April 2009 - Haskell-Cafe

Announce: A pragmatic Haskell .NET interop layer, 0.4.0
by Sigbjorn Finne 19 Apr '09

19 Apr '09

A new version of a Haskell .NET interop layer, hs-dotnet, has just been released and is now available for download, http://haskell.forkIO.com/dotnet It lets you access .NET functionality from Haskell and vice versa. Tool support is included in this release to aid such interop. The new version includes development done since the start of the year. Apart from rewriting the internals completely to put it all on a sounder footing, this release includes proper support for .NET generic types (classes and interfaces), mapping them naturally on to Haskell parameterized types. The support for generics enables for instance mixed Haskell-.NET LINQ programming; see the distribution for examples of this along with some other interesting applications of the hs-dotnet interop layer. enjoy --sigbjorn

2 1

Using GHC as an arrows preprocessor?
by Peter Verswyvelen 19 Apr '09

19 Apr '09

I'm making some custom arrows and I'm getting bugs when using the GHC preprocessor, but not when using the old standalone preprocessor. In order to debug this, it would be nice to use GHC as an arrows preprocessor (so converting the arrow notation into plain Haskell and outputting the converted source code); is this possible? I also noticed that GHC's builtin preprocessor generates far more intermediate combinators than the old preprocessor. The benchmark in this paper <http://www.cs.yale.edu/homes/hl293/download/NEPLS-talk.pdf> seems to confirm this. Is this a known issue? Thanks, Peter

2 2

Matrices
by Cetin Sert 19 Apr '09

19 Apr '09

http://ccnmtl.columbia.edu/projects/qmss/the_chisquare_test/about_the_chisq… given two matrices, Prelude Data.Matrix.Dense Data.Vector.Dense> m listMatrix (2,2) [46.0,37.0,71.0,83.0] Prelude Data.Matrix.Dense Data.Vector.Dense> es listMatrix (2,2) [40.9746835443038,42.0253164556962,76.0253164556962,77.9746835443038] how can I flatten them to do: [ (o-e)^2 / e | o <- m, e <- es ] or use a function that will apply a given function to every corresponding elements of 2 or n matrices and create a result matrix? If I should use a different matrix type or library altogether, please specify which one and how ^__^" Best Regards, Cetin Sert

3 2

Zippers for gui elements?
by Gü?nther Schmidt 18 Apr '09

18 Apr '09

Hi all, I need to control some gui elements, ie. stuff like when dis button is clicked make dat one inactive. Some time ago, wish I'd remember where, I read about using Zippers for that sort of thing. Does anybody know where I might have read that? Günther

1 0

Re: [Haskell-cafe] General function to count list elements?
by nowgate＠yahoo.com 18 Apr '09

18 Apr '09

Hi John, Yes, I see now what the hullabaloo was about. I've only been hacking Haskell for a few weeks, so I'm still seeing the "forest" rather than the "trees". But I'm coming along. Thanks. Michael --- On Sat, 4/18/09, John Dorsey <haskell(a)colquitt.org> wrote: From: John Dorsey <haskell(a)colquitt.org> Subject: Re: [Haskell-cafe] General function to count list elements? To: "michael rice" <nowgate(a)yahoo.com> Cc: haskell-cafe(a)haskell.org Date: Saturday, April 18, 2009, 2:36 PM Michael, > I had a count function that worked fine for an enum type, but thought > why not go polymorphic with it. So I changed it and got the error I > originally inquired about. For variety, I'll go a slightly different direction. If you generalize count to use any predicate, instead of always equality... gcount :: (a -> a -> Bool) -> a -> [a] -> Int gcount pred x0 xs = length (filter (pred x0) xs) count = gcount (==) This will work with any type that you can write a predicate for with the type (a -> a -> Bool). I can even use this with functions, if I'm careful. ghci> gcount (\f g -> True) (*2) [id,(const 1),(*3)] 3 ghci> gcount (\f g -> f 1 == g 1) (^2) [id,(const 1),(*3)] 2 By the way, do you see why everyone's bothing you about comparing functions? The type you gave count, which didn't have an Eq constraint, was an assertion that you could compare two values of *any* type. If there's a type that's not comparable, then count's type was wrong. Functions are the canonical example of an incomparable type. When you're bored some time, read a bit about the Curry-Howard correspondence. It's interesting, even if (like me) you don't grok all of its implications. Regards, John

