Re: [Haskell-cafe] Template Haskell tutorials?
On Wed, Apr 13, 2011 at 10:43 PM, Kenneth Hoste kenneth.ho...@gmail.com wrote: Hi, The links to the supposedly brilliant Template Haskell tutorials by Bulat are broken. http://www.haskell.org/bz/thdoc.htm http://www.haskell.org/bz/th3.htm Does anyone know if these tutorials moved to somewhere else? greetings, Kenneth If all else fails, it's still accessable via the wayback machine. http://replay.waybackmachine.org/20090418034412/http://www.haskell.org/bz/thdoc.htm and http://replay.waybackmachine.org/20090421010040/http://www.haskell.org/bz/th3.htm -Zoe (hopefully I've figured out this mailing list thing right) ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Haskell on-line judge for Programming Challenges / Contests ?
Hello, I am looking for a site providing on-line automatic judge for Programming Challenges that can be coded in Haskell. For example this site: http://www.programming-challenges.com provides lots of interesting programming provlems: http://www.programming-challenges.com/pg.php?page=index Yet, unfortunately only C, C++ and Java code can be submitted, judge on that site can't compile and run Haskell code. Is there any similar online judge for Haskell? Thanks, Dmitri ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Haskell on-line judge for Programming Challenges / Contests ?
Hi On 14/04/11 10:29, Dmitri O.Kondratiev wrote: I am looking for a site providing on-line automatic judge for Programming Challenges that can be coded in Haskell. For example this site: http://www.programming-challenges.com provides lots of interesting programming provlems: http://www.programming-challenges.com/pg.php?page=index Yet, unfortunately only C, C++ and Java code can be submitted, judge on that site can't compile and run Haskell code. Is there any similar online judge for Haskell? You can try http://www.spoj.pl/ Cheers, Ángel -- http://www.iac.es/galeria/angelv/ High Performance Computing Support PostDoc Instituto de Astrofísica de Canarias - ADVERTENCIA: Sobre la privacidad y cumplimiento de la Ley de Protección de Datos, acceda a http://www.iac.es/disclaimer.php WARNING: For more information on privacy and fulfilment of the Law concerning the Protection of Data, consult http://www.iac.es/disclaimer.php?lang=en ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] IO and Cont as monads
On 13 April 2011 21:26, Tim Chevalier catamorph...@gmail.com wrote: IO doesn't obey the monad laws, due to the presence of seq in Haskell. Sad but true... See also a previous discussion about IO and the Monad laws: http://www.haskell.org/pipermail/haskell-cafe/2010-March/074001.html ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Haskell on-line judge for Programming Challenges / Contests ?
On Thu, 14 Apr 2011 13:29:48 +0400 Dmitri O.Kondratiev doko...@gmail.com wrote: Hello, I am looking for a site providing on-line automatic judge for Programming Challenges that can be coded in Haskell. For example this site: http://www.programming-challenges.com provides lots of interesting programming provlems: http://www.programming-challenges.com/pg.php?page=index Yet, unfortunately only C, C++ and Java code can be submitted, judge on that site can't compile and run Haskell code. Is there any similar online judge for Haskell? Thanks, Dmitri Althpugh not Haskell-specific, the Mooshak system for managing programming contests on the Web allows you to configure most programming languages: http://mooshak.dcc.fc.up.pt/ It's been developed our CS department and used internally and externally for programming contests and courses. Pedro ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Programming Chalenges: The 3n+1 problem
3n+1 is the first, warm-up problem at Programming Chalenges site:
http://www.programming-challenges.com/pg.php?page=downloadproblemprobid=110101format=html
(This problem illustrates Collatz conjecture:
http://en.wikipedia.org/wiki/3n_%2B_1#Program_to_calculate_Collatz_sequences
)
As long as the judge on this site takes only C and Java solutions, I
submitted in Java some add-hock code (see at the end of this message) where
I used recursion and a cache of computed cycles. Judge accepted my code and
measured 0.292 sec with best overall submissions of 0.008 sec to solve the
problem.
*** Question: I wonder how to implement cache for this problem in Haskell?
At the moment, I am not so much interested in the speed of the code, as in
nice implementation.
To illustrate my question I add the problem description and my Java solution
at the end of this message.
Thanks!
*** Problem
Consider the following algorithm to generate a sequence of numbers. Start
with an integer *n*. If *n* is even, divide by 2. If *n* is odd, multiply by
3 and add 1. Repeat this process with the new value of *n*, terminating
when *n* = 1. For example, the following sequence of numbers will be
generated for *n* = 22:
22 11 34 17 52 26 13 40 20 10 5 16 8 4 2 1
It is *conjectured* (but not yet proven) that this algorithm will terminate
at *n* = 1 for every integer *n*. Still, the conjecture holds for all
integers up to at least 1, 000, 000.
For an input *n*, the *cycle-length* of *n* is the number of numbers
generated up to and *including* the 1. In the example above, the cycle
length of 22 is 16. Given any two numbers *i* and *j*, you are to determine
the maximum cycle length over all numbers between *i* and *j*, *including* both
endpoints.
InputThe input will consist of a series of pairs of integers *i* and *j*,
one pair of integers per line. All integers will be less than 1,000,000 and
greater than 0.
OutputFor each pair of input integers *i* and *j*, output *i*, *j* in the
same order in which they appeared in the input and then the maximum cycle
length for integers between and including *i* and *j*. These three numbers
should be separated by one space, with all three numbers on one line and
with one line of output for each line of input.
Sample Input
1 10
100 200
201 210
900 1000
Sample Output
1 10 20
100 200 125
201 210 89
900 1000 174
*** my Java solution
import java.io.BufferedReader;
import java.io.InputStreamReader;
public class Main {
final static BufferedReader reader_ = new BufferedReader(new
InputStreamReader(System.in));
/**
* @param args
*/
public static void main(String[] args) {
new Problem().run();
}
static String[] ReadLn() {
String[] tokens = null;
try {
String line = reader_.readLine();
String REGEX_WHITESPACE = \\s+;
String cleanLine =
line.trim().replaceAll(REGEX_WHITESPACE, );
tokens = cleanLine.split(REGEX_WHITESPACE);
} catch (Exception e) {}
return tokens;
}
}
class Problem implements Runnable {
long CACHE_SIZE = 65536;
private final long[] cache_ = new long[(int) CACHE_SIZE];
/**
* Compute cycle length for a single number
*
* @param n number for which we find cycle length
* @return cycle length
*/
long cycleLen(long n) {
long len = 1;
if (n != 1) {
len = getFromCache(n);
if (len == 0) { //not yet in cache
// Recursively compute and store all
intermediate values of cycle length
if ((n 1) == 0) {
len = 1 + cycleLen(n 1);
} else {
len = 1 + cycleLen(n * 3 + 1);
}
putInCache(n, len);
}
}
return len;
}
void putInCache(long n, long len) {
if(n CACHE_SIZE) {
cache_[(int)n] = len;
}
}
long getFromCache(long n) {
long result = 0;
if(n CACHE_SIZE) {
result = cache_[(int)n];
}
return result;
}
/**
* Find max cycle on interval
*
* @param from interval start
* @param to interval end
* @return max cycle
*/
Long maxCycle(Long from, Long to) {
Long result = 0L;
Long cycle = 0L;
// Get all values of cycle length on the interval and put these
values into a sorted set
Re: [Haskell-cafe] Haskell on-line judge for Programming Challenges / Contests ?
