January 2005 - Haskell-Cafe

fastest Fibonacci numbers in the West
by William Lee Irwin III 28 Jan '05

28 Jan '05

Inspired by a discussion on freenode #haskell, I tried to write the fastest Fibonacci number function possible, i.e. given a natural number input n to compute F_n. mlton seems to enjoy substantially better speed despite equivalent algorithms; it may be enlightening to determine the causes of this, at least for those concerned about performance and the inner workings of the Haskell runtime system(s). In general, I am not usually very concerned about performance, nor am I in this case. But it's something of a mindless microbenchmark or similar. The things I'd normally think of to speed this up would be getting some primitives to find the highest bit of an integer (floor . lg) and to test a given bit of an integer (something vaguely like fromEnum . fromIntegral . (`mod` 2) . (/2^k)), both in some lightweight O(1) with low-hidden-constant manner. This doesn't appear to be a factor in the testing I did, as they're largely 1:1 translations of each other. Still, such things would be useful in various other algorithms, of far greater importance than this one. For the moment, mlton-generated binaries crash computing fib (10^8-1), and there is a 6:1 speed difference for fib (10^7-1) between the two, where mlton-generated binaries take just under 1 minute, and ghc- generated binaries take just under 6 minutes. Anyway, thoughts on how to improve all this from the programmer's point of view, or otherwise explaining what's going on or ameliorating whatever effect is at work here would be appreciated. -- wli

2 2

wierd type errros with MPTCs
by S. Alexander Jacobson 28 Jan '05

28 Jan '05

This code gives me a Kind error because IVal isn't applied to enough type arguments. import qualified Set class Table table indexVal | indexVal->table where --insertIndex::item->indexVal item -> table item ->table item union::table item -> table item -> table item --union t1 t2 = t1 data DBTable item = DBTable data IVal item = Name item instance Table DBTable (IVal ) where Weirdly, when I uncomment the insertIndex function, things work. But, if I then uncomment the default implementation of union, I get: No instance for (Table DBTable indexVal) arising from use of `Main.$dmunion' at example.hs:13 In the definition of `union': union = Main.$dmunion In the definition for method `union' In the instance declaration for `Table DBTable IVal' I don't know what this error even means. But it goes away if I put the union implementation in the instance rather than in the class. Bot these error messages seem unreasonable. Can someone clarify? Note: I am using GHC 6.2.2 -Alex- ______________________________________________________________ S. Alexander Jacobson tel:917-770-6565 http://alexjacobson.com

3 2

Question on "case x of g" when g is a function
by yeoh＠cs.wisc.edu 28 Jan '05

28 Jan '05

Can a kind soul please enlighten me on why f bit0 and f bit1 both return 0? > bit0 = False > bit1 = True > f x = case x of > bit0 -> 0 > bit1 -> 1

4 3

Converting from Int to Double
by Dmitri Pissarenko 28 Jan '05

28 Jan '05

Hello! I have a list of integer numbers (grayscale values from 0 to 255) and want to convert them to a list of double numbers, so that each number is 0 <= x <= 1, where 0 is completely black and 1 is completely white. Before I convert the numbers, I need to convert them to a list of double values (then I can use map to re-scale each element from 0..255 to 0..1). How can I convert an Int into a Double? TIA Dmitri Pissarenko -- Dmitri Pissarenko Software Engineer http://dapissarenko.com

5 8

RE: [Haskell-cafe] File path programme
by Simon Marlow 28 Jan '05

28 Jan '05

On 25 January 2005 19:45, Duncan Coutts wrote: > On Tue, 2005-01-25 at 19:12 +0000, Ben Rudiak-Gould wrote: >> My concern here is that someone will actually use the library once it >> ships, with the following consequences: >> >> 1. Programs using the library will have predictable >> (exploitable) bugs in pathname handling. >> >> 2. It will never be possible to change the current weird >> behavior, because it might break legacy code. The System.FilePath >> library will >> have to remain in GHC forever in its current form, enticing >> programmers with its promise of easy pathname handling and then >> cruelly breaking its contract. >> >> If no one uses it in production code then we can fix it at our >> leisure, and having it out there with "experimental" status isn't >> necessarily a bad thing in that case. It just feels like we're >> playing a dangerous game. > > That's a sufficiently persuasive argument for me! > > Could we just punt this library for this release. After all we can add > libraries in a later point release (eg 6.4.1) you just can't change > existing APIs. We can't add libraries in a point release, because there's no way for code to use conditional compilation to test the patchlevel version number. This seems to be a common misconception, probably brought about by the fact that the time between major releases of GHC is getting quite long. Perhaps I should stop writing email and get some work done :) Cheers, Simon