1 0

Re: [Haskell-cafe] General function to count list elements?
by michael rice 18 Apr '09

18 Apr '09

I wasn't comparing functions, you were. I was just trying to answer your questions. I had a count function that worked fine for an enum type, but thought why not go polymorphic with it. So I changed it and got the error I originally inquired about. Thanks. Michael --- On Sat, 4/18/09, Eugene Kirpichov <ekirpichov(a)gmail.com> wrote: From: Eugene Kirpichov <ekirpichov(a)gmail.com> Subject: Re: [Haskell-cafe] General function to count list elements? To: "michael rice" <nowgate(a)yahoo.com> Cc: haskell-cafe(a)haskell.org Date: Saturday, April 18, 2009, 1:18 PM 2009/4/18 michael rice <nowgate(a)yahoo.com>: >> To compare two functions in C, I would compare their machine addresses. >> > > Why would you need that at all? > > How would *you* do it? Do what? As for comparing functions, I would not do it at all, because comparing functions makes no sense: one way of comparison (extensional equality) is not implementable, whereas the other (object identity) breaks referential transparency and is impossible to reason about and generally doesn't have a single reason to work correctly at all. Object identity has no reason to exist in a language without mutable objects, like Haskell. That's probably the key. As for implementing the count function, I would implement it in the way that people already suggested to you: by putting a constraint onto its type, namely a constraint that makes it polymorphic *over types whose values can be compared for equality*, t.i. over members of the Eq class. This is perfectly correct and does not cause any mathematical unsoundness. > > Michael > > ================ > > --- On Sat, 4/18/09, Eugene Kirpichov <ekirpichov(a)gmail.com> wrote: > > From: Eugene Kirpichov <ekirpichov(a)gmail.com> > Subject: Re: [Haskell-cafe] General function to count list elements? > To: "michael rice" <nowgate(a)yahoo.com> > Cc: haskell-cafe(a)haskell.org > Date: Saturday, April 18, 2009, 12:39 PM > > 2009/4/18 michael rice <nowgate(a)yahoo.com>: >> I know functions can be compared in Scheme >> >> Welcome to DrScheme, version 4.1 [3m]. >> Language: Swindle; memory limit: 128 megabytes. >>> (equal? equal? equal?) >> #t >>> >> > > That's not the functions being compared, but the memory addresses of > the code implementing them. If your goal is comparing functions to > answer a question "Are these two values indistinguishable?", equal? > doesn't help you, because it may answer 'false' even if these two > values are indistinguishable from a mathematical point of view. > >> >> but apparently not in Haskell >> >> [michael@localhost ~]$ ghci >> GHCi, version 6.10.1: http://www.haskell.org/ghc/ :? for help >> Loading package ghc-prim ... linking ... done. >> Loading package integer ... linking ... done. >> Loading package base ... linking ... done. >> Prelude> (==) (==) (==) >> >> <interactive>:1:0: >> No instance for (Eq (a -> a -> Bool)) >> arising from a use of `==' at <interactive>:1:0-13 >> Possible fix: add an instance declaration for (Eq (a -> a -> Bool)) >> In the expression: (==) (==) (==) >> In the definition of `it': it = (==) (==) (==) >> Prelude> >> >> though I'm new at Haskell and may not be posing the question properly. >> >> I would think a language with 1st-class support for functions would >> certainly include comparing them. >> > > Again, this is first-class