http://codechef.com accepts Haskell entries. pgpa3li6zz0gF.pgp Description: PGP signature ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Programming Chalenges: The 3n+1 problem
This is very similar to the Problem 14 on Project Euler [1]. Should you really want a nice solution in Haskell, and noting that this is a big spoiler, there are some on the wiki [2]. But I highly recommend you to try coding your own solutions before looking at the site =). Cheers! [1] http://projecteuler.net/index.php?section=problemsid=14 [2] http://www.haskell.org/haskellwiki/Euler_problems/11_to_20#Problem_14 -- Felipe. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Haskell on-line judge for Programming Challenges / Contests ?
Angel de Vicente wrote: On 14/04/11 10:29, Dmitri O.Kondratiev wrote: * I am looking for a site providing on-line automatic judge for ** Programming Challenges that can be coded in Haskell. ** For example this site: ** http://www.programming-challenges.com ** provides lots of interesting programming provlems: ** http://www.programming-challenges.com/pg.php?page=index * * Yet, unfortunately only C, C++ and Java code can be submitted, judge on ** that site can't compile and run Haskell code. ** Is there any similar online judge for Haskell? * You can try http://www.spoj.pl/ Cheers, Ángel Angel, thanks! This site looks good. Do you know if spoj.pl has a search utility for problems? It would be great if they have 3n+1 problem I wrote about in a separate thread here. -- All the best, Dmitri O. Kondratiev This is what keeps me going: discovery doko...@gmail.com http://sites.google.com/site/dokondr/welcome ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Programming Chalenges: The 3n+1 problem
Hi Dmitri, As a reference you might want to take a look at the Project Euler problem #14 and the corresponding dubious entry in the HaskellWiki: http://www.haskell.org/haskellwiki/Euler_problems/11_to_20#Problem_14 . *** Question: I wonder how to implement cache for this problem in Haskell? At the moment, I am not so much interested in the speed of the code, as in nice implementation. a while ago I wondered about the same question: memoization and more specifically tabulation (DP) in Haskell. After some time I came up with code which does it IMO in a fairly nice way using an array of lazy computations which gives you top-down DP for free at the expense of performance. Another way is to use lazily built tries ala MemoTrie: http://www.haskell.org/haskellwiki/MemoTrie . Specific implementation: import Data.Array import Data.MemoTrie I write the collatz function using open recursion in order to separate the recursive function from the memo scheme. type Gen a = a - a type Fix a = Gen a - a collatzGen :: Gen (Integer - Integer) collatzGen c 1 = 1 collatzGen c n | even n= 1 + c (div n 2) | otherwise = 1 + c (3*n + 1) We can use plain old Data.Function.fix to get the usual stupid recursive function or use our custom fixpoint combinators. |fixtab| for example uses an array to cache the result in the given range. When the argument is outside of the range the normal recursive scheme is applied. fixtab :: Ix i = (i, i) - Fix (i - r) fixtab b f = f' where a= listArray b [ f f' i | i - range b ] f' i = if inRange b i then a!i else f f' i Another option is to use the MemoTrie package where we can write a fixpoint combinator like this: fixmemo :: HasTrie i = Fix (i - r) fixmemo f = let m = memo (f m); in m And of course the final collatz function: collatz :: Integer - Integer collatz = fixtab (1,100) collatzGen Cheers, Steven ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Programming Chalenges: The 3n+1 problem
Thanks, Felipe! Sure I will try to code solution myself. Doing it all by yourself is the main fun of everything, isn't it? I was just asking about the *idea* of implementing cache in Haskell, in general, not just in this case only. On Thu, Apr 14, 2011 at 3:11 PM, Felipe Almeida Lessa felipe.le...@gmail.com wrote: This is very similar to the Problem 14 on Project Euler [1]. Should you really want a nice solution in Haskell, and noting that this is a big spoiler, there are some on the wiki [2]. But I highly recommend you to try coding your own solutions before looking at the site =). Cheers! [1] http://projecteuler.net/index.php?section=problemsid=14 [2] http://www.haskell.org/haskellwiki/Euler_problems/11_to_20#Problem_14 -- Felipe. -- All the best, Dmitri O. Kondratiev This is what keeps me going: discovery doko...@gmail.com http://sites.google.com/site/dokondr/welcome ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Programming Chalenges: The 3n+1 problem
So if we were to emulate your Java solution, we'd do
import Data.Array
cacheSize :: Int
cacheSize = 65536
table :: Array Int Integer
table = listArray (1,cacheSize) (1 : map go [2..cacheSize]) where
go n
| even n = 1 + lookup (n `div` 2)
| otherwise = 1 + lookup (3 * n + 1)
lookup :: Integer - Integer
lookup n
| n cacheSize = table ! (fromInteger n)
| even n = 1 + lookup (n `div` 2)
| otherwise = 1 + lookup (3 * n + 1)
The rest of the code is just some simple i/o.
The table is filled up lazily as you request values from it.
On Thu, Apr 14, 2011 at 3:29 AM, Dmitri O.Kondratiev doko...@gmail.comwrote:
3n+1 is the first, warm-up problem at Programming Chalenges site:
http://www.programming-challenges.com/pg.php?page=downloadproblemprobid=110101format=html
(This problem illustrates Collatz conjecture:
http://en.wikipedia.org/wiki/3n_%2B_1#Program_to_calculate_Collatz_sequences
)
As long as the judge on this site takes only C and Java solutions, I
submitted in Java some add-hock code (see at the end of this message) where
I used recursion and a cache of computed cycles. Judge accepted my code and
measured 0.292 sec with best overall submissions of 0.008 sec to solve the
problem.
*** Question: I wonder how to implement cache for this problem in Haskell?
At the moment, I am not so much interested in the speed of the code, as in
nice implementation.
To illustrate my question I add the problem description and my Java
solution at the end of this message.
Thanks!
