- Beginners - Haskell.org

Haskell, Ray Tracing, Parallel Computation
by Jon Harrop 16 Jan '10

16 Jan '10

Someone has tried to parallelize Lennart's Haskell from my ray tracer language comparison and they are claiming awesome performance results for Haskell, even beating HLVM: http://poorlytyped.blogspot.com/2010/01/haskell-ray-tracing-parallel.html Frankly, I don't believe their claims for a second. So I'd like to test their code for myself. However, apparently their code needs some third-party library that can be installed with: cabal install AC-Vector but, on my machine, that only seems to make the library accessible to GHC 6.10 and not to GHC 6.12 (which is needed to compile this benchmark). What is the easiest way for me to get this library working under GHC 6.12? -- Dr Jon Harrop, Flying Frog Consultancy Ltd. http://www.ffconsultancy.com/?e

4 5

using "read"
by Luca Ciciriello 16 Jan '10

16 Jan '10

Hi All. in the function: func :: String -> Int func str = read str An absolute beginner as i am, want to use "read" to convert a String in a Int, but I don't know how to protect my function func from an exception when str is not a number like "12A". How can I do? Thanks in advance for any answer to this my simply question. Luca.

4 3

ANN: Lite Haskell IDE
by Mambo Banda 15 Jan '10

15 Jan '10

Hi Just started an open source project called Hoodoo. It is meant to be a Lite IDE for Haskell. I wrote it to learn Haskell, so if you are a beginner (like me) it might interest you. If you're interested follow this link (http://hoodoo.kenai.com/) and if you like what you see, grab the code and add something:) Mambo

1 0

Design Question
by Jonathan Fischoff 14 Jan '10

14 Jan '10

I am writing graphics code, and there is a pattern that I am repeating. I imagine that there could be some way to automate the code that I am creating. I have a collection of type classes like the ones below (minus fundeps, and other elements to make it compile). class Vertex a ... class (Coordinate b, Vertex a) => Positionable a b where getPosition :: a -> b class (Vertex a) => Colorable a where getColor :: a -> Color class (Vertex a) => Textureable a where getUVs :: a -> (Double, Double) etc... Basically I have a bunch of type classes with one method getSomeType. The following would be an instance of all of the type classes: data ColorVertex = ColorVertex Cart3Coordinate Color (Double, Double) Additionally I have collection type classes as follows: class (Coordinate b) => PositionCollection a b where getPositionList :: a -> [b] class ColorList a where getColorList :: a -> [Color] class UVList a where getUVList :: a -> [(Double, Double)] So, for every getSomeType class, there is a getSomeTypeList class. An instance of these types would look like data StreamCollection3 = StreamCollection3 [Cart3Coordinate] [Color] [(Double, Double)] add it would also an instance of the following class, and I would like to figure out how to generate the functions for the class. class (Vertex b) => VertexCollection a b where getElement :: a -> Int -> b getVertexList :: a -> [b] fromVertexList :: [b] -> a The Vertex in this case would be ColorVertex. I was thinking I could reduce the code by making the classes class Gets a b where gets :: a -> b class GetsList a b where getsList :: a -> [b] But I think I could go further. I'm wondering if there is a way to completely automate the construction of these classes, data types and methods, such that all I would provide is definition of the element type such as data PositionVertex = PositionVertex Cart3Coordinate and everything else would get generated. I can see how I would write code to generate this. Is the time to learn Template Haskell, or is there another way to generate my boilerplate? -Jonathan Fischoff

1 0

Trivial function of type (a -> b) -> Maybe a -> Maybe b
by Eugen Zagorodniy 14 Jan '10

14 Jan '10

Greetings! I looked at documentation of Maybe type and suprisingly found no function like maybize :: (a -> b) -> Maybe a -> Maybe b maybize f Nothing = Nothing maybize f (Just x) = Just (f x) I believe there should be a good reason for absence of such function, say: 1) there is an idiomatic way (probably utilizing higher level abstractions) of doing the same what 'maybize' does, 2) no one really needs it... except me, 3) ?.. Did I miss something? (It is rather about curiosity than a real problem though). Thanks in advance for any comments!

