- Beginners - Haskell.org

Compilation error
by mukesh tiwari 21 Jul '11

21 Jul '11

Kindly some one please tell me why this code in not compiling . I have to round a Double value up to two decimal places and i wrote this code for this problem [ http://www.spoj.pl/problems/QCJ4 ] . Thank you Mukesh Tiwari import Data.List import qualified Data.Sequence as DS import Text.Printf data Point a = P a a deriving ( Show , Eq , Ord ) data Turn = S | L | R deriving ( Show , Eq , Ord , Enum ) -- straight left right compPoint :: ( Num a , Ord a ) => Point a -> Point a -> Ordering compPoint ( P x1 y1 ) ( P x2 y2 ) | compare x1 x2 == EQ = compare y1 y2 | otherwise = compare x1 x2 findMinx :: ( Num a , Ord a ) => [ Point a ] -> [ Point a ] findMinx xs = sortBy ( \x y -> compPoint x y ) xs compAngle ::(Num a , Ord a ) => Point a -> Point a -> Point a -> Ordering compAngle ( P x1 y1 ) ( P x2 y2 ) ( P x0 y0 ) = compare ( ( y1 - y0 ) * ( x2 - x0 ) ) ( ( y2 - y0) * ( x1 - x0 ) ) sortByangle :: ( Num a , Ord a ) => [ Point a ] -> [ Point a ] sortByangle (z:xs) = z : sortBy ( \x y -> compAngle x y z ) xs convexHull ::( Num a , Ord a ) => [ Point a ] -> [ Point a ] convexHull xs = reverse . findHull [y,x] $ ys where (x:y:ys) = sortByangle.findMinx $ xs findTurn :: ( Num a , Ord a , Eq a ) => Point a -> Point a -> Point a -> Turn findTurn ( P x0 y0 ) ( P x1 y1 ) ( P x2 y2 ) | ( y1 - y0 ) * ( x2- x0 ) < ( y2 - y0 ) * ( x1 - x0 ) = L | ( y1 - y0 ) * ( x2- x0 ) == ( y2 - y0 ) * ( x1 - x0 ) = S | otherwise = R findHull :: ( Num a , Ord a ) => [ Point a ] -> [ Point a ] -> [ Point a ] findHull [x] ( z : ys ) = findHull [ z , x ] ys --incase of second point on line from x to z findHull xs [] = xs findHull ( y : x : xs ) ( z:ys ) | findTurn x y z == R = findHull ( x : xs ) ( z:ys ) | findTurn x y z == S = findHull ( x : xs ) ( z:ys ) | otherwise = findHull ( z : y : x : xs ) ys --from here on testing part for SPOJ format::(Num a , Ord a ) => [[a]] -> [Point a] format xs = map (\[x0 , y0] -> P x0 y0 ) xs helpSqrt :: ( Floating a ) => Point a -> Point a -> a helpSqrt ( P x0 y0 ) ( P x1 y1 ) = sqrt $ ( x0 - x1 ) ^ 2 + ( y0 - y1 ) ^ 2 solve :: ( Num a , RealFrac a , Floating a ) => [ Point a ] -> a solve xs = d where d = snd . foldl ( \( P x0 y0 , s ) ( P x1 y1 ) -> ( P x0 y0 , max s $ 2.0 * helpSqrt ( P x0 y0 ) ( P x1 y1 ) ) ) ( P x y , 0 ) $ xs where ( P x y ) = cMass xs cMass :: ( Num a , RealFrac a , Floating a ) => [ Point a ] -> Point a cMass xs = P x y where ( P x0 y0 ) = foldl ( \( P x1 y1 ) (P x2 y2 ) -> P ( x1 + x2 ) ( y1 + y2 ) ) ( P 0 0 ) xs n = genericLength xs x = x0 / n y = x0 / n readInt ::( Num a , Read a ) => String -> a readInt = read main = do let l = solve . convexHull . format . map ( map readInt . words ) . tail . lines printf "%.2f\n" l return () {-- main = interact $ solve . convexHull . format . map ( map readInt . words ) . tail . lines --} The error is [1 of 1] Compiling Main ( qcj4_6044.hs, qcj4_6044.o ) qcj4_6044.hs:69:1: No instance for (PrintfArg (String -> a)) arising from a use of `printf' at qcj4_6044.hs:69:1-18 Possible fix: add an instance declaration for (PrintfArg (String -> a)) In a stmt of a 'do' expression: printf "%.2f\n" l In the expression: do { let l = solve . convexHull . format . map (map readInt . words) . tail . lines; printf "%.2f\n" l; return () } In the definition of `main': main = do { let l = ...; printf "%.2f\n" l; return () }