*** Problem
Consider the following algorithm to generate a sequence of numbers. Start
with an integer *n*. If *n* is even, divide by 2. If *n* is odd, multiply
by 3 and add 1. Repeat this process with the new value of *n*, terminating
when *n* = 1. For example, the following sequence of numbers will be
generated for *n* = 22:
22 11 34 17 52 26 13 40 20 10 5 16 8 4 2 1
It is *conjectured* (but not yet proven) that this algorithm will
terminate at *n* = 1 for every integer *n*. Still, the conjecture holds
for all integers up to at least 1, 000, 000.
For an input *n*, the *cycle-length* of *n* is the number of numbers
generated up to and *including* the 1. In the example above, the cycle
length of 22 is 16. Given any two numbers *i* and *j*, you are to
determine the maximum cycle length over all numbers between *i* and *j*, *
including* both endpoints.
Input The input will consist of a series of pairs of integers *i* and *j*,
one pair of integers per line. All integers will be less than 1,000,000 and
greater than 0.
OutputFor each pair of input integers *i* and *j*, output *i*, *j* in the
same order in which they appeared in the input and then the maximum cycle
length for integers between and including *i* and *j*. These three numbers
should be separated by one space, with all three numbers on one line and
with one line of output for each line of input.
Sample Input
1 10
100 200
201 210
900 1000
Sample Output
1 10 20
100 200 125
201 210 89
900 1000 174
*** my Java solution
import java.io.BufferedReader;
import java.io.InputStreamReader;
public class Main {
final static BufferedReader reader_ = new BufferedReader(new
InputStreamReader(System.in));
/**
* @param args
*/
public static void main(String[] args) {
new Problem().run();
}
static String[] ReadLn() {
String[] tokens = null;
try {
String line = reader_.readLine();
String REGEX_WHITESPACE = \\s+;
String cleanLine =
line.trim().replaceAll(REGEX_WHITESPACE, );
tokens = cleanLine.split(REGEX_WHITESPACE);
} catch (Exception e) {}
return tokens;
}
}
class Problem implements Runnable {
long CACHE_SIZE = 65536;
private final long[] cache_ = new long[(int) CACHE_SIZE];
/**
* Compute cycle length for a single number
*
* @param n number for which we find cycle length
* @return cycle length
*/
long cycleLen(long n) {
long len = 1;
if (n != 1) {
len = getFromCache(n);
if (len == 0) { //not yet in cache
// Recursively compute and store all
intermediate values of cycle length
if ((n 1) == 0) {
len = 1 + cycleLen(n 1);
} else {
len = 1 + cycleLen(n * 3 + 1);
}
putInCache(n, len);
}
}
return len;
}
void putInCache(long n, long len) {
if(n CACHE_SIZE) {
Re: [Haskell-cafe] Haskell on-line judge for Programming Challenges / Contests ?
Hi, On 14/04/11 12:16, Dmitri O.Kondratiev wrote: Angel, thanks! This site looks good. Do you know if spoj.pl http://spoj.pl has a search utility for problems? It would be great if they have 3n+1 problem I wrote about in a separate thread here. yes they have a search utility. Searching for 3n+1, one of the problems that comes up is the problem with ID: 4073 (http://www.spoj.pl/problems/PROBTNPO/), but I don't know if that is the problem you were talking about. Ángel -- http://www.iac.es/galeria/angelv/ High Performance Computing Support PostDoc Instituto de Astrofísica de Canarias - ADVERTENCIA: Sobre la privacidad y cumplimiento de la Ley de Protección de Datos, acceda a http://www.iac.es/disclaimer.php WARNING: For more information on privacy and fulfilment of the Law concerning the Protection of Data, consult http://www.iac.es/disclaimer.php?lang=en ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Programming Chalenges: The 3n+1 problem
Thanks, everybody! Your feedback is a great food for my mind (as Lewis Carroll once wrote :) When asking how to implement cache in Haskell I was hopping that there exists some solution without using Data.Array, more functional approach, if I may say so ... I must be wrong, though (need more time to fully comprehend solution that Steven described in this thread ). On Thu, Apr 14, 2011 at 3:22 PM, Ryan Ingram ryani.s...@gmail.com wrote: So if we were to emulate your Java solution, we'd do import Data.Array cacheSize :: Int cacheSize = 65536 table :: Array Int Integer table = listArray (1,cacheSize) (1 : map go [2..cacheSize]) where go n | even n = 1 + lookup (n `div` 2) | otherwise = 1 + lookup (3 * n + 1) lookup :: Integer - Integer lookup n | n cacheSize = table ! (fromInteger n) | even n = 1 + lookup (n `div` 2) | otherwise = 1 + lookup (3 * n + 1) The rest of the code is just some simple i/o. The table is filled up lazily as you request values from it. -- All the best, Dmitri O. Kondratiev This is what keeps me going: discovery doko...@gmail.com http://sites.google.com/site/dokondr/welcome ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Programming Chalenges: The 3n+1 problem
Am 14.04.2011 12:29, schrieb Dmitri O.Kondratiev: 3n+1 is the first, warm-up problem at Programming Chalenges site: http://www.programming-challenges.com/pg.php?page=downloadproblemprobid=110101format=html http://www.programming-challenges.com/pg.php?page=downloadproblemprobid=110101format=html (This problem illustrates Collatz conjecture: http://en.wikipedia.org/wiki/3n_%2B_1#Program_to_calculate_Collatz_sequences) As long as the judge on this site takes only C and Java solutions, I submitted in Java some add-hock code (see at the end of this message) where I used recursion and a cache of computed cycles. Judge accepted my code and measured 0.292 sec with best overall submissions of 0.008 sec to solve the problem. *** Question: I wonder how to implement cache for this problem in Haskell? At the moment, I am not so much interested in the speed of the code, as in nice implementation. I'ld use something like: import qualified Data.Map as Map addToMap :: Integer - Map.Map Integer Integer - Map.Map Integer Integer addToMap n m = case Map.lookup n m of Nothing - let l = if even n then div n 2 else 3 * n + 1 p = addToMap l m Just s = Map.lookup l p in Map.insert n (s + 1) p Just _ - m addRangeToMap :: Integer - Integer - Map.Map Integer Integer - Map.Map Integer Integer addRangeToMap i j m = if j i then m else addRangeToMap i (j - 1) $ addToMap j m getMaxLength :: Integer - Integer - Map.Map Integer Integer - Integer getMaxLength i j = Map.foldWithKey (\ k l - if i k || k j then id else max l) 0 -- putting it all togeter getRangeMax :: Integer - Integer - Integer getRangeMax i j = getMaxLength i j $ addRangeToMap i j $ Map.singleton 1 1 ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Type Classes in Haskell - how can I make GHC make a choice of types, when the type chosen doesn't matter?
This is a question about the use of type classes in Haskell.
I get an error (below) when trying to compile the code (below and at
https://github.com/chrisdew/haskell-sandbox/blob/master/not_working_but_clean.hs
).
As someone just learning Haskell, I have tried following GHC's advice,
but I think the cause is different.
