrewarding >>>> work, in >>>> a way. >>>> >>>> Ian >>>> >>>> >>>> On Tue, Feb 2, 2010 at 5:56 AM, bob therriault >>>> <bobtherria...@mac.com> wrote: >>>>> Thanks for the review, Ian >>>>> >>>>> It definitely needs to be sped up and, although music could >> be >>>> added, it was going to be finished with a voice over and >> sound >>>> effects. I hadn't got around to final version, but when I >> read >>>> Oleg's comment about animation, I exported the 'work in >>>> progress' to YouTube. YouTube does provide a string for >>>> embedding, but I didn't think it would work within an email. >>>>> >>>>> I would be happy to work on animations as they were needed >> for >>>> specific primitives. My experience says that there is a >> certain >>>> level of fundamental understanding that is required before >>>> animations are really effective, before that they tend >> towards >>>> eye-candy (or worse misleading). If you have a primitive in >> mind >>>> let me know and we can go through some iterations to see how >>>> productive we are. >>>>> >>>>> Cheers, bob >>>>> >>>>> ps. you can put videos into J labs as well, by using the >>>> jbrowser file and calling specific URL's hiding it within the >>>> PREPARE section. >>>>> eg. >>>>> PREPARE >>>>> launch_jbrowser_ 'http://www.apple.com/ca/ipodtouch/what- >>>> is/pocket-computer.html' >>>>> PREPARE >>>>> It has been a while since I have done this and the result is >>>> that a whole browser window will open, but when refined this >> may >>>> have some possibilities. I'm not as sure how this will work >> with >>>> the new browser interface. bt >>>>> >>>>> On -Feb1-2010, at -Feb1-20106:43 PM, Ian Clark wrote: >>>>> >>>>>> I like it, Bob. >>>>>> >>>>>> The minimalism may arise from it being a first effort, but >> it shows >>>>>> it's neither necessary nor wise to over-egg the pudding. >>>>>> >>>>>> Minor crits: could do with being sped up. Needs music. >>>> YouTube has a >>>>>> library of free soundtracks: you can just attach one for now. >>>>>> >>>>>> YouTube movies can be easily embedded in your own html >> file. >>>> Stefano & >>>>>> I have done it. They generate you an <object> to >>>> copy/paste. See >>>>>> source of >> http://www.maxclark.me.uk/undeadtree/interspex.htm >>>> for an >>>>>> example. >>>>>> >>>>>> But how to do it with MoinMoin? >>>>>> >>>>>> So... are you offering to do a 15 second movie for each J >>>> primitive?>> >>>>>> Ian >>>>>> >>>>>> >>>>>> On Mon, Feb 1, 2010 at 10:08 PM, bob therriault >>>> <bobtherria...@mac.com> wrote: >>>>>>> Hi Oleg and Ian, >>>>>>> >>>>>>> http://www.youtube.com/watch?v=ODxv498p4ME >>>>>>> >>>>>>> This isn't about a specific function, but is an animation >> i >>>> put together to explain why it is useful to organize >> information >>>> into arrays. I developed it on keynote, but haven't yet put a >>>> soundtrack to it. It's not really high end, but it wasn't too >>>> hard to put together either. Let me know if you have more >>>> specific ideas for animations. >>>>>>> >>>>>>> Cheers, bob >>>>>>> >>>>>>> On -Feb1-2010, at -Feb1-201012:50 PM, Oleg Kobchenko wrote: >>>>>>> >>>>>>>>> From: Ian Clark <earthspo...@googlemail.com> >>>>>>>> >>>>>>>>> >>>>>>>>>> One winning instructional strategy, including Gilman >> and Rose, >>>>>>>>>> is telling a story for each piece of material. Such stories, >>>>>>>>>> among other things, may revolve around exo-paradigms. >>>>>>>>> >>>>>>>>> When I used to program in FORTH (sign of a mis-spent >>>> youth) one >>>>>>>>> well-beloved primer in the FORTH community was >> illustrated with >>>>>>>>> engaging but instructive cartoons. Thus the standard >> word >>>> SWOP was a >>>>>>>>> little two-headed dragon that did what you'd guess with >>>> objects on the >>>>>>>>> stack. Can anyone remember the book and remind me of its >> title?>>>>>> >>>>>>>> Starting FORTH, by Leo Brodie >>>>>>>> >>>>>>>> http://www.forth.com/starting-forth/sf2/sf2.html >>>>>>>> >>>>>>>> >>>>>>>> This is fun stuff all right. And a good read for a vacation. >>>>>>>> >>>>>>>> >>>>>>>> What I was thinking for APL/J for a long time was >>>>>>>> an interactive or animated illustrator of the operations, >>>>>>>> especially those that manipulate multidimensional and >>>>>>>> nested structures. >>>>>>>> >>>>>>>> >>>>>>>>> I know someone (http://www.leelamaria.com/) who could do >>>> us a wodge of >>>>>>>>> cartoons like that. Shall I try to get them interested? > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