support for memory addresses of code > representing functions. > >> To compare two functions in C, I would compare their machine addresses. >> > > Why would you need that at all? > >> Michael >> >> >> --- On Sat, 4/18/09, Eugene Kirpichov <ekirpichov(a)gmail.com> wrote: >> >> From: Eugene Kirpichov <ekirpichov(a)gmail.com> >> Subject: Re: [Haskell-cafe] General function to count list elements? >> To: "michael rice" <nowgate(a)yahoo.com> >> Cc: haskell-cafe(a)haskell.org >> Date: Saturday, April 18, 2009, 11:39 AM >> >> Could you then provide an example of two functions that *are* equal, >> or, even better, a definition of equality for arbitrary functions? >> Since Haskell may be compiled into C, this must be a definition that >> is implementable in C. >> >> 2009/4/18 michael rice <nowgate(a)yahoo.com>: >>> Though I haven't tried it out, it's trying to use my function to count >>> functions. >>> >>> The first argument is the identity function. >>> >>> The second argument is a list of a different form of the identity >>> function. >>> >>> Though the two identity functions, given the same input, would produce >>> the >>> same output, I doubt they would be equal. >>> >>> So my guess at an answer would be zero. >>> >>> Michael >>> >>> --- On Sat, 4/18/09, Eugene Kirpichov <ekirpichov(a)gmail.com> wrote: >>> >>> From: Eugene Kirpichov <ekirpichov(a)gmail.com> >>> Subject: Re: [Haskell-cafe] General function to count list elements? >>> To: "michael rice" <nowgate(a)yahoo.com> >>> Cc: haskell-cafe(a)haskell.org >>> Date: Saturday, April 18, 2009, 11:03 AM >>> >>> What should >>> >>> count (\x -> x) (replicate 10 (\y -> if 1==1 then y else undefined)) >>> >>> return? >>> >>> 2009/4/18 michael rice <nowgate(a)yahoo.com>: >>>> Is there a general function to count list elements. I'm trying this >>>> >>>> count :: a -> [a] -> Int >>>> count x ys = length (filter (== x) ys) >>>> >>>> with this error upon loading >>>> >>>> ============= >>>> >>>> [michael@localhost ~]$ ghci count >>>> GHCi, version 6.10.1: http://www.haskell.org/ghc/ :? for help >>>> Loading package ghc-prim ... linking ... done. >>>> Loading package integer ... linking ... done. >>>> Loading package base ... linking ... done. >>>> [1 of 1] Compiling Main ( count.hs, interpreted ) >>>> >>>> count.hs:2:29: >>>> Could not deduce (Eq a) from the context () >>>> arising from a use of `==' at count.hs:2:29-32 >>>> Possible fix: >>>> add (Eq a) to the context of the type signature for `count' >>>> In the first argument of `filter', namely `(== x)' >>>> In the first argument of `length', namely `(filter (== x) ys)' >>>> In the expression: length (filter (== x) ys) >>>> Failed, modules loaded: none. >>>> Prelude> >>>> >>>> ============= >>>> >>>> Not sure what it's trying to tell me other than I need an (Eq a) >>>> somewhere. >>>> >>>> Michael >>>> >>>> >>>> >>>> >>>> _______________________________________________ >>>> Haskell-Cafe mailing list >>>> Haskell-Cafe(a)haskell.org >>>> http://www.haskell.org/mailman/listinfo/haskell-cafe >>>> >>>> >>> >>> >>> >>> -- >>> Eugene Kirpichov >>> Web IR developer, market.yandex.ru >>> >>> >> >> >> >> -- >> Eugene Kirpichov >> Web IR developer, market.yandex.ru >> >> > > > > -- > Eugene Kirpichov > Web IR developer, market.yandex.ru > > -- Eugene Kirpichov Web IR developer, market.yandex.ru

2 1

Re: [Haskell-cafe] General function to count list elements?
by michael rice 18 Apr '09