I believe the problem is that either of the types 'IO String' or plain
'String' could be the type of 'lhello - lbracket', but GHC doesn't
know which.
The problem is that it doesn't matter, either type would work fine.
I have posted a working version of the code at
https://github.com/chrisdew/haskell-sandbox/blob/master/working_but_ugly.hs
. This replaces one of the - operators with a new (non type class)
operator '-' which forces 'lhello - lbracket' to be of type 'IO
String'.
* Is my analysis correct? Or is there something else going on here?
* Is there any way of informing GHC what the type of 'lhello -
lbracket' doen't matter and that is should just chose either of the
two possibilities. Or perhaps theres a LANGUAGE option which will let
me specify that 'lastest declared matching instance of the class wins'
if anything is undecided.
Thanks,
Chris.
Error:
chris@chris-linux-desktop:~/nonworkspace/haskell-sandbox$ ghc
not_working_but_clean.hs
not_working_but_clean.hs:40:16:
No instance for (Stream (IO String) (IO String) (IO String) d)
arising from a use of `-' at not_working_but_clean.hs:40:16-34
Possible fix:
add an instance declaration for
(Stream (IO String) (IO String) (IO String) d)
In the first argument of `(-)', namely `lhello - lbracket'
In the second argument of `($)', namely
`lhello - lbracket - putStrLn'
In a stmt of a 'do' expression:
forkIO $ lhello - lbracket - putStrLn
not_working_but_clean.hs:40:16:
No instance for (Stream d String (IO ()) (IO ()))
arising from a use of `-' at not_working_but_clean.hs:40:16-47
Possible fix:
add an instance declaration for (Stream d String (IO ()) (IO ()))
In the second argument of `($)', namely
`lhello - lbracket - putStrLn'
In a stmt of a 'do' expression:
forkIO $ lhello - lbracket - putStrLn
In the expression:
do { forkIO $ (bracket $ hello) - putStrLn;
forkIO $ lhello - lbracket - putStrLn;
forkIO $ bracket hello - putStrLn;
forkIO $ lbracket lhello - putStrLn;
}
not_working_but_clean.hs:
{-# LANGUAGE MultiParamTypeClasses, FlexibleInstances,
TypeSynonymInstances, OverlappingInstances #-}
{-# OPTIONS_GHC #-}
module Main (
main
)
where
import Control.Concurrent (forkIO, MVar, newEmptyMVar, putMVar,
takeMVar, ThreadId, threadDelay)
import Control.Monad (forever, liftM)
class Stream a b c d where
(-) :: a - (b - c) - d
instance Stream (IO d) d (IO c) (IO c) where
f - g = f = g
instance Stream d d (IO c) (IO c) where
f - g = g f
instance Stream d d c c where
x - y = y $ x
-- This simply wraps a string in brackets.
bracket :: String - String
bracket x = ( ++ x ++ )
lbracket :: IO String - IO String
lbracket x = liftM bracket x
hello :: String
hello = Hello World!
lhello :: IO String
lhello = do return hello
main :: IO ()
main = do
forkIO $ (bracket $ hello) - putStrLn
forkIO $ lhello - lbracket - putStrLn
forkIO $ bracket hello - putStrLn
forkIO $ lbracket lhello - putStrLn
threadDelay 1000 -- Sleep for at least 10 seconds before exiting.
working_but_ugly.hs:
{-# LANGUAGE MultiParamTypeClasses, FlexibleInstances,
TypeSynonymInstances, OverlappingInstances #-}
{-# OPTIONS_GHC #-}
module Main (
main
)
where
import Control.Concurrent (forkIO, MVar, newEmptyMVar, putMVar,
takeMVar, ThreadId, threadDelay)
import Control.Monad (forever, liftM)
class Stream a b c d where
(-) :: a - (b - c) - d
instance Stream (IO d) d (IO c) (IO c) where
f - g = f = g
instance Stream d d (IO c) (IO c) where
f - g = g f
instance Stream d d c c where
x - y = y $ x
x - y = y $ x
-- This simply wraps a string in brackets.
bracket :: String - String
bracket x = ( ++ x ++ )
lbracket :: IO String - IO String
lbracket x = liftM bracket x
hello :: String
hello = Hello World!
lhello :: IO String
lhello = do return hello
main :: IO ()
main = do
forkIO $ (bracket $ hello) - putStrLn
forkIO $ lhello - lbracket - putStrLn
forkIO $ bracket hello - putStrLn
forkIO $ lbracket lhello - putStrLn
threadDelay 1000 -- Sleep for at least 10 seconds before exiting.
___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Type Classes in Haskell - how can I make GHC make a choice of types, when the type chosen doesn't matter?
On 14/04/11 13:00, Chris Dew wrote: class Stream a b c d where (-) :: a - (b - c) - d instance Stream (IO d) d (IO c) (IO c) where f - g = f= g instance Stream d d (IO c) (IO c) where f - g = g f instance Stream d d c c where x - y = y $ x I notice that in all your instances, the last two types are the same. So do you need the final type parameter? Could you not make it: class Stream a b c where (-) :: a - (b - c) - c I quickly tried this, and it fixes the errors you were getting. If that doesn't hold for all instances you have in mind, then you may want to use functional dependencies or type families to specify a relationship between the types. Thanks, Neil. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Type Classes in Haskell - how can I make GHC make a choice of types, when the type chosen doesn't matter?