4 3

RE: [Haskell-cafe] Problem with Hyperlinking Haddock-documentation
by Simon Marlow 28 Jan '05

28 Jan '05

On 27 January 2005 16:54, daniel.is.fischer(a)web.de wrote: > BTW, would it be worthwhile to let haddock use a list of default > interfaces, say on installation it would create a file with > -i /usr/...../base.haddock > -i /usr/...../parsec.haddock &cetera > in it, later a user might add other often used interfaces ? > > Okay, writing one's own script isn't too hard, so there's no real > need for it, but it would be nice for dummies like me. I'd like Haddock to know about installed packages, and in fact we're part of the way there in GHC 6.4 because the package database now includes the locations of the .haddock file and the HTML for each package. I just have to modify Haddock to query the package database... Cheers, Simon

1 0

RE: [Haskell-cafe] Problem with Hyperlinking Haddock-documentation
by Simon Marlow 28 Jan '05

28 Jan '05

On 27 January 2005 09:03, Daniel Fischer wrote: > Am Mittwoch, 26. Januar 2005 21:49 schrieben Sie: >> On Wed, 26 Jan 2005, Daniel Fischer wrote: >>> Maybe somebody can enlighten me. >>> >>> When I run haddock and put the html files e.g. in directory >>> ~/bar/foo, any references to things defined in the Prelude or the >>> libraries are linked to, say ~/bar/foo/Prelude.html#t%3AFractional, >>> which of course does not exist, because the documentation for the >>> Prelude is elsewhere. >>> >>> How can I tell haddock to link such references to where Prelude.html >>> actually is? >> >> I add an option like >> >> -i /usr/local/share/ghc-6.2/html/libraries/base/base.haddock >> > The problem still stands, the link is to ~/bar/foo/Prelude.html, > not to /usr/local/.... > where Prelude.html is. > > Does anybody know what to do? You probably want to invoke Haddock like this: haddock -i/usr/share/ghc-6.2.2/html/libraries/base,/usr/share/ghc-6.2.2/librarie s/base/base.haddock (all on one line). The first path specifies the directory of the HTML for the base package, the second path specifies the .haddock file for the base package. Cheers, Simon

2 1

RE: [Haskell-cafe] Re: library documentation
by Simon Marlow 27 Jan '05

27 Jan '05

On 27 January 2005 10:46, Peter Simons wrote: > Isaac Jones writes: > > > http://www.haskell.org/hawiki/LibraryDocsNeedingHelp > > This is a great idea. > > I have been thinking .... you know what would make > contribution to the library efforts even simpler? If they > were available in a Darcs repository. > > Saying "darcs push" after you've spontaneously added a > three-line explanation into a standard library is a lot more > attractive than "cvs diff", cutting and pasting that diff > into an editor, and e-mailing it to the -libraries list. > (Who submits a diff for three lines of documentation?) > > One of the reasons why Darcs is great is that it makes > submitting changes very simple for _everybody_. This would be great, but unfortunately the activation energy is quite high: we're already pretty committed to CVS, so we either move over to darcs wholesale, or we keep GHC in CVS and libraries in darcs. Neither is particularly attractive right now - darcs isn't ready for projects the size of GHC (at least, last I looked), and having a split repository would be a nightmare. Cheers, Simon

2 2

RE: [Haskell-cafe] Convert to unboxed value? How?
by Simon Marlow 27 Jan '05

27 Jan '05

On 27 January 2005 12:36, Dimitry Golubovsky wrote: > What function should be used to convert an integer value to Int#? This function: f :: Int -> Int# f (I# x) = x > A character to Char#? Similarly: f :: Char -> Char# f (C# c) = c there aren't any library functions that do this for you. Cheers, Simon

1 0

Convert to unboxed value? How?
by Dimitry Golubovsky 27 Jan '05

27 Jan '05

Hi, What function should be used to convert an integer value to Int#? A character to Char#? Dimitry Golubovsky Middletown, CT

1 0