- Beginners - Haskell.org

Stockholm Haskell User Group?
by Obscaenvs 74 18 Aug '11

18 Aug '11

I am posting this again although it might be the wrong forum to do so: I feel there is definitely a need for a Haskell User Group in Stockholm. If you are interested, contact me: obscaenvs(a)gmail.com . My name is Fredrik. /F

1 0

[m [a]] -> m [a]
by Guy 18 Aug '11

18 Aug '11

Is there a function which can do this, where m is not mplus? I would like all the lists to be joined inside the monad with ++.

4 5

How to find a "best alternative"
by Martin Drautzburg 17 Aug '11

17 Aug '11

Hello all, I want to write transformations on musical motives and I came upon the following algorithmic question: For every note of the original motiv I can compute a sequence of transformend notes which are ordered by their "fitness" with respect of a certain criterion. If might for example compute the "fitness with respect to harmonic function". I have a whole set of such criteria. So in the end the questions are: what should the first transformend note be? Harmonic function tells me it should be a G, but could also be a C or a D ..., some other function tells me it should be a C, but could also be a C# or a Cb. It kind of reminds me of the Levenshtein Disance, but this gives me a "distance" from two samples. I need a function which returns a new sample when one sample is given, i.e. the one with the smallest distance. Any idea in which direction is shall focus my thoughts? -- Martin

2 1

flatten
by Sunil S Nandihalli 16 Aug '11

16 Aug '11

Hello everybody, can somebody tell me as to what is wrong with the following code myflatten::[(a,a)]->[a] myflatten [] = [] myflatten (fx,sx):xs = fx:sx:(myflatten xs) It says that there is a parse error on the last line.. Thanks, Sunil.

2 3

the first argument of take is an Int and not Integral..
by Sunil S Nandihalli 16 Aug '11

16 Aug '11

Hi everybody, I feel forcing the first argument of take to be an Int is unnecessarily restrictive .. Is there a rationale behind not making it just (Integral a) ? right now it is take::Int->[a]->[a] I was thinking may be take::(Integral b)=>b->[a]->[a] would be better... Thanks, Sunil.

6 7

abstype?
by Christopher Howard 15 Aug '11

15 Aug '11

In my older (1992) intro programming book, the author uses the keyword "abstype" instead of "data" inside of module code. From what I can gather, abstype is supposed to hide constructors. But I've never seen this keyword used in any modern tutorials I've gone through. Is "abstype" still a valid keyword? Is is still relevant? -- frigidcode.com theologia.indicium.us

2 1

a simple little problem
by Dennis Raddle 14 Aug '11

14 Aug '11

Can someone suggest an elegant way to write the following? fn :: [Maybe Float] -> Maybe Float in which, if the input list has all Nothing, then the result is Nothing if the input list has one or more Just x, then the result is Just x (in which the x is picked arbitrarily, could be the first one or last one) I have something like import Data.Maybe fn list = case catMaybes list of [] -> Nothing [x:_] -> fromJust x Next, an augmentation of this idea (or similar idea). fn2 :: [Maybe Float] -> Map Int Float When a Just x appears at position n in the list, then put (key=n, value=x) into the map. I have: import qualified Data.Map as M f2 list = M.fromList [(n,x) | (n,Just x) <- zip [0..] list]

6 8

Invitation to connect on LinkedIn
by Andrew Smith B.Sc(Hons),MBA 14 Aug '11

14 Aug '11

LinkedIn ------------ I'd like to add you to my professional network on LinkedIn. - Andrew Andrew Smith B.Sc(Hons),MBA Founder and CEO at VTRL - Value Technology Research Ltd Edinburgh, United Kingdom Confirm that you know Andrew Smith B.Sc(Hons),MBA https://www.linkedin.com/e/-3dpk6t-grc8g5b9-3s/isd/3853823396/wkuHdNPh/ -- (c) 2011, LinkedIn Corporation

1 0

QuickCheck with monadic IO
by Hartmut 13 Aug '11

13 Aug '11

Everyone, I want to run an QuickCheck on "a most simple" IO monad. When I compile and run :main, it says: ....................................................................................................+++ OK, passed 100 tests. Well, thats fine, since *1 mapped on any Int gives the same Int back. But it lasts 10 seconds (!) for doing those 100 Tests. Something must be wrong?! I'd appreciate your ideas. Hartmut module MonadicQuickCheck where import IO import Random import Test.QuickCheck import Test.QuickCheck.Monadic main = do quickCheck x x :: Property x = monadicIO $ do a <- pick arbitrary pre $ not(null a) b <- run $ testfunction a assert $ b == a testfunction :: [Int] -> IO [Int] testfunction x = do putStr "." return $ map (*1) x

2 1

Relationship between graphs and queues?
by Christopher Howard 13 Aug '11

13 Aug '11

I working through the practice problems in Davie's book (by myself) and there is one that is not quite clear to me: "Construct an abstract data type for a queue. Attempt to supply an implementation (hidden from the user of the type) which allows access to the head of the queue and the construction of a new queue by adding elements to the end of an existing one. Both these interface functions should operate in time independent of the length of the queue. What implications does the this exercise suggest for search and 'updating' (by copying all unchanged nodes) a general graph?" There are two parts to the problem here: 1) defining the data type and the interface functions, and 2) indicating the implications for general graphs. On the first part I'm a little unclear because I am not sure what he means by "allows access to the head of the queue"; i.e., does he mean to just get the value in the head of the queue, or pop the value off the head of the queue and return the new queue? If the former, a solution is not hard, but if the latter, then I am not sure how to solve this without somehow transversing the queue, which of course is not a time-independent activity. (Even in the implementation for a queue in the Data.Graph.Inductive.Internal.Queue, eventually a "reverse" call must be made.) Anyway, maybe if someone could explain the relationship between this problem and graphs, I'll see the bigger picture and have a clearer view of the problem. -- frigidcode.com theologia.indicium.us

2 1