Mailman 3 October 2002 - Edu-sig

re: Types and true division (was Re: strange output
by Arthur Oct. 15, 2002

Oct. 15, 2002

Tim writes - >I value your presence in our little community. Glad to hear that. Because, porting PyGeo to Ocaml(?) was beginning to sound like quite the pain-in-the-ass. :) Art

1 0

re: Types and true division (was Re: strange output
by Arthur Oct. 15, 2002

Oct. 15, 2002

Tim writes - >Those folks were the first to bring up the issue forcefully, and Guido >recognized the correctness of their conclusion at once. Figuring out *why* >the conclusion was correct stretched out over years, and Guido didn't move >to change the language until after his brain caught up with his designer's >intuition. If you are saying that there was a totally serendipitous concurrence by which where Randy's and Bruce's forceful statement of the more than obvious issues related to 1/2 =0 and the newbie gets Guido to thinking, which thinking then lights on: "It has to do with substitutability of equal values with different types. When a==A and b==B, then a+b should be == A+B, at least within reasonable precision. " OK, I guess, if you say so. But we agree that one thing has absolutely nothing to do with the other. Don't we? And we understand how when total serendipity comes into play, us reasonably minded observers can get confused. No? So I'm forgivven. Yes? Art

2 1

the legality of naked expressions
by Walter A. Aprile Oct. 15, 2002

Oct. 15, 2002

Hello, I have noticed (or rather: my students have painfully noticed) that in Python it is legal but confusing to have certain expressions as statements. One that seems to bite people regularly in my class is the function-reference-that-is-not-a-function-call: def bemoan(): print "Woe is me!" a=1 if (a==1): bemoan else: print "I am happy!" When run this code, obviously and correctly, produces absolutely nothing --- but puzzlement. Demonstrating that, at the prompt, >>> bemoan <function bemoan at 0x008B3C40> helps somewhat. Is there perhaps some warning that can be turned on in the interpreter? In the same vein, certain students find it perplexing that if (a==1): 1+1 else: 2+2 is perfectly legal yet looks a bit pointless - even if I realize that __add__ could be performing untold magic behind the scenes. I hope that I am not rehashing an old point here, but I haven't found this in the PEPs. Thank you for your attention, Walter

3 2

EducationGuardian.co.uk Schools Before he had no future. Now he is a talented
by Jason Cunliffe Oct. 13, 2002

Oct. 13, 2002

http://education.guardian.co.uk/schools/story/0,5500,811079,00.html

1 0

Re: rationals
by Lloyd Hugh Allen Oct. 12, 2002

Oct. 12, 2002

> From: "Dennis E. Hamilton" <dennis.hamilton(a)acm.org> > To: <edu-sig(a)python.org> > Subject: RE: [Edu-sig] Re: rationals > Date: Fri, 11 Oct 2002 18:25:24 -0700 > > The biggest difference is that some floating-point calculations that will > converge (disappear at 0 or converge on a fixed value) will not terminate in > rational computation. > > Since this sort of thing is not a very good way to use floats either, it > should provoke more-appropriate numerical techniques. Right--I thought that the more accepted technique was to (if we're going for zero) pick an epsilon, and then exit when abs(arg)<epsilon (and if we converge to something else, perhaps to pick an epsilon, and quit when the last ten arguments were within epsilon of each other). re: int/int --> float Wha? But the definition of rational is the set of numbers that can be represented as int/int! I understand the speediness argument, but please, let us not go float unless we coerce there! I thought the whole point of the rational thread was for "1/3" to simplify to something meaningful. Would it work (just throwing this out, having no real idea of how rats are implemented elsewhere) to set each of the numerator and the denominator to be a dict of the prime factorization thereof (assuming we are working with <100 digit numerals)? Or we could say that "int/int --> float, int//int --> int, and int///int --> rational". But this feels like a dangerous road. It'll be neat to have a set of numbers, though, where division is cheaper than addition.

3 2

re: Types and true division (was Re: strange output
by Arthur Oct. 12, 2002

Oct. 12, 2002

>As I understand it, you wish there'd be some acknowledgement that >at some point in the evolutionary trajectory, there was a time >when newbie concerns seemed to weigh too heavily in the balance. Actually what I *wish* is that Michael Williams ruled the world. There I am in high asshole mode, contending with Williams of the "/" issue. And in the course of a few exchanges he made the following clear: 1) Whether he liked or approved of me was not to a point. 2) That he understood my position, in substance. 2) That he understood my position , in substance (double credit, because that seems to be of some importance to me) 3) That he disagreed with me. 4) That he was willing tacitly to acknowledge that he did not see my position as motivated by one-up-man-ship, by some interest to draw attention to myself, or to make someone else look stupid or be wrong. It was what I believed (and he disagreed) was the right answer, based on a decent understanding of the issues that were on his mind as an instructor of Python to physics students. 5) That he took something away from the exchange that was useful to him. I would like to at least nominate Mr. Williams to the post of Chancelor for Foreign Affairs. I will be the Foreign Affiars. Art