Hi Chris What does the Stream class *do* though? class Stream a b c d where (-) :: a - (b - c) - d Even with Neil's change its still quite unusual: class Stream a b c where (-) :: a - (b - c) - c In the first formulation there is an input of type a, a function (b - c) and a result of a completely different type d. In Neil's class the function relates to the type of the answer but not to the input. The difficult type classes in Haskell - Applicative, Monad, and Arrows / Category - are related to some degree to fairly standard combinators on functions. But they generalize the combinators to operate on other types than the function type (-). As there isn't a relation between input and output, I don't quite see how the Stream type could start as a combinator. Best wishes Stephen ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Programming Chalenges: The 3n+1 problem
Hi Dimitri, When asking how to implement cache in Haskell I was hopping that there exists some solution without using Data.Array, more functional approach, if I may say so ... Steven's second solution is purely functional. It uses so-called tries to cache results instead of mutable arrays. Here is an alternative definition of Steven's `fixmemo` function: fixmemo :: HasTrie a = ((a - b) - (a - b)) - (a - b) fixmemo f = fix (memo . f) It uses the standard fixpoint combinator Data.Function.fix :: (a - a) - a and has a similar (but more restrictive) type. In order to understand Steven's solution you need to know how to define recursive functions using a fixpoint combinator. For example, you can define the Fibonacci function as follows: fibonacci :: Int - Integer fibonacci = fix fib fib :: (Int - Integer) - (Int - Integer) fib fib 0 = 0 fib fib 1 = 1 fib fib n = fib (n-1) + fib (n-2) Note how the first argument `fib` of the function `fib` shadows the name of the global function `fib`. The advantage of this complicated definition is that you get a memoized version of the `fibonacci` function simply by using `fixmemo` instead of `fix`: memoFib :: Int - Integer memoFib = fixmemo fib This definition avoids to recompute the same recursive calls over and over again in is therefore much faster than the original version. That said, memoization helps to solve your original problem only if you reuse the cache for different top-level calls of the Collatz length function. According to the Collatz conjecture no argument appears again when computing the Collatz length of a number (otherwise the computation would not terminate). Ryan's solution achieves reuse of the cache by defining it at the top level and you can do the same with tries. For the `fib` example it would look like this: fibCache :: Int :-: Integer fibCache = trie (fib cachedFib) cachedFib :: Int - Integer cachedFib = untrie fibCache Now even independent calls to `cachedFib` reuse previously computed results. In my experiments, caching did not pay off, even for computing the maximum Collatz length of a given range. Without caching it took less than 5 seconds to compute the maximum Collatz length of all numbers between 1 and 100. With caching, it took 30 seconds. Cheers, Sebastian ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Munich functional programmers' UG
Dear all, last week, there have been thoughts to create a user group for functional programmers in Munich, Germany. People who are interessted can join the new mailing list: http://www.fs.lmu.de/cgi-bin/mailman/listinfo/high-order-munich The idea is to have meetings every now and then with talks on aspects of functional programming. However, it's not restricted to Haskell. Please spread the word to anyone who could be interested. Thank you, Heinrich -- hoerde...@funktional.info www.funktional.info -- ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] parallel-haskell mailing list
Hi everybody, I thought I should point out that there is a mailing list dedicated to parallelism and concurrency in Haskell. The group exists to bring together the various groups in the Haskell community that are working on parallelism. It is intended to provide some visibility into the various efforts and to encourage collaboration. So if you want to know what's going on in the parallel Haskell world, this is the place to hang out. You can join the list by sending an email to parallel-haskell+subscr...@googlegroups.com You can also visit the group on the web at http://groups.google.com/group/parallel-haskell Or for those of you who prefer GMane http://dir.gmane.org/gmane.comp.lang.haskell.parallel Enjoy! Eric -- Got news for the Parallel Haskell Digest? Send a mail to e...@well-typed.com! pgp2azpIL6ffL.pgp Description: PGP signature ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Questioning seq
A couple of questions: 1. Why is the existence of polymorphic seq bad? 2. Why would the existence of polymorphic rnf be worse? 3. How is pseq different from seq? That is all... ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Programming Chalenges: The 3n+1 problem
On Thu, Apr 14, 2011 at 4:29 AM, Dmitri O.Kondratiev doko...@gmail.comwrote: 3n+1 is the first, warm-up problem at Programming Chalenges site: http://www.programming-challenges.com/pg.php?page=downloadproblemprobid=110101format=html (This problem illustrates Collatz conjecture: http://en.wikipedia.org/wiki/3n_%2B_1#Program_to_calculate_Collatz_sequences ) As long as the judge on this site takes only C and Java solutions, I submitted in Java some add-hock code (see at the end of this message) where I used recursion and a cache of computed cycles. Judge accepted my code and measured 0.292 sec with best overall submissions of 0.008 sec to solve the problem. *** Question: I wonder how to implement cache for this problem in Haskell? At the moment, I am not so much interested in the speed of the code, as in nice implementation. This is the exact problem data-memocombinators was written to solve. http://hackage.haskell.org/packages/archive/data-memocombinators/0.4.1/doc/html/Data-MemoCombinators.html For this problem, it is too slow to memoize everything; you have to use a bounded memo table. That's why I use a combinator-based memo approach as opposed to the type-directed approach used in eg. MemoTrie. The memo table you need is something like switch (10^6) integral id Luke ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Questioning seq
Am 14.04.2011 18:57:35 schrieb Andrew Coppinandrewcop...@btinternet.com: A couple of questions: 1. Why is the existence of polymorphic seq bad? See http://www.iai.uni-bonn.de/~jv/acta.pdf, Sections 1 and 2 and pointers therein. Also, numerous discussions on this list over the years. Best, Janis. 2. Why would the existence of polymorphic rnf be worse? 3. How is pseq different from seq? That is all... -- Jun.-Prof. Dr. Janis Voigtländer http://www.iai.uni-bonn.de/~jv/ mailto:j...@iai.uni-bonn.de ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Questioning seq
As usual, I'm foolish and forget to hit 'reply to all'. Original message unedited below, so it can be sent to -cafe. To answer question #3, pseq and seq are semantically equivalent (indeed, if you look at the source for Control.Parallel, if you are not using GHC, pseq is defined as 'pseq = seq'.) There is a subtle operational difference however: seq is strict in both of its arguments, while pseq is only strict in its first argument as far as GHC is concerned. When you annotate code for parallelism, seq is a little more problematic, because you want more control over the evaluation order. For example, given a `seq` b, the compiler can freely rearrange this in a number of ways - but if we are evaluating code in parallel, that's a little too strict. So typically you want to say a `pseq` b - you will strictly evaluate 'a' before 'b', presumably because 'b' has already been sparked off in parallel using `par`. If you were using seq, the compiler can rearrange a `seq` b into say, b `seq` a `seq` b. Which won't gain you anything from a parallel perspective, for the most part. For more info, see the Control.Parallel module: http://hackage.haskell.org/packages/archive/parallel/3.1.0.1/doc/html/Control-Parallel.html Also see Simon Simon's paper, 'Runtime support for Multicore Haskell', particularly section 2, 'Background: programming model'. http://community.haskell.org/~simonmar/papers/multicore-ghc.pdf On Thu, Apr 14, 2011 at 12:57 PM, Andrew Coppin andrewcop...@btinternet.com wrote: A couple of questions: 1. Why is the existence of polymorphic seq bad? 2. Why would the existence of polymorphic rnf be worse? 3. How is pseq different from seq? That is all... ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe -- Regards, Austin ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Type Classes in Haskell - how can I make GHC make a choice of types, when the type chosen doesn't matter?
@Neil Brown - That did it. It's not the ideal solution, as all - are 'coerced' into being 'IO x' (if the rightmost term is an 'IO x'. But it'll do for the time being. Many thanks, Chris. On 14 April 2011 13:50, Neil Brown nc...@kent.ac.uk wrote: On 14/04/11 13:00, Chris Dew wrote: class Stream a b c d where (-) :: a - (b - c) - d instance Stream (IO d) d (IO c) (IO c) where f - g = f= g instance Stream d d (IO c) (IO c) where f - g = g f instance Stream d d c c where x - y = y $ x I notice that in all your instances, the last two types are the same. So do you need the final type parameter? Could you not make it: class Stream a b c where (-) :: a - (b - c) - c I quickly tried this, and it fixes the errors you were getting. If that doesn't hold for all instances you have in mind, then you may want to use functional dependencies or type families to specify a relationship between the types. Thanks, Neil. ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Type Classes in Haskell - how can I make GHC make a choice of types, when the type chosen doesn't matter?