3 2

What book is being referred to: "... in SPJs beautiful code chapter ..."
by Costello, Roger L. 21 Jul '11

21 Jul '11

Hi Folks, I saw this mention [1]: I like the terminology in SPJs beautiful code chapter. He talks about IO *actions* and how IO actions can be glued together into larger IO actions, and that main is then an IO action in itself. Do you know what book is being referred to? /Roger [1] A comment by Sebastian Sylvan in: http://book.realworldhaskell.org/read/io.html

1 0

(Implicit) equality testing using multiple function definitions
by Tom Murphy 21 Jul '11

21 Jul '11

Hi list! When I define an algebraic datatype without an instance for Eq, I'm obviously unable to use the (==) function on it. I can pattern-match with a series of function definitions (f [] = False; f x = True) on the expression, though. Why is that? I understand that in the second case I'm not literally using the (==) function, but it seems like there would be instances where you'd intentionally not want to be able to test for equality, and pattern-matching with multiple function definitions circumvents that. Thanks for your time, Tom

7 12

loop problem
by Roelof Wobben 20 Jul '11

20 Jul '11

Hello, I need to make a sort of loop so a text is displayed several times. So this idea : repeat 3 true will be True True True. So in another language I would have programmed. for teller = 1 to a print text But I need to use list comprehession. So I thougt I use this : replicate :: Int -> a -> [a] replicate a xs = [xs | x' <= a <-xs] But now I get this message : oefening.hs:2:24: Parse error in pattern: x' <= a Where did I go wrong here ? Roelof

2 1

Parse error: naked expression at top level
by Roelof Wobben 20 Jul '11

20 Jul '11

Hello, I made this little script : replicate :: Int -> a -> [a] replicate a xs = [xs | x' <= a <--xs] replicate 3 True But when I let it run in GHCi, version 7.0.3 on a FreeBSD box I see this error : oefening.hs:4:1: Parse error: naked expression at top level Failed, modules loaded: none. How can I solve this and what does it mean. Roelof

5 5

Re: [Haskell-beginners] Can fields in a record be optional?
by Antoine Latter 20 Jul '11

20 Jul '11

I would expect I would run into trouble with this approach - when a function receives one of these records as an argument, how does it know if it is safe to acces the record fields in question? I would prefer using Maybe types, different types altogether for summed contracts, or imaking a Contract a type with two cases (as suggested earlier). On Jul 17, 2011 5:19 PM, "David Place" <d(a)vidplace.com> wrote:

2 1

Can fields in a record be optional?
by Costello, Roger L. 20 Jul '11

20 Jul '11

Hi Folks, Can fields in a record be made "optional"? Let me motivate my question: I have a data type called "Contract". It is a record. It has one field that holds a currency value, a second field that holds payments, and a third field to hold sub-Contracts. data Contract = Contract { currency :: Currency , payments :: Double , contracts :: [Contract] } deriving (Show) Here is a function that creates a Contract with currency c and payments = 1: one :: Currency -> Contract one c = Contract { currency = c, payments = 1, contracts = [] } Since there are no sub-Contracts I'd rather not specify that last field. Here is a function that AND's two Contracts, c1 and c2, and returns one Contract: and :: Contract -> Contract -> Contract (and) c1 c2 = Contract { currency = undefined, payments = undefined, contracts = [c1, c2] } I'd rather not specify those first two fields. See why I desire optional fields? Perhaps I am going about it completely wrong and there is a better way to express Contract? I am eager to learn! /Roger

8 15

How to correctly define data types?
by Davi Santos 19 Jul '11

19 Jul '11

Hello (first post), I have spent so many time learning typeclasses and thinking it was part of Haskell essence... and suddenly I read the discussion "[Haskell-beginners] Can fields in a record be optional?". So typeclasses are not always recomended? I'm implementing a Machine Learning framework and I am in a sort of related dilemma. I found three ways of implementing the same distance function between "examples" (aka "attribute vectors" or simply "Float vectors" for mere mortals :) ): [obs: "Example" datatype will be added more fields later] --------------first------------------------------------ module ML where data Example = Example [Float] deriving (Show) class ExampleClass a where (distance) :: a → a → Float instance ExampleClass Example where (Example atts1) distance (Example atts2) = sqrt $ sum $ map (λ(x, y) → (x-y)↑2) $ zip atts1 atts2 ================================= --------------second------------------------------------ module ML where data Example = Example {attributes :: [Float]} deriving (Show) distance :: Example → Example → Float distance ex1 ex2 = sqrt $ sum $ map (λ(x, y) → (x-y)↑2) $ zip (attributes ex1) (attributes ex2) ================================= --------------third------------------------------------ module ML where data Example = Example [Float] deriving (Show) distance :: Example → Example → Float distance (Example att1) (Example att2) = sqrt $ sum $ map (λ(x, y) → (x-y)↑2) $ zip (att1) (att2) ================================= All three reserves the word "distance" for itself and the second reserves also the word "attributes". How could I implement the module ML and which would be the best way to set "attributes" outside the module? Thanks Davi

