- Haskell-Cafe - Haskell.org

How to convert Num a => a to Int
by Sergey Perminov 15 Mar '07

15 Mar '07

As a beginner I want to implement such fuction: plusSecs :: Num a => a -> ClockTime -> ClockTime plusSecs n = addToClockTime ( secTimeDiff (n) ) where secTimeDiff n = TimeDiff { tdYear = 0, tdMonth = 0, tdDay = 0, tdHour = 0, tdMin = 0, tdSec = n, tdPicosec = 0 } The problem here that it's type signature is illegal because tdSec of System.Time.TimeDeff must be Int type only, but I wish to apply plusSecs to any Num or (at least) any Integral type number. The reason for this - to have less compilation errors (that are mostly about error types) and more abstractedness (generality) of this function. So, the question is: Is it possible somehow to convert any Num (or at least Integral) to Int. Please, tell me if I don't get smth in Haskell type system.

3 2

Re: small boys performance
by apfelmus＠quantentunnel.de 14 Mar '07

14 Mar '07

-- redirected from haskell.general Andrzej Jaworski wrote: > As I regretfully pointed out earlier in [Fwd: Re: Computer Language > Shootout] > large search and simulations are not for Haskell. This is equally true with > GHC 6.5 http://eric_rollins.home.mindspring.com/haskellAnt.html. Indeed, directly translated ML code is not for Haskell. Or did you really expect Array (Int,Int) Double, ubiquitous tail recursion and accumulating nodes at the end of a list (nextPath = path ++ [nextCity]) to give you simulation wonders? Btw, there are a lot of opportunities for abstraction, separation of concerns and reuse in the code. There are much to many to list them all, but here are few: - Most importantly, iterations are best separated into generation and selection: bestPath = last . iterate (evolve) There is really no need (and it hurts performance, you know) to have this parameter hell. - By grouping paths with their total scores like in data Path = Path { nodes :: [Node], score :: Int } you can define a very convenient best :: Path -> Path -> Path that chooses the one with the greater score. More general, you can define maxBy :: (a -> Double) -> a -> a -> a maxBy f x y if f x >= f y then x else y - Considering a minor stylistic point, incrs = [ (fromIntegral boost) | n <- pairs ] pairsWithInc = zip pairs incrs is best known as pairsWithInc = zip pairs $ repeat (fromIntegral boost) - Abstraction from the concrete graph implementation enables switching from Array (Int,Int) Double to a sparse representation if demanded, namely a nested IntMap (IntMap Double). - etc. > calculating a tensors with symbolic indices (without components) > so that one could have components calculated for specific cases on the end > of general calculation. http://www.nt.ntnu.no/users/haugwarb/Programming/haskell_automatic_differen… Regards, apfelmus

1 0

Parallelism on concurrent?
by Dusan Kolar 14 Mar '07

14 Mar '07

Hello all, I'm googling around haskell.org to get some deeper knowledge about Control.Parallel.Strategies than it is presented on http://www.haskell.org/ghc/docs/latest/html/libraries/base/Control-Parallel… BTW, could someone point me to some more deeper doc. about it? During googling I've discovered that since GHC 6.6, par, forkIO, and forkOS should make the stuff run in parallel if I have more than one CPU core. Is that right? I think not, because on my machine only par makes the things run in parallel and only during the computation (GC runs in a single thread). If it should work, how can I verify that my installation is correct? If it should not work, will it be working someday? Thanks for your patience, responses, and tips Dusan