18 Apr '09

> To compare two functions in C, I would compare their machine addresses. > Why would you need that at all? How would *you* do it? Michael ================ --- On Sat, 4/18/09, Eugene Kirpichov <ekirpichov(a)gmail.com> wrote: From: Eugene Kirpichov <ekirpichov(a)gmail.com> Subject: Re: [Haskell-cafe] General function to count list elements? To: "michael rice" <nowgate(a)yahoo.com> Cc: haskell-cafe(a)haskell.org Date: Saturday, April 18, 2009, 12:39 PM 2009/4/18 michael rice <nowgate(a)yahoo.com>: > I know functions can be compared in Scheme > > Welcome to DrScheme, version 4.1 [3m]. > Language: Swindle; memory limit: 128 megabytes. >> (equal? equal? equal?) > #t >> > That's not the functions being compared, but the memory addresses of the code implementing them. If your goal is comparing functions to answer a question "Are these two values indistinguishable?", equal? doesn't help you, because it may answer 'false' even if these two values are indistinguishable from a mathematical point of view. > > but apparently not in Haskell > > [michael@localhost ~]$ ghci > GHCi, version 6.10.1: http://www.haskell.org/ghc/ :? for help > Loading package ghc-prim ... linking ... done. > Loading package integer ... linking ... done. > Loading package base ... linking ... done. > Prelude> (==) (==) (==) > > <interactive>:1:0: > No instance for (Eq (a -> a -> Bool)) > arising from a use of `==' at <interactive>:1:0-13 > Possible fix: add an instance declaration for (Eq (a -> a -> Bool)) > In the expression: (==) (==) (==) > In the definition of `it': it = (==) (==) (==) > Prelude> > > though I'm new at Haskell and may not be posing the question properly. > > I would think a language with 1st-class support for functions would > certainly include comparing them. > Again, this is first-class support for memory addresses of code representing functions. > To compare two functions in C, I would compare their machine addresses. > Why would you need that at all? > Michael > > > --- On Sat, 4/18/09, Eugene Kirpichov <ekirpichov(a)gmail.com> wrote: > > From: Eugene Kirpichov <ekirpichov(a)gmail.com> > Subject: Re: [Haskell-cafe] General function to count list elements? > To: "michael rice" <nowgate(a)yahoo.com> > Cc: haskell-cafe(a)haskell.org > Date: Saturday, April 18, 2009, 11:39 AM > > Could you then provide an example of two functions that *are* equal, > or, even better, a definition of equality for arbitrary functions? > Since Haskell may be compiled into C, this must be a definition that > is implementable in C. > > 2009/4/18 michael rice <nowgate(a)yahoo.com>: >> Though I haven't tried it out, it's trying to use my function to count >> functions. >> >> The first argument is the identity function. >> >> The second argument is a list of a different form of the identity >> function. >> >> Though the two identity functions, given the same input, would produce the >> same output, I doubt they would be equal. >> >> So my guess at an answer would be zero. >> >> Michael >> >> --- On Sat, 4/18/09, Eugene Kirpichov <ekirpichov(a)gmail.com> wrote: >> >> From: Eugene Kirpichov <ekirpichov(a)gmail.com> >> Subject: Re: [Haskell-cafe] General function to count list elements? >> To: "michael rice" <nowgate(a)yahoo.com> >> Cc: haskell-cafe(a)haskell.org >> Date: Saturday, April 18, 2009, 11:03 AM >> >> What should >> >> count (\x -> x) (replicate 10 (\y -> if 1==1 then y else undefined)) >> >> return? >> >> 2009/4/18 michael rice <nowgate(a)yahoo.com>: >>> Is there a general function to count list elements. I'm trying this >>> >>> count :: a -> [a] -> Int >>> count x ys = length (filter (== x) ys) >>> >>> with this error upon loading >>> >>> ============= >>> >>> [michael@localhost ~]$ ghci count >>> GHCi, version 6.10.1: http://www.haskell.org/ghc/ :? for help >>> Loading package ghc-prim ... linking ... done. >>> Loading package integer ... linking ... done. >>> Loading package base ... linking ... done. >>> [1 of 1] Compiling Main ( count.hs, interpreted ) >>> >>> count.hs:2:29: >>> Could not deduce (Eq a) from the context () >>> arising from a use of `==' at count.hs:2:29-32 >>> Possible fix: >>> add (Eq a) to the context of the type signature for `count' >>> In the first argument of `filter', namely `(== x)' >>> In the first argument of `length', namely `(filter (== x) ys)' >>> In the expression: length (filter (== x) ys) >>> Failed, modules loaded: none. >>> Prelude> >>> >>> ============= >>> >>> Not sure what it's trying to tell me other than I need an (Eq a) >>> somewhere. >>> >>> Michael >>> >>> >>> >>> >>> _______________________________________________ >>> Haskell-Cafe mailing list >>> Haskell-Cafe(a)haskell.org >>> http://www.haskell.org/mailman/listinfo/haskell-cafe >>> >>> >> >> >> >> -- >> Eugene Kirpichov >> Web IR developer, market.yandex.ru >> >> > > > > -- > Eugene Kirpichov > Web IR developer, market.yandex.ru > > -- Eugene Kirpichov Web IR developer, market.yandex.ru