2 4

Error in converting the function from let in clause to where
by mukesh tiwari 18 Jul '11

18 Jul '11

Hello all This is my first post on Haskell-beginner so i am not completely aware of protocols here . Pardon me for my ignorance . I am trying to implement elliptic curve prime factorisation and initially wrote function "addPoints" using let in clause. It works fine. When i converted the "addPoint" using where clause then its not working and not adding the points correctly . I run both functions from ghci and for same point " addPoints ( Conelliptic 1222 34 5678 ) ( Conpoint 123 42567 ) ( Conpoint 1234 4322222222222222) " addPoints function using where clause return Right 2 while let in clause return Left (Conpoint 4840 4122) . Could some one please help me to find out the bug . In case of indentation problem , source code on http://hpaste.org/49174 . Regards Mukesh Tiwari --y^2= x^3+ax+b mod n import Random import Control.Monad import Data.List import Data.Bits data Elliptic = Conelliptic Integer Integer Integer deriving ( Eq , Show ) data Point = Conpoint Integer Integer | Identity deriving ( Eq , Show ) powM :: Integer -> Integer -> Integer -> Integer powM a d n | d == 0 = 1 | d == 1 = mod a n | otherwise = mod q n where p = powM ( mod ( a^2 ) n ) ( shiftR d 1 ) n q = if (.&.) d 1 == 1 then mod ( a * p ) n else p calSd :: Integer -> IO ( Integer , Integer ) calSd n = return ( s , d ) where s = until ( \x -> testBit ( n - 1) ( fromIntegral x ) ) ( +1 ) 0 d = div ( n - 1 ) ( shiftL 1 ( fromIntegral s ) ) isProbable::Integer->IO Bool isProbable n | n <= 1 = return False | n == 2 = return True | even n = return False | otherwise = calSd n >>= (\( s , d ) -> rabinMiller 0 n s d ) rabinMiller::Integer->Integer->Integer->Integer->IO Bool rabinMiller cnt n s d | cnt>=5= return True | otherwise = randomRIO ( 2 , n - 2 ) >>= (\a -> case powM a d n == 1 of True -> rabinMiller ( cnt + 1 ) n s d _ -> if any ( == pred n ) . take ( fromIntegral s ) . iterate (\e -> mod ( e^2 ) n ) $ powM a d n then rabinMiller ( cnt + 1) n s d else return False ) {-- --add points of elliptic curve addPoints :: Elliptic -> Point -> Point -> Either Point Integer addPoints _ Identity p_2 = Left p_2 addPoints _ p_1 Identity = Left p_1 addPoints ( Conelliptic a b n ) ( Conpoint x_p y_p ) ( Conpoint x_q y_q ) | x_p /= x_q = case ( ( Conpoint x_r y_r ) , d ) of ( _ , 1 ) -> Left ( Conpoint x_r y_r ) ( _ , d' ) -> Right d' where [ u , v , d ] = extended_gcd ( x_p - x_q ) n s = mod ( ( y_p - y_q ) * u ) n x_r = mod ( s*s - x_p - x_q ) n y_r = mod ( -y_p - s * ( x_r - x_p ) ) n | otherwise = if mod ( y_p + y_q ) n == 0 then Left Identity else case ( ( Conpoint x_r y_r ) , d ) of ( _ , 1 ) -> Left ( Conpoint x_r y_r ) ( _ , d' ) -> Right d' where [ u , v , d ] = extended_gcd ( 2 * y_p ) n s = mod ( ( 3 * x_p * x_p + a ) * u ) n x_r = mod ( s * s - 2 * x_p ) n y_r = mod ( -y_p - s * ( x_r - x_p ) ) n --} --add points of elliptic curve addPoints::Elliptic->Point->Point-> Either Point Integer addPoints _ Identity p_2 = Left p_2 addPoints _ p_1 Identity = Left p_1 addPoints ( Conelliptic a b n ) ( Conpoint