4 4

Randomness woes
by Erlend Hamberg 12 Jan '10

12 Jan '10

Hello, everyone. I'm experimenting with making a simple genetic algorithm test in Haskell and have some questions about how to “propagate” randomness to the various functions and about how to do the final iteration. I have used MonadRandom which already have simplified the program quite a bit. The main functions in my program are: (Fitness is a Double, Genome is String of ones and zeros) ------------------------------------------------------------------------ -- calculates the fitness of the given genome fitness :: Genome -> Fitness -- takes a genome and a list of booleans and flips the bits according to the -- bool list mutate :: Genome -> [Bool] -> Genome -- takes two genomes and swap the genes after the given point crossOver :: (Genome, Genome) -> Int -> Genome -- takes two genoms and a mutation rate and returns a new genome reproduce :: (RandomGen g) => (Genome, Genome) -> Double -> Rand g (Genome) -- takes a list of genomes and their corresponding fitness and creates a new -- population by selecting two individuals by roulette wheel selection and -- create an offspring until the new population is the same size as the -- original makePopulation :: (RandomGen g) => [(Genome, Fitness)] -> Rand g [Genome] -- takes a list of genomes and creates the next generation by using makePop nextGeneration :: (RandomGen g) => [Genome] -> Rand g [Genome] -- returns a list of random genomes each with a given number of genes. used to -- create the initial population randomGenomes :: (RandomGen g) => Int -> Int -> Rand g [Genome] -- picks a random element from a list of (a, “score”). elements with a higher -- score are more likely to be picked rouletteSelect :: (RandomGen g) => [(a, Double)] -> Rand g a ------------------------------------------------------------------------ I have two concerns about my architecture. The first one is that I have several “layers” of randomness: The main function will call makePopulation which in turn will call rouletteSelect and reproduce. The latter will then call mutate, which also involves randomness. As can be seen above, I have rewritten mutate to take a list of booleans so that I can use getRandomRs in reproduce instead of having to have yet another layer of g <- getSplit let genome' = evalRand (mutate genome) g Handling the random generator explicitly like this feels a bit ugly. Or is this okay? Calling getSplit and then evalRand recursively is what I do in makePopulation, to make a new population. I would greatly appreciate comments on how this “should” be done. I realise that I could just get a list of random numbers in makePopulation and then use this as a source of randomness for the called functions, but I fear that this would just lead to inelegant code managing (chunks of) this list. My second stumbling point is how to produce a list of successive generations. I made the following test function to try to achieve this: generations :: (RandomGen g) => [Genome] -> g -> [[Genome]] generations pop g = pop : generations (evalRand (nextGeneration pop) g1) g2 where (g1,g2) = split g However, it doesn't feel right at all to do it like this. It also seems that the fitness values of all generations except the first one are identical, which is a good indication that that approach won't work. :-) The full program, as it is now, can be found here: http://hpaste.org/fastcgi/hpaste.fcgi/view?id=15782 The problem the algorithm is trying to solve can be found here: http://www.ai-junkie.com/ga/intro/gat3.html I'm grateful for all comments and suggestions, but mainly the two aforementioned points. -- Erlend Hamberg “Everything will be ok in the end. If its not ok, its not the end.” GPG/PGP: 0xAD3BCF19 45C3 E2E7 86CA ADB7 8DAD 51E7 3A1A F085 AD3B CF19

3 3

prefix instead of infix
by John Moore 11 Jan '10

11 Jan '10

Hi, Can anyone explain how to turn the code from infix to Prefix. I want to include a let statement below. I not sure of how prefix works. expr :: Parser Expr expr = buildExpressionParser table factor <?> "expression" table :: [[Operator Char st Expr]] table = [[op "*" Mul AssocLeft, op "/" Div AssocLeft] ,[op "+" Add AssocLeft, op "-" Sub AssocLeft] ] where op s f assoc = Infix (do{ string s; return f}) assoc factor :: Parser Expr factor = do{ char '(' ; x <- expr ; char ')' ; return x } <|> number <?> "simple expression" number :: Parser Expr number = do{ ds <- many1 alphaNum ; return (Val $ read ds) } John

3 2

Re: [Haskell-beginners] prefix instead of infix
by Daniel Fischer 11 Jan '10

11 Jan '10

Am Sunday 10 January 2010 21:24:40 schrieben Sie: > Hi Daniel, > I want to add a let expression to the program so as I can > enter let x = y in z or something to that effect. I was going along the > lines of using prefix as I cant use infix so I thought I would make it > easier by using prefix notation. Probably a bad idea. > > John If you want only one binding per let, I think letExpr = do keyword "let" -- implementation of keyword is left as an exercise bind <- variable -- binding patterns like (x,y) is left for later -- maybe check for "=" ? rhs <- expression keyword "in" use <- expression return (Let bind rhs use) expression = letExpr <|> expr factor = between (char '(') (char ')') expression <|> atom atom = variable <|> number would work (untested). > > On Sun, Jan 10, 2010 at 8:14 PM, Daniel Fischer <daniel.is.fischer(a)web.de>wrote: > > Am Sonntag 10 Januar 2010 20:33:40 schrieb John Moore: > > > Hi, > > > Can anyone explain how to turn the code from infix to Prefix. I > > > want to include a let statement below. I not sure of how prefix > > > works. > > > > Can you elaborate? I don't understand what it is you want to do. > > Do you want to parse expressions in prefix notation (aka Polish > > notation), such as +(3,4) (or, more Lispy, (+ 3 4))? > > That'd be simple, but you wouldn't use buldExpressionParser for that > > (in Polish notation, neither associativity nor precedence have a > > meaning). Or do you want to add a parser for let-expressions to expr? > > > > > expr :: Parser Expr > > > expr = buildExpressionParser table factor > > > <?> "expression" > > > table :: [[Operator Char st Expr]] > > > table = [[op "*" Mul AssocLeft, op "/" Div AssocLeft] > > > ,[op "+" Add AssocLeft, op "-" Sub AssocLeft] > > > ] > > > where > > > op s f assoc > > > = Infix (do{ string s; return f}) assoc > > > factor :: Parser Expr > > > factor = do{ char '(' > > > ; x <- expr > > > ; char ')' > > > ; return x > > > } > > > <|> number > > > <?> "simple expression" > > > number :: Parser Expr > > > number = do{ ds <- many1 alphaNum > > > ; return (Val $ read ds) > > > } > > > > > > John

