- Haskell-Cafe - Haskell.org

Very crazy
by Andrew Coppin 26 Sep '07

26 Sep '07

While using Haskell, I often find myself writing convoluted constructions such as this: show_system = unlines . zipWith (\l ms -> "Eq" ++ show l ++ ": " ++ (concat $ intersperse " + " $ zipWith (\n x -> x ++ " x" ++ show n) [1..] (init ms)) ++ " = " ++ last ms ) [1..] . map (map (take 8 . show)) And people complain that *Perl* is bad? This function is quite obviously absurd. I mean, it works, but can *you* figure out what it does without running it? The question is, can anybody think of a better way to write this function? (And more generally, functions like it.)

12 27

Re: [Haskell-cafe] Very crazy
by Andrew Coppin 26 Sep '07

26 Sep '07

Chaddaï Fouché wrote: > 2007/9/25, Andrew Coppin <andrewcoppin(a)btinternet.com>: > >>> printf don't always perform IO : if you ask it for a String it will >>> happily turn into sprintf for you, if you use it in the IO Monad, it >>> will indeed perform IO, but there's nothing fundamentally IO bound in >>> printf logic. >>> >>> >> That's even *more* impossible... o_O >> > > Why ? > How can one function have more than one type signature?

3 2

Re: [Haskell-cafe] Haskell Cheat Sheet?
by Ben 26 Sep '07

26 Sep '07

+1 > Message: 9 > Date: Tue, 25 Sep 2007 13:04:56 -0700 > From: Evan Klitzke <evan(a)yelp.com> > Subject: [Haskell-cafe] Haskell Cheat Sheet? > To: haskell-cafe(a)haskell.org > Message-ID: <1190750696.5459.7.camel@thinkpad> > Content-Type: text/plain > > Has anybody made (or have a link to) a Haskell reference cheat sheet? > I'm thinking of a nice LaTeXed PDF in the 1-10 page range (e.g. > something like this http://www.tug.org/texshowcase/cheat.pdf) with the > basics of the language syntax, the type declarations for the common type > classes, the type signatures of the most commonly used functions in the > Prelude and other common modules, and so forth? The Haskell standard > library is very large for a newcomer (even just the Prelude!), and as a > learner of the language I find myself spending a lot of time looking up > Prelude functions and syntax details -- having all of this in a short > PDF document that I could have offline would be very useful. > > -- > Evan Klitzke <evan(a)yelp.com> > > > > ------------------------------ > > _______________________________________________ > Haskell-Cafe mailing list > Haskell-Cafe(a)haskell.org > http://www.haskell.org/mailman/listinfo/haskell-cafe > > > End of Haskell-Cafe Digest, Vol 49, Issue 118 > ********************************************* >

1 0

Re: [Haskell-cafe] Troubles understanding memoization in SOE
by Peter Verswyvelen 26 Sep '07

26 Sep '07

Thanks Paul & Paul for the answers. I'll certainly read the paper Paul Liu reported. I just deleted 100 lines of text which explained my problem in more detail, and while I was explaining it, I answered it myself. Typical. I thought the lambda function that memo1 returns would be called over and over again, and instead of reevaluating the stream from the beginning, it would just return the stream since it is in the cache, but actually it just gets called twice in recursive situations: the first time it evaluates y = f x, stores the thunk in the cache, and returns the thunk, the second time it finds the same thunk in the cache, and then computation of the rest of the stream continues without consulting the cache anymore right? From my clumsy explanation you can see that I'm still thinking too imperative, too eager. Haskell is more lazy than I am, which is an incredible achievement :-) It would really help if I could see the lazy computation; do you think this kind of memo code is traceable using HAT? I'll guess I'll have to check out arrows / yampa again. A year ago I did not understand a single thing in those papers, but I should try it again now I read the SOE book :-) Thanks a lot, Peter