3 2

General function to count list elements?
by michael rice 18 Apr '09

18 Apr '09

Is there a general function to count list elements. I'm trying this count :: a -> [a] -> Int count x ys = length (filter (== x) ys) with this error upon loading ============= [michael@localhost ~]$ ghci count GHCi, version 6.10.1: http://www.haskell.org/ghc/ :? for help Loading package ghc-prim ... linking ... done. Loading package integer ... linking ... done. Loading package base ... linking ... done. [1 of 1] Compiling Main ( count.hs, interpreted ) count.hs:2:29: Could not deduce (Eq a) from the context () arising from a use of `==' at count.hs:2:29-32 Possible fix: add (Eq a) to the context of the type signature for `count' In the first argument of `filter', namely `(== x)' In the first argument of `length', namely `(filter (== x) ys)' In the expression: length (filter (== x) ys) Failed, modules loaded: none. Prelude> ============= Not sure what it's trying to tell me other than I need an (Eq a) somewhere. Michael

4 3

Computing a sorted list of products lazily
by Jason Dagit 18 Apr '09

18 Apr '09

Hello, A colleague of mine recently asked if I knew of a lazy way to solve the following problem: Given two sets of sorted floating point numbers, can we lazily generate a sorted list of the products from their Cartesian product? The algorithm should return the same result as: sortProduct a b = sort [ x * y | x <- a, y <- b ] But, my friend is not satisfied with the strictness of sort. In particular, he doesn't want to have to hold the whole list in memory to sort it. He would like to compute the results one product at a time and in the correct order. Is there an existing algorithm for this problem? We searched but we don't seem to know the right search terms. Additionally, my friend is not writing this in Haskell, but a Haskell solution is fine because we should be able to transform it into a Java solution by hand. We think we know an algorithm that will produce the right result but it's a bit messy and we doubt we're the first to tackle this problem. Any help would be greatly appreciated. Thanks in advance! Jason

7 7

Synchronising cabal package version with self-reported version
by Denis Bueno 18 Apr '09

18 Apr '09

Hi all, In a command-line app of mine I want to have a --version flag, like many GNU apps do to report their version. The only way I know to provide the version number is to hardcode it in the source code somewhere. That means I have the version number in two places: the .cabal file and the source code. I just know that one day they'll get unsynchronised, and --version will be wrong. Is there any way to put the version number in _one_ place, and have both the .cabal and source refer to it? Or even any easier way to synchronise both version numbers? Thanks, Denis

2 2