6 8

c2hs and pthreadtypes.h
by Magnus Therning 14 Mar '07

14 Mar '07

I'm trying to use c2hs but get stuck when including sys/types.h (though the problem really resides in pthreadtypes.h): % ./Setup.hs build -v Preprocessing executables for kowasu-0.1... /usr/bin/c2hs -C -D__GLASGOW_HASKELL__=606 -C -Icsrc -o src/Kowasu/PTrace.hs src/Kowasu/PTrace.chs c2hs: Error in C header file. /usr/include/bits/pthreadtypes.h:69: (column 6) [FATAL] >>> Syntax error! The symbol `;' does not fit here. The part of pthreadtypes.h where it chokes seems to be 65 __extension__ union 66 { 67 int __spins; 68 __pthread_slist_t __list; 69 }; 70 } __data; 71 char __size[__SIZEOF_PTHREAD_MUTEX_T]; 72 long int __align; 73 } pthread_mutex_t; Any pointers on how to solve this? (Besides manually resolving all defined types down to the C builtins.) /M -- Magnus Therning (OpenPGP: 0xAB4DFBA4) magnus(a)therning.org Jabber: magnus.therning(a)gmail.com http://therning.org/magnus

4 4

Defining types (newbie)
by John Fouhy 14 Mar '07

14 Mar '07

Hi all, I studied Haskell a few years ago at uni, but I haven't used it since, so I'm a bit rusty. But I've been inspired by this paper: http://citeseer.ist.psu.edu/585421.html which talks about building a combinator language for financial contracts, and which uses Haskell as the implementation language. ...anyway. Say I have a data type: data Thing = Thing { field_one :: String, field_two :: String, field_three :: Integer } I'd like to build a datatype that represents comparison rules on Things. A comparison rule would be something like "thing1.field_two == thing2.field_two" or "thing1.field_three > thing2.field_three". If I was writing python, I might do this as: class ThingCompare: def __init__(self, op, field, thing1, thing2): self.op = op self.field = field I could evaluate instances of ThingCompare as: def evalThingCompare(tc, thing1, thing2): return tc.op(getattr(thing1, tc.field), getattr(thing2, tc.field)) In Haskell, I envisage writing something like: data ThingCompare = TC Op Field So that if I had: eq x y = x == y then I could create a ThingCompare by saying: tc = eq field_two And I could evaluate my ThingCompares by defining: evaltc :: ThingCompare -> Thing -> Thing -> Bool evaltc (TC o f) t1 t2 = o (f t1) (f t2) But I can't figure out what types "Op" and "Field" should be. Any suggestions? Am I on the wrong track? (have I been corrupted by python's dynamic typing? :-) ) -- John.

4 7

Re: Maybe and partial functions
by oleg＠pobox.com 14 Mar '07

14 Mar '07

Neil Mitchell wrote: > newtype N1 = N1 Int > (put that in a module and don't export N1) > > define the constant 2, define the increment operator, change div and mod. That is precisely what I would have done. > Now we've mainly got a proof in the type checker, but we still don't > actually have a proof for our core N1 definition - but I think that is > probably ok, provided it is a common core which can be reused and is > minimal. Any proof of totality of the original digits-of-e code must first prove or postulate that incrementing a positive number gives the positive number. So, one must prove/verify/assume the basic properties one way or another. In the lightweight static capabilities approach, these basic lemmas/properties/axioms are collected in modules, which can be verified once and for all and then used in various parts of the program. Accommodating the security kernel in those parts of the program does not seem to be a burden. In return, we get assurances of the simple kernel in those parts of the program without the need to run the verification algorithm all over again (on the possibly complex pieces of code). The idea that data type abstraction can enforce complex invariants was advocated back in 1973 by James H. Morris (the father of the guy who released the `Morris worm'), in the CACM paper "Protection in Programming Languages". http://www.erights.org/history/morris73.pdf

1 0

Re: [Haskell] Haskell Weekly News: March 12, 2007
by Conrad Parker 14 Mar '07

14 Mar '07

On 13/03/07, Wolfgang Jeltsch <g9ks157k(a)acme.softbase.org> wrote: > Am Montag, 12. März 2007 03:52 schrieb Donald Bruce Stewart: > > * [41]Why Publish CS Papers Without Code? > > 41. http://billmill.org/why_no_code > > Interesting! > > This leads me to the question how copyright of code fragments included in > conference papers is handled. Say, I submit a paper containing some code > fragments to the ICFP. I have to assign the copyright of the paper to the > ACM. Does this mean that the ACM will hold the copyright of the code > fragments? Does this mean that I'm not allowed to publish the source code > containing these code fragments as open source software somewhere? (I am not a lawyer, etc) If your paper simply cites, and quotes from, code that you have already published elsewhere (eg. on your own web site), then I think that would simply count as a citation. If on the other hand your paper contains your code verbatim and doesn't cite it (ie. the code is previously unpublished) then perhaps you are assigning the copyright of it. So it may be a good idea to always make available open source projects before publishing about them in academic papers, and to always include a citation to the open source project web site or tech report. Besides, <tshirt>If it's not open source, it's not computer science</tshirt>. Science demands repeatable results, computer science demands literate programming. The solution is not to shy away from including code, or else the IP lawyers have won, science is banned and we get plunged into another Dark Age. cheers, Conrad.

4 3

Maybe and partial functions
by Dougal Stanton 14 Mar '07

14 Mar '07

The Maybe construction is very useful for explicitly handling circumstances where the function cannot produce a sensible answer. But how far should this notion be taken? When you're writing a function which you realise may not produce what you want, in what circumstances would you use a Maybe, and when would you just throw an error? I'm wondering both in terms of good engineering practise and also for correctness. Should elementary partial functions like > 5 `div` 0 or > head [] return Nothing? I guess it's a bit of a silly suggestion, but it helps to highlight why we use Maybe in the first place. So --- where's the cutoff point in your code? Cheers, D. -- Dougal Stanton

5 4

Mixed-type arithmetic, or type coercion
by Dave Hinton 13 Mar '07

13 Mar '07

Arithmetic operators in haskell appear to require their operands to have the same type. How can I do arithmetic between operands of different types? Alternatively, how do I coerce a value from being an Int to being a Double? I am new to haskell. I have read a few tutorials but they do not mention this issue. I have read the haddock documentation for Prelude and Numeric, but cannot find anything that helps here. (I saw a couple of numeric conversion functions, but they won't turn an Int into a Double.) Toy example: $ cat eg.hs calc :: Int -> Double -> Double calc count weight = count * weight $ ghc eg.hs eg.hs:2: Couldn't match `Int' against `Double' Expected type: Int Inferred type: Double In the second argument of `(*)', namely `weight' In the definition of `calc': calc count weight = count * weight

4 3

Maybe and partial functions
by oleg＠pobox.com 13 Mar '07

13 Mar '07

Neil Mitchell wrote: > I suggest you try rewriting this program to be complete: > > http://darcs.haskell.org/nofib/imaginary/digits-of-e2/Main.lhs > > (if you do, please post the result to the list) As Gen Zhang noted, the problem seems to be quite straightforward: just express in types the fact that we deal with infinite streams and so Nil just cannot happen. One way of doing this as follows. Showing totality (that is, termination of recursion in carry propagate) is harder. But re-writing the program so to avoid head/tail on the empty list is easy: module E where import Prelude hiding (head, tail, map, iterate, take) data I a = C a (I a) deriving Show -- these are total functions: they give WHNF in one step provided non-bottom -- arguments head (C a _) = a tail (C _ l) = l map f (C a l) = C (f a) (map f l) rept a = C a (rept a) take n (C a l) | n > 0 = a:take (n-1) l take _ _ = [] iterate f z = C z (iterate f (f z)) -- There are no pattern-matching failures here. -- The totality is harder to see: all digits are roughly of the same range, -- but each recursive call increments base. Eventually, base becomes bigger -- than d+9 and so the first alternative will be selected, which is in the -- WHNF and so recursion terminates. carryPropagate base (C d ds) | carryguess == (d+9) `div` base = C carryguess (C (remainder+nextcarry) fraction) | otherwise = (C (dCorrected `div` base) (C (dCorrected `mod` base) fraction)) where carryguess = d `div` base remainder = d `mod` base C nextcarry fraction = carryPropagate (base+1) ds dCorrected = d + nextcarry e = map (show.head) $ iterate (carryPropagate 2 . map (10*) . tail) $ C 2 (rept 1) en n = "2." ++ concat (take n (tail e))

4 5