2 2

Shouldnt this be lazy too?
by Vimal 25 Sep '07

25 Sep '07

Hi all, I was surprised to find out that the following piece of code: > length [1..] > 10 isnt lazily evaluated! I wouldnt expect this to be a bug, but in this case, shouldnt the computation end when the length function evaluation goes something like: > 10 + length [11..] ? -- -- Vimal

13 26

representing differencial equations in haskell
by Thomas Girod 25 Sep '07

25 Sep '07

Hi there. Let's say I have mathematical model composed of several differential equations, such as : di/dt = cos(i) dc/dt = alpha * (i(t) - c(t)) (sorry my maths are really bad, but I hope you get the point) I would like to approximate the evolution of such a system iteratively. How would you do that in haskell ? cheers, Thomas

4 3

Template Haskell newbie questions
by Hugo Pacheco 25 Sep '07

25 Sep '07

Hi all, I'm try to write some function in TH that I don't even know if it is possible. Consider the example from the tutorials sel 1 2 = [| \(x,_) -> x |] sel 2 2 = [| \(_,x) -> x |] If I want to write some function that will dynamically create a selection function according to its arguments and do something with it. fsel :: Int -> Int -> x fsel x y = $(sel x y) ... Imagine a stupid case for fsel :: Int -> Int -> Int fsel x y = $(sel x y) (x,y) The compiler complains that GHC stage restriction: `x' is used in a top-level splice, and must be imported, not defined locally In the first argument of `sel', namely `x' In the expression: $[splice](sel x y) (x, y) In the definition of `fsel': fsel x y = $[splice](sel x y) (x, y) what, indeed, makes sense. However, is there a possible way to generate Haskell functions from TH without having to instanciate its arguments? Thanks, hugo

3 4

Re: [Haskell] Re: [Haskell-cafe] PROPOSAL: Rename haskell@to haskell-announce@
by Andrzej Jaworski 25 Sep '07

25 Sep '07

> (...) I subscribe to haskell so as not to miss > anything important, and when something I'm interested in moves to > haskell-cafe, (...) Wrong assumption. Even if you know the author it makes more sense to wait and see how others respond to his newest input. This means that you should check up Cafe first to see if it is worth moving to Cafe;-) Another false believe is that best things appear first in Haskell@. On the contrary, long threads in Cafe indicate that participants know what to expect of each other and pick up a thread on the same principle some pick up new Stephen King novel. Before reviews on the amazon.com publishers had to result to tricks like labelling a collection of science fiction stories by various lesser known authors with Isaac Asimov name even when his only cotribution was a preface. As I said ealier - really big traffic will need intelligent screening and lebbeling. And to what pathology it leads when unattended see comp.soft-sys.matlab. Cheers, -Andrzej Jaworski

1 0

Re: [Haskell-cafe] Very crazy
by Miguel Mitrofanov 25 Sep '07

25 Sep '07

> How can one function have more than one type signature? It's called "polymorphism".

1 0

Re: [Haskell] simple function: stack overflow in hugs vs none in ghc
by Isaac Dupree 25 Sep '07

25 Sep '07

(redirecting to haskell-cafe) Tom Pledger wrote: > | > sqnc p ts = let ( r, ts' ) = p ts in case r of > | > Nothing -> ([],ts') > | > Just x -> let (r',ts'') = (sqnc p ts') in > (x:r', ts'' ) > : > > > I don't know how ghc is avoiding the stack overflow, but it looks like > it's caused by strict pattern matching. The first call to sqnc wants > reassurance that the second call is returning a pair (as opposed to _|_) > before deciding that it's OK to return a pair. > > Try putting a tilde in the recursive call, to make the pattern lazy. > > let ~(r',ts'') = ... "let" already is lazy -- putting a tilde there has no effect. In fact that tuple is produced from (x:r', ts'' ) even if (sqnc p ts') is _|_. (The "let" at the beginning of the function can be discovered to be strict by strictness analysis (which GHC does), because "r" is analyzed by the "case", and the pair needs to be deconstructed to find the value of "r".) Isaac

4 9