1 0

variable parser
by John Moore 10 Jan '10

10 Jan '10

Hi, I just want to add a new parser for variables but this doesn't seem to work can anybody tell me what I'm doing wrong? module ExprSyn where import Text.ParserCombinators.Parsec import Text.ParserCombinators.Parsec.Expr runExpr :: String -> IO () runExpr str = case runParser expr () " " str of Left err -> putStrLn "Error:" >> print err Right val -> print val demo1 = runExpr "2+5" data Expr = Mul Expr Expr | Div Expr Expr | Add Expr Expr | Sub Expr Expr | Val Float |Let String Expr Expr | Var String deriving (Eq,Show) expr :: Parser Expr expr = variable <|> buildExpressionParser table factor <?> "expression" table :: [[Operator Char st Expr]] table = [[op "*" Mul AssocLeft, op "/" Div AssocLeft] ,[op "+" Add AssocLeft, op "-" Sub AssocLeft] ] where op s f assoc = Infix (do{ string s; return f}) assoc factor :: Parser Expr factor = do{ char '(' ; x <- expr ; char ')' ; return x } <|> number <|> variable <?> "simple expression" variable :: Parser Expr variable = do { b <- letter ; do { cs <- word ;return (c:cs) <|> return [c] <|> number } number :: Parser Expr number = do{ ds <- many1 digit ; return (Val $ read ds) } John

2 1

Help! I can't compile an example program
by Tim Perry 09 Jan '10

09 Jan '10

Hi all, I found an example program that computes grades here: http://donsbot.wordpress.com/?s=grades I copied it to my computer and tried to compile it and if failed. Strangely enough, it works on the command line using 'runhaskell'. This makes me think I didn't set up the libraries correctly. So I downloaded and re-installed the Haskell Platform from: http://hackage.haskell.org/platform/2009.2.0.2/HaskellPlatform-2009.2.0.2-s… The problem remains. I included the errors, the source and the results of runhaskell below. I hope someone can help me sort out what went wrong. Thanks, Tim C:\Documents and Settings\perry\hs>ghc -O Grades.hs Grades.o:fake:(.text+0x2f7): undefined reference to `containerszm0zi2zi0zi1_DataziMap_Tip_closure' Grades.o:fake:(.text+0x930): undefined reference to `containerszm0zi2zi0zi1_DataziMap_lvl47_closure' Grades.o:fake:(.text+0xa8b): undefined reference to `__stginit_containerszm0zi2zi0zi1_DataziMap_' Grades.o:fake:(.text+0x219): undefined reference to `containerszm0zi2zi0zi1_DataziMap_insertWithKey_info' Grades.o:fake:(.text+0x504): undefined reference to `containerszm0zi2zi0zi1_DataziMap_lookup_info' Grades.o:fake:(.text+0x90e): undefined reference to `containerszm0zi2zi0zi1_DataziMap_polyzugo1_info' Grades.o:fake:(.text+0x93f): undefined reference to `containerszm0zi2zi0zi1_DataziMap_foldr_info' Grades.o:fake:(.data+0x50): undefined reference to `containerszm0zi2zi0zi1_DataziMap_insertWithKey_closure' collect2: ld returned 1 exit status C:\Documents and Settings\perry\hs>cat Grades.hs import Data.Char import Data.Maybe import Data.List import Data.Map hiding (map) import Text.Printf main = do src <- readFile "grades.txt" let pairs = map (split.words) (lines src) grades = foldr insert empty pairs mapM_ (draw grades) (sort (keys grades)) where insert (s, g) = insertWith (++) s [g] split [name,mark] = (name, read mark) draw g s = printf "%s\t%s\tAverage: %f\n" s (show marks) avg where marks = findWithDefault (error "No such student") s g avg = sum marks / fromIntegral (length marks) :: Double C:\Documents and Settings\perry\hs>runhaskell Grades.hs Alonzo [70.0,90.0,95.0] Average: 85.0 Bob [80.0] Average: 80.0 Eugenio [69.0] Average: 69.0 Henk [79.0] Average: 79.0 Oleg [77.0] Average: 77.0 Philip [73.0] Average: 73.0 Simon [94.0,99.0,93.0] Average: 95.33333333333333 C:\Documents and Settings\perry\hs>

2 1