x_p y_p ) ( Conpoint x_q y_q ) | x_p /= x_q = let [ u , v , d ] = extended_gcd (x_p-x_q) n s = mod ( ( y_p - y_q ) * u ) n x_r = mod ( s * s - x_p - x_q ) n y_r= mod ( -y_p - s * ( x_r - x_p ) ) n in case ( ( Conpoint x_r y_r ) , d ) of ( _ , 1 ) -> Left ( Conpoint x_r y_r ) ( _ , d' ) -> Right d' | otherwise = if mod ( y_p + y_q ) n == 0 then Left Identity else let [ u , v , d ] = extended_gcd ( 2*y_p ) n s = mod ( ( 3 * x_p * x_p + a ) * u ) n x_r = mod ( s * s - 2 * x_p ) n y_r = mod ( -y_p - s * ( x_r - x_p ) ) n in case ( ( Conpoint x_r y_r ) , d ) of ( _ , 1 )-> Left (Conpoint x_r y_r) ( _ , d' ) -> Right d' extended_gcd::Integer->Integer->[Integer] extended_gcd u v= helpGcd u v 0 [ [ 1 , 0 ] , [ 0 , 1 ] ] where helpGcd u v n m @ ( [ [ a , b ] , [ c , d ] ] ) | v == 0 = if u<0 then [ - ( ( -1 ) ^ n ) * ( m !! 1 !! 1 ) , - ( ( -1 ) ^ ( n + 1 ) ) * ( m !! 0 !! 1 ) , -u ] else [ ( ( -1 ) ^ n ) * ( m !! 1 !! 1 ) , ( ( -1 ) ^ ( n + 1 ) ) * ( m !! 0 !! 1 ) , u ] | otherwise = helpGcd u' v' ( n + 1 ) m' where ( q , v' ) = quotRem u v t = [ [q , 1 ] , [ 1 , 0 ] ] m' = [ [ q * a + b , a ] , [ q * c + d , c ] ] --mult m t u' = v multiEC :: Elliptic -> Point -> Integer -> IO ( Either Point Integer ) multiEC _ _ 0 = return $ Left Identity multiEC ecurve point k | k>0 = return $ helpEC Identity point k where helpEC p _ 0 = Left p helpEC p q n = case (p',q') of ( Left p'' , Left q'' ) -> helpEC p'' q'' (div n 2) ( Right p'' , _ ) -> Right p'' ( _ , Right q'' ) -> Right q'' where p' =if odd n then addPoints ecurve p q else Left p q' = addPoints ecurve q q dscrmntEC a b = 4 * a * a * a + 27 * b * b randomCurve :: Integer -> IO [Integer] randomCurve n = randomRIO ( 1 , n ) >>= ( \a -> randomRIO ( 1 , n ) >>= ( \u -> randomRIO (1 , n) >>= (\v -> return [a , mod ( v*v - u*u*u - a*u ) n , n , u , v ] ) ) ) factor :: Integer -> Integer -> IO [Integer] factor 1 _ = return [] factor n m = isProbable n >>= (\x -> if x then return [n] else randomCurve n >>= (\[ a , b , n , u , v ] -> multiEC ( Conelliptic a b n ) ( Conpoint u v ) m >>= ( \p -> case p of Left p' -> factor n m Right p'-> factor ( div n p' ) m >>= ( \x -> factor p' m >>= (\y -> return $ x ++ y ) ) ) ) ) solve :: Integer -> IO [ Integer ] solve n = factor n ( foldl lcm 1 [1..10000] ) main = liftM read getLine >>=( \n -> solve n ) >>= print

2 1

FW: question
by Roelof Wobben 18 Jul '11

18 Jul '11

Oke, I tried with Ghci 7.03 on the prompt but then I see this: [x^2 | x <-- [1..100] not in scope x not in scope <-- When I put the text in a file I see these errors. parse error : naked expression at top level. Roelof > > > > ---------------------------------------- > > Date: Wed, 13 Jul 2011 23:04:58 +0200 > > From: andres.loeh(a)googlemail.com > > To: beginners(a)haskell.org > > Subject: Re: [Haskell-beginners] question > > > > I've just tried to evaluate 2 ^ 5 in codepad and it worked for me, so > > (^) seems to be supported. > > > > Cheers, > > Andres > > > > _______________________________________________ > > Beginners mailing list > > Beginners(a)haskell.org > > http://www.haskell.org/mailman/listinfo/beginners

8 12