@Stephen Tetley - The stream class exists simply to allow for the creation of a - operator which can be used to 'Stream' data through multiple pure and IO functions, on the way to some form of output. It's probably not a great idea, as there are more idiomatic solutions in Haskell - I'm sure someone will mention arrows. I want the result of (-) to be what the following function requires, either an 'a' or and 'IO a'. This is too unconstrained if the following function is flexible in it's input. (e.g. another application of (-)). Hence my original problem. a and b have are related, but not in a way I know how to express in Haskell. They are constrained to: a == b || IO a == b || a == IO b. c and d have a similar constraint. Could you suggest how these constraints could be expressed in the Haskell type system? Thanks, Chris. On 14 April 2011 14:28, Stephen Tetley stephen.tet...@gmail.com wrote: Hi Chris What does the Stream class *do* though? class Stream a b c d where (-) :: a - (b - c) - d Even with Neil's change its still quite unusual: class Stream a b c where (-) :: a - (b - c) - c In the first formulation there is an input of type a, a function (b - c) and a result of a completely different type d. In Neil's class the function relates to the type of the answer but not to the input. The difficult type classes in Haskell - Applicative, Monad, and Arrows / Category - are related to some degree to fairly standard combinators on functions. But they generalize the combinators to operate on other types than the function type (-). As there isn't a relation between input and output, I don't quite see how the Stream type could start as a combinator. Best wishes Stephen ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Type Classes in Haskell - how can I make GHC make a choice of types, when the type chosen doesn't matter?
On 14 April 2011 20:35, Chris Dew cms...@gmail.com wrote: Could you suggest how these constraints could be expressed in the Haskell type system? Hi Chris I'm afriad I'd have to decline - generally in Haskell implicit lifters are problematic, so it isn't something I'd be looking to solve. There was a thread on Haskell Cafe about them last November called Making monadic code more concise, that you might find interesting - especially Oleg Kiselyov's comments: http://www.haskell.org/pipermail/haskell-cafe/2010-November/086445.html Best wishes Stephen ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Type Classes in Haskell - how can I make GHC make a choice of types, when the type chosen doesn't matter?
Thanks, that link's very relevant to what I'm trying. For the time being I'll accept a partial solution where the last two types are now the same, and try to improve it when my knowledge of Haskell improves. I really want (hello - bracket) in (hello - bracket - putStrLn) to have a type of String. Using the partial solution which Neil Brown proposed, the code will work, but (hello - bracket) will have a type of IO String which *seems* like it will be less efficient. All the best, Chris. On 14 April 2011 21:22, Stephen Tetley stephen.tet...@gmail.com wrote: On 14 April 2011 20:35, Chris Dew cms...@gmail.com wrote: Could you suggest how these constraints could be expressed in the Haskell type system? Hi Chris I'm afriad I'd have to decline - generally in Haskell implicit lifters are problematic, so it isn't something I'd be looking to solve. There was a thread on Haskell Cafe about them last November called Making monadic code more concise, that you might find interesting - especially Oleg Kiselyov's comments: http://www.haskell.org/pipermail/haskell-cafe/2010-November/086445.html Best wishes Stephen ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Questioning seq
On 11-04-14 01:57 PM, Andrew Coppin wrote: 3. How is pseq different from seq? An example to show that there are non-unique evaluation orders to fulfill the mere strictness promise of seq: import Data.List(foldl') () () = () main = print (foldl' () () (replicate 250 ())) with ghc with -O0 : fast, no stack overflow with ghc with -O or -O2: slow, stack overflow ghc versions 6.12.3, 7.0.2, 7.0.3 Look at the core code (-ddump-simpl) to verify evaluation orders. Repeat the experiment and observations with pseq: import GHC.Conc(pseq) () () = () foldl' :: (a - b - a) - a - [b] - a foldl' op z xs = go z xs where go z [] = z go z (x:xs) = let z' = op z x in z' `pseq` go z' xs main = print (foldl' () () (replicate 250 ())) ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
[Haskell-cafe] Experimental 64 bit installer fails
I downloaded the experimental 64 bit installer for from Haskell Platformhttp://hackage.haskell.org/platform/mac.html. It fails with the message The installation failed. In case Haskell Platform i386 was interfering, I uninstalled it. But the 64 bit installer still yields The installation failed. Specs - Haskell Platform 2011.2.0.0-x86_64 installer - Haskell Platform 2011.2.0.0-i386 present - Mac OS X 10.6.7 - MacBook Pro 5,1 /var/log/install.log Apr 14 17:58:38 wonko Installer[18657]: Haskell Platform Installation Log Apr 14 17:58:38 wonko Installer[18657]: Opened from: /Users/andrew/Downloads/Haskell Platform 2011.2.0.0-x86_64.pkg Apr 14 17:58:46 wonko runner[18669]: Administrator authorization granted. Apr 14 17:58:46 wonko Installer[18657]: Apr 14 17:58:46 wonko Installer[18657]: User picked Standard Install Apr 14 17:58:46 wonko Installer[18657]: Choices selected for installation: Apr 14 17:58:46 wonko Installer[18657]: Upgrade: Haskell Platform Apr 14 17:58:46 wonko Installer[18657]: Upgrade: Glasgow Haskell Compiler Apr 14 17:58:46 wonko Installer[18657]: Haskell Platform 2011.2.0.0-x86_64.pkg#ghc.pkg : org.haskell.HaskellPlatform.GHC.pkg : 1 Apr 14 17:58:46 wonko Installer[18657]: Upgrade: Haskell Platform Libraries Apr 14 17:58:46 wonko Installer[18657]: Haskell Platform 2011.2.0.0-x86_64.pkg#haskell.pkg : org.haskell.HaskellPlatform.Libraries.pkg : 1 Apr 14 17:58:46 wonko Installer[18657]: Haskell Platform 2011.2.0.0-x86_64.pkg#haskellPlatformPostflight.pkg : org.haskell.haskellPlatform.postflight.pkg : 1.0 Apr 14 17:58:46 wonko Installer[18657]: Apr 14 17:58:46 wonko Installer[18657]: It took 0.00 seconds to summarize the package selections. Apr 14 17:58:46 wonko Installer[18657]: -[IFDInstallController(Private) _buildInstallPlan]: location = file://localhost Apr 14 17:58:46 wonko Installer[18657]: -[IFDInstallController(Private) _buildInstallPlan]: file://localhost/Users/andrew/Downloads/Haskell%20Platform%202011.2.0.0-x86_64.pkg#ghc.pkg Apr 14 17:58:46 wonko Installer[18657]: -[IFDInstallController(Private) _buildInstallPlan]: file://localhost/Users/andrew/Downloads/Haskell%20Platform%202011.2.0.0-x86_64.pkg#haskell.pkg Apr 14 17:58:46 wonko Installer[18657]: -[IFDInstallController(Private) _buildInstallPlan]: file://localhost/Users/andrew/Downloads/Haskell%20Platform%202011.2.0.0-x86_64.pkg#haskellPlatformPostflight.pkg Apr 14 17:58:46 wonko Installer[18657]: Will use PK session Apr 14 17:58:46 wonko Installer[18657]: Starting installation: Apr 14 17:58:46 wonko Installer[18657]: Configuring volume Wonko Apr 14 17:58:46 wonko Installer[18657]: Preparing disk for local booted install. Apr 14 17:58:46 wonko Installer[18657]: Free space on Wonko: 52.51 GB (52508946432 bytes). Apr 14 17:58:46 wonko Installer[18657]: Create temporary directory /var/folders/N-/N-IxaaKuFae5ik1WKxV6wE+++TI/-Tmp-//Install.18657bWOdda Apr 14 17:58:46 wonko Installer[18657]: IFPKInstallElement (3 packages) Apr 14 17:58:46 wonko installd[18673]: PackageKit: - Begin install - Apr 14 17:58:46 wonko installd[18673]: PackageKit: request=PKInstallRequest 3 packages, destination=/ Apr 14 17:58:46 wonko installd[18673]: PackageKit: packages=(\n PKLeopardPackage file://localhost/Users/andrew/Downloads/Haskell%20Platform%202011.2.0.0-x86_64.pkg#ghc.pkg,\n PKLeopardPackage file://localhost/Users/andrew/Downloads/Haskell%20Platform%202011.2.0.0-x86_64.pkg#haskell.pkg,\n PKLeopardPackage file://localhost/Users/andrew/Downloads/Haskell%20Platform%202011.2.0.0-x86_64.pkg#haskellPlatformPostflight.pkg\n) Apr 14 17:58:47 wonko installd[18673]: PackageKit: Extracting file://localhost/Users/andrew/Downloads/Haskell%20Platform%202011.2.0.0-x86_64.pkg#ghc.pkg (destination=/var/folders/zz/zzzivhrRnAmviuee+++/Cleanup At Startup/PKInstallSandbox-tmp/Root/Library/Frameworks, uid=0) Apr 14 17:59:06 wonko installd[18673]: PackageKit: Extracting file://localhost/Users/andrew/Downloads/Haskell%20Platform%202011.2.0.0-x86_64.pkg#haskell.pkg (destination=/var/folders/zz/zzzivhrRnAmviuee+++/Cleanup At Startup/PKInstallSandbox-tmp/Root/Library, uid=0) Apr 14 17:59:12 wonko installd[18673]: PackageKit: Extracting file://localhost/Users/andrew/Downloads/Haskell%20Platform%202011.2.0.0-x86_64.pkg#haskellPlatformPostflight.pkg (destination=/var/folders/zz/zzzivhrRnAmviuee+++/Cleanup At Startup/PKInstallSandbox-tmp/Root, uid=0) Apr 14 17:59:14 wonko installd[18673]: PackageKit: Shoving /var/folders/zz/zzzivhrRnAmviuee+++/Cleanup At Startup/PKInstallSandbox-tmp/Root (1 items) to / Apr 14 17:59:16 wonko installd[18673]: PackageKit: Executing script ./postinstall in /private/tmp/PKInstallSandbox.e59JtF/Scripts/org.haskell.HaskellPlatform.GHC.pkg.k6SuIR Apr 14 17:59:16 wonko installd[18673]: PackageKit: Executing script ./postinstall in
[Haskell-cafe] ghc7 bug? associated type + implicit parameters
Hi,
I have stumbled across some strange behaviour in ghc7.
The following compiles fine with ghc 6.12.3:
{-# LANGUAGE ImplicitParams #-}
{-# LANGUAGE TypeFamilies #-}
{-# LANGUAGE NoMonomorphismRestriction #-}
module Test where
class C t where
type TF t
ttt :: TF t - t
b :: (C t, ?x :: TF t) = t
b = ttt ?x
but ghc7 says:
Could not deduce (?x::TF t)
arising from a use of implicit parameter `?x'
from the context (C t, ?x::TF t)
bound by the type signature for b :: (C t, ?x::TF t) = t
at Test.hs:13:1-10
In the first argument of `ttt', namely `?x'
In the expression: ttt ?x
In an equation for `b': b = ttt ?x
Moreover, when I comment out the type declaration for b, it compiles and the
inferred type for b is identical to the one in the above program:
*Test :t b
b :: (C t, ?x::TF t) = t
It feels to me like a bug but I am not entirely confident. Any ideas?
Michal
--
|o| Michal Konecny mikkone...@gmail.com
|o|http://www-users.aston.ac.uk/~konecnym/
|o|office: (+42) (0)121 204 3462
|o| PGP key http://www-users.aston.ac.uk/~konecnym/ki.aston
--
|o| Michal Konecny mikkone...@gmail.com
|o|http://www-users.aston.ac.uk/~konecnym/
|o|office: (+42) (0)121 204 3462
|o| PGP key http://www-users.aston.ac.uk/~konecnym/ki.aston
--
|o| Michal Konecny mikkone...@gmail.com
|o|http://www-users.aston.ac.uk/~konecnym/
|o|office: (+42) (0)121 204 3462
|o| PGP key http://www-users.aston.ac.uk/~konecnym/ki.aston
signature.asc
Description: This is a digitally signed message part.
___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Experimental 64 bit installer fails
Well that's no fun! The install looks like it mostly worked, execept that the final registration of the installed packages failed because for some reason the script has them out of order. You can fix up your install by doing this: cd /Library/Haskell/ghc-7.0.2/lib/registrations for c in *.conf; do echo == $c ==; ghc-pkg register --force $c; done I'll have to look into why that build of the package got the files out of order... - Mark ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Experimental 64 bit installer fails
I can confirm that Mark's Haskell 2011.2.0.0-x86_64 Mac OS X installer fix works :) cd /Library/Haskell/ghc-7.0.2/lib/registrations for c in *.conf; do echo == $c ==; ghc-pkg register --force $c; done My goal is to get Haskell ncurses working. c2hs is required, so I did: $ cabal --version cabal-install version 0.10.2 using version 1.10.1.0 of the Cabal library $ cabal install c2hs $ c2hs -bash: c2hs: command not found It appears Cabal never added ~/Library/Haskell/ghc-7.0.2/lib/c2hs-0.16.3/bin to PATH. It's easy enough for me personally to fix, but Haskell 2010 didn't have this problem. Cheers, Andrew Pennebaker www.yellosoft.us On Thu, Apr 14, 2011 at 7:28 PM, Mark Lentczner mark.lentcz...@gmail.com wrote: Well that's no fun! The install looks like it mostly worked, execept that the final registration of the installed packages failed because for some reason the script has them out of order. You can fix up your install by doing this: cd /Library/Haskell/ghc-7.0.2/lib/registrations for c in *.conf; do echo == $c ==; ghc-pkg register --force $c; done I'll have to look into why that build of the package got the files out of order... - Mark ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] ghc7 bug? associated type + implicit parameters
Someone on the ghc-users list might have an idea. I recommend filing a
bug report:
http://hackage.haskell.org/trac/ghc/wiki/ReportABug
Even if the behavior isn't wrong, in could use a better error message!