4 3

re: Typed arrays (not about div)
by Arthur Oct. 12, 2002

Oct. 12, 2002

Kirby - >Same issue, yet very clearly this has nothing whatever to do with the behavior of div, either way. Thanks for the change in subject line anyway. But yikes and duh. What is so difficult about the point that I am trying to make and you keep walking right past. A point that I am willingly to say that I am blowing out of proportion, that I am missing for the trees, that have misweighed and misjudged. Once you tell me you see that it is there. 1/2 = 0 brings us to attention that there is numeric typing and that it matters. And reminds us of it now and again. And we go from there to deeper matters if and when they arise. With less surprise. >To bring div into it only confuses the issue. 'a' is being coerced to an int: *I* am not bringing div into the behavior of typed arrays. Why the hell would I. What I said was: >Huh? under new "/" >ouch - I should have seen that coming under old "/" The "fun" I keep referring to is the fun of making the whirlygig go 'round. What I am pointing to here is a practical point of the most fundamental kind. It *doesn't* make me right about the "/" operator. But you keep telling me you have *no* idea what I am saying, and I can't understand why. Art

3 2

re: Types and true division (was Re: strange output
by Arthur Oct. 12, 2002

Oct. 12, 2002

>I conclude that you don't yourself understand the importance of typing, >or you wouldn't wish div to remain so ambiguous. Conclude what you like. After you understand my position. Python, I thought, was always interested in the practical outcome rather than theoretical purity. Which is why my mind does shut off I bit when you and Guido and Tim explain to me the *real* reason the div operator had to change. It's theoretical "right". I'm sure. Its symetrical, and all that. Does it help someone trying to get a handle on things? How? My position has *no* theoretical basis. Beyond the fact that - and please don't jump all over me if there is something technical amiss in this wording - it better exposes the reality of what we are dealing with (a pretty thin wrapper around standard C, no?). And that is only my attempt to say in higher falutin language, that my point is experiental - not theoretical. And not unlikely wrong. Which you are well entitled to tell me. But I will definitely take it better if I think you have first taken in the practical point I am making. Michael did. You haven't. Art

2 1

Re: [Edu-sig] Re: rationals
by Lloyd Hugh Allen Oct. 12, 2002

Oct. 12, 2002

> But it would look like this: > > >>> a = 3/2r > >>> b = 1/2r > >>> a * b > (3/2r) That won't do at all. (3/2)(1/2)==(3/4). I got terribly confused going from here to the decimal representation. (And I suppose I'll never be able to get implied division without a '*', while people are whining about ninny things). > >>> float(a*b) > 0.75 Here's a question from someone who hasn't used rationals outside of Mathematica, TI-89/92, some sort of weird Casio, and the like, mostly in an interactive basis, on an issue that people had later in the thread: besides the representation given by the "print" command, how would code be affected if the programmer expected a float and the engine had been upgraded to return a rational instead? What would break?

3 2

re: Types and true division (was Re: strange output)
by Arthur Oct. 12, 2002

Oct. 12, 2002

>But I have the ability to foresee where Bruce's and Michael's students are >most likely to get caught with *their* pants down, under the new operator. >Which are ways even more subtle, IMO, but just as likely to arise as with my >raised to a power issue. So, on the gotcha score - at best a wash, IMO. Kirby writes - >(You don't spell out the gotchas with the new behavior -- examples >might have been nice.) Gotchas with the new behavior are not directly gotchas with the new behavior. They are gotchas that arise when being blindsided by the fact that numeric typing matters. Something clear to anyone functioning under the old system. The one I raised to Michael was: >>> fromNumeric import * >>> a=array([1,2,3]) >>> a[1]=.5 >>> a array([1, 0, 3]) Huh? under new "/" ouch - I should have seen that coming under old "/" Michael acknowledged that that exact situation had already arisen and thrown folks for a loop in his class. Which is what was so bizarre to me about the discussion. That 1/2 = 0 would be an intial surprise to Bruce's students is so obviously true that it isn't worth mentioning. We could have all agreed as a starting point - relying on nothing but our common sense - that we got an issue on our hands. What might have been productive and interesting and actually a little scientific, is to test the full reprecussions of the*change* to 1/2 = .5 for Bruce's students. And then *begin* a discussion of pros and cons. And *I'm* the English major. Art

2 1