Antoine
2011/4/14 Michal Konečný mikkone...@googlemail.com:
Hi,
I have stumbled across some strange behaviour in ghc7.
The following compiles fine with ghc 6.12.3:
{-# LANGUAGE ImplicitParams #-}
{-# LANGUAGE TypeFamilies #-}
{-# LANGUAGE NoMonomorphismRestriction #-}
module Test where
class C t where
type TF t
ttt :: TF t - t
b :: (C t, ?x :: TF t) = t
b = ttt ?x
but ghc7 says:
Could not deduce (?x::TF t)
arising from a use of implicit parameter `?x'
from the context (C t, ?x::TF t)
bound by the type signature for b :: (C t, ?x::TF t) = t
at Test.hs:13:1-10
In the first argument of `ttt', namely `?x'
In the expression: ttt ?x
In an equation for `b': b = ttt ?x
Moreover, when I comment out the type declaration for b, it compiles and the
inferred type for b is identical to the one in the above program:
*Test :t b
b :: (C t, ?x::TF t) = t
It feels to me like a bug but I am not entirely confident. Any ideas?
Michal
--
|o| Michal Konecny mikkone...@gmail.com
|o| http://www-users.aston.ac.uk/~konecnym/
|o| office: (+42) (0)121 204 3462
|o| PGP key http://www-users.aston.ac.uk/~konecnym/ki.aston
--
|o| Michal Konecny mikkone...@gmail.com
|o| http://www-users.aston.ac.uk/~konecnym/
|o| office: (+42) (0)121 204 3462
|o| PGP key http://www-users.aston.ac.uk/~konecnym/ki.aston
--
|o| Michal Konecny mikkone...@gmail.com
|o| http://www-users.aston.ac.uk/~konecnym/
|o| office: (+42) (0)121 204 3462
|o| PGP key http://www-users.aston.ac.uk/~konecnym/ki.aston
___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
___
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Programming Chalenges: The 3n+1 problem
Hi Dmitri, *** Question: I wonder how to implement cache for this problem in Haskell? At the moment, I am not so much interested in the speed of the code, as in nice implementation. Yet another option for memoization implementation: to use monad-memo package [1] which provides memoization for monadic function (using Data.Map by default). To use it we need to define our recursive function in monadic form and add `memo` in place of its recursive call: import Control.Applicative import Control.Monad.Memo -- calculates the length of sequence (with memoization) calcM 1 = return 1 calcM n = succ $ memo calcM (if even n then n `div` 2 else n*3 + 1) Here is a helper-function to run this calculation (we don't really need it here since `calcM` is going to be called from other monadic function directly): runCalc :: Integer - Integer runCalc = startEvalMemo . calcM NB: the inferred type for `calcM` might look a little bit.. verbose, but we don't really need to specify it or expose `calcM` from our module: calcM :: (MonadMemo a1 a m, Num a, Functor m, Integral a1, Enum a) = a1 - m a The implementation of the function to calculate the maximum length of the sequence for all numbers in a range is straightforward (and also monadic): -- NB: memoization cache is shared among all 'calcM' calls (very efficient) calcRangeM f t = maximum $ forM [f..t] calcM and a similar helper run-function (is not needed for the task either): runCalcRange :: Integer - Integer - Integer runCalcRange f t = startEvalMemo $ calcRangeM f t To run `calcRangeM` for the input list of values (map the function over it) we can define yet another simple monadic function which calls `calcRangeM` directly (thus reusing/building the same memoization cache): solM = mapM (uncurry calcRangeM) and a helper run-function: runSol :: [(Integer, Integer)] - [Integer] runSol = startEvalMemo . solM Composing all these together results in the following test code (I hard-coded the input for the sake of simplicity): import Control.Applicative import Control.Monad.Memo calcM 1 = return 1 calcM n = succ $ memo calcM (if even n then n `div` 2 else n*3 + 1) calcRangeM f t = maximum $ forM [f..t] calcM solM = mapM (uncurry calcRangeM) runSol = startEvalMemo . solM main = print $ runSol [ (1, 10), (100, 200), (201, 210), (900, 1000) ] As for the performance, `main = print $ runSol [(1, 100)]` takes ~40 seconds with -O2 on my laptop. [1] http://hackage.haskell.org/package/monad-memo ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
Re: [Haskell-cafe] Problem building qtHaskell
Thanks for the tip! I reinstalled everything from scratch here, and then I got it working. On Wed, Apr 13, 2011 at 2:26 AM, Mark Wright markwri...@internode.on.net wrote: On Tue, 12 Apr 2011 12:14:50 +0200, Øystein Kolsrud kols...@gmail.com wrote: Hi! I am trying to install qtHaskell on a Windows 7 machine, but I am running into some problems while running the build script. I believe I have followed all the steps stated in the installation instructions, but get errors of this type: Creating library file: ..\bin\libqtc_core.a c:\Prog\Qt\2010.05\qt\lib/libQtUiTools.a(quiloader.o):quiloader.cpp:(.text+0x3c3): undefined reference to `_Unwind_Resume' Hi Øystein, Maybe this might give some hints: http://www.qtcentre.org/threads/33394-Link-errors-undefined-reference-to-_Unwind_Resume-and-__gxx_personality_v0 It suggests checking if c:\Prog\Qt\2010.05\qt\lib/libQtUiTools.a was built with the same g++ version. Regards, Mark Can anyone help me figure out what is missing? I have the following versions installed which should meet the requirements: C:\Prog\QTHASK~1.4g++ --version g++ (GCC) 4.4.3 C:\Prog\QTHASK~1.4gmake --version GNU Make 3.81 This program built for Windows32 C:\Prog\QTHASK~1.4ghc --version The Glorious Glasgow Haskell Compilation System, version 7.0.2 Thanks! Best regards, Øystein Kolsrud ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe -- Mvh Øystein Kolsrud ___ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe
