00:00:00 --- log: started haskell/02.02.06 00:08:42 --- join: shapr (~user@195.156.199.178) joined #haskell 01:35:51 --- join: Yurik (~yrashk@gw.telcos.net.ua) joined #haskell 01:38:20 hiall 01:38:22 hi all 02:04:17 --- join: cuelebre (~cuelebre@212.85.32.125) joined #haskell 02:32:27 * shapr bounces 02:32:30 hi Yurik 02:32:54 shapr hi 02:33:16 what's happening Yurik? 02:33:33 did you think of a better application data exchange protocol? 02:34:08 currently, I'm still covinced that lisp style syntax expressions (sexps) are the best overall 02:39:57 nope, haven't found yet 02:40:58 the most complex thing that I should exchange Erlang modules with non-Erlang modules and reduce data to Erlang "data types 02:41:09 s/"data types/"datatypes"/ 02:45:42 the Twisted server framework from TwistedMatrix.com has a sexp exchanger 02:46:08 currently it has modules for at least python, elisp, and java. I think I heard rumors of an erlang module 02:46:50 is it implemented in C and than wrapped for python,elisp and java? 02:47:58 actually, I think it's implemented in Python, Elisp, and Java 02:48:11 I think there's also a C version, but I don't know any details 02:48:13 let me check 02:54:46 hm 02:55:17 Yurik: check it out: http://twistedmatrix.com/spread.epy 03:10:58 * Yurik is browsing 03:35:20 --- join: ChoJin (~ask@cha213245038031.chello.fr) joined #haskell 03:35:23 hi 03:35:31 hello ChoJin 03:35:32 what's up? 03:35:48 anyone here ever coded a neuronal network in haskell ? :) 03:36:00 not me 03:36:10 I am a beginner in haskell, and I need to code this, I have some problem :) 03:36:12 but there are several experience haskellers here 03:36:32 I love haskell, I just need experience I guess :)) 03:36:37 me too 03:37:02 for both :) 03:37:09 I'm happy to find a haskell channel :) 03:37:18 k brb in 10min, need to eat :) 03:37:49 sure :) 03:48:50 back :) 03:49:38 ok I have a type neuron: type Neuron = (Value, [Value] -> Value), and I want to code a printNeuron 03:49:44 oh 03:49:50 well 03:50:00 as far as I understand it, you have two choices 03:50:23 you can either define a show instance for that type synonym 03:50:33 or you can use newtype and make your own show instance for that 03:50:51 how can I define a show instance for that type synonym ? 03:51:44 You can't make show instances for type synonyms. You'd need to use newtype 03:52:11 Or you could just write a function of another name that takes a Neuron 03:52:31 What should printNeuron do out of interest? 03:52:39 debugging :) 03:52:50 I build a list of neuron 03:52:56 and I want to print the state 03:53:10 The state is the first value? 03:53:17 yes 03:53:27 How is Value defined? 03:53:38 type Value = Float 03:53:39 :) 03:54:24 atm I use this: print $ buildLayer 5 42.0 sumNeurons 03:54:28 but of course it doesn't work 03:54:35 No instance for (Show ([Value] -> Value)) 03:54:50 Right, you want a fst in there 03:55:02 (fst is defined as fst (x, y) = x) 03:55:06 buildLayer just build a list of n neuron with value x and function sumNeurons :) 03:55:19 k :) 03:55:30 Oh, if it's a list you'll need to map fst across it 03:55:57 thx 03:55:59 it works :) 03:56:38 a problem I often have, is the library report doesn't seem to have all function inside 03:56:51 I need a description of _all_ standard function 03:57:00 ? 03:57:16 http://haskell.org/onlinelibrary/ 03:57:16 ? 03:57:28 http://www.haskell.org/definition/haskell98-library.ps.gz 03:57:32 I read this :) 03:57:35 You're missing the prelude 03:57:43 In the language report 03:59:21 Igloo: did you have ever coded a neuronal network ? :) 03:59:27 (just in case ;) ) 03:59:55 No 04:00:28 Sorry 04:02:56 you dont have to be sorry :))) 04:03:12 Sorry :-) 04:14:46 where is fst in the onlinelibrary ? 04:45:31 --- quit: Luke_ ("Client Exiting") 06:07:14 --- quit: shapr (Remote closed the connection) 06:48:46 http://research.microsoft.com/Users/simonpj/haskell98-revised/haskell98-report-html/standard-prelude.html#$vfst 06:55:03 --- quit: Yurik (Remote closed the connection) 07:28:19 --- join: Vutral (~ss@port-212.169.145.223.flat4all.de) joined #haskell 07:37:37 --- quit: Vutral (Remote closed the connection) 07:51:35 Where would I find examples of reading a file line-by-line in haskell? 07:58:05 --- join: Vutral (~ss@port-212.169.145.223.flat4all.de) joined #haskell 08:09:55 --- join: tmoertel (~chatzilla@pa-mtlebanon2a-268.pit.adelphia.net) joined #haskell 08:32:09 --- part: Vutral left #haskell 11:49:23 jewel: There are several ways... 11:49:31 I've been meaning to try out the interact function. 11:49:40 interact :: (String -> String) -> IO () 11:49:40 interact f = do s <- getContents 11:49:40 putStr (f s) 11:49:55 You can use the lines function to split a String at newlines. 11:51:43 * Logan goes to get food. 11:51:49 bye 11:52:24 jewel: I have whole bunch of programs that read data from file or stdin and process it and output results. 11:52:33 jewel: I can put them up for you when I get back if you want. 11:52:38 that would be kind 11:52:41 They are more verbose than necessary, as I'm still learning. 11:53:52 I won't critique :-) 11:56:12 hmmm- I'd like to see those, too 11:59:43 interact plays weirdly with a lot of stuff. 12:04:39 http://loganh.com/haskell/20020209/ 12:05:18 The problems they solve are at http://www.ntnu.edu.tw/acm/ProblemSetArchive/A_FINALS/1996/Finals96html/ 12:13:38 still noone who ever coded a neuronal network ? :) 13:16:27 --- join: shapr (~user@p-c2fbab34.easy.inet.fi) joined #haskell 13:40:22 --- quit: shapr (Read error: 54 (Connection reset by peer)) 13:40:55 --- join: shapr (~user@p-c2fbab34.easy.inet.fi) joined #haskell 14:23:06 --- quit: smkl (Remote closed the connection) 14:27:10 --- join: smkl (~sami@glubimox.yok.utu.fi) joined #haskell 15:15:36 --- quit: pHa (Read error: 104 (Connection reset by peer)) 15:16:44 * shapr wrestles with regexes 15:19:24 --- join: pHa (sjh@Sprint215.tbaytel.net) joined #haskell 15:30:00 shapr: What about them? 15:54:35 --- quit: ChoJin ("bye !!!") 17:16:34 --- quit: jemfinch (Remote closed the connection) 17:20:22 --- join: jemfinch (~jemfinch@rnie-99-168.resnet.ohio-state.edu) joined #haskell 17:53:00 --- nick: juhp_ -> juhp 18:26:27 --- quit: jemfinch (Remote closed the connection) 18:27:05 --- join: jemfinch (~jemfinch@rnie-99-168.resnet.ohio-state.edu) joined #haskell 18:54:36 --- quit: shapr (Read error: 60 (Operation timed out)) 20:33:24 --- quit: Igloo (carter.openprojects.net irc.openprojects.net) 20:33:24 --- quit: Heffalump (carter.openprojects.net irc.openprojects.net) 20:35:01 --- join: Heffalump (ganesh@munchkin.comlab.ox.ac.uk) joined #haskell 20:35:01 --- join: Igloo (~igloo@c93.keble.ox.ac.uk) joined #haskell 20:43:59 * tmoertel just finished the 10-20-30 card game from the 1996 ACM programming comp 20:47:25 tmoertel: Awesome! 20:50:01 Link: http://www.moertel.com/~thor/pragprog/acm-contest-1996/ 20:50:57 tmoertel: Try the bicycle delivery problem (if you know your graph theory). :P 20:51:08 Though that's hard if you haven't done any graphs in haskell and don't have a graph library. 20:51:23 I found the trellis problem was really really easy. And the prelim problem is even easier. 20:51:45 I think that I solved the bicycle program while in the shower today. 20:53:53 And, yes, looking at it like a graph problem rather than a geom search problem made the light bulb go off. 20:54:36 When I get a few minutes I'll code up the solution w/ FGL. 20:55:37 * xbill notes graph algorithms always come back to haunt you. 20:56:21 I've been emaning to test and fix my bicycle solution. I can think of a way to break it. :P 20:56:30 I'll have fun breaking people's submissions during practice on Sat. 20:56:40 Yeah, FGL makes it easy (once you fix FGL so it compiles). 20:56:57 What's on Sat? 20:58:34 We have programming team practices. 20:58:40 Practicing for the international in March in Hawaii. 20:58:53 I'm assistant coaching (so I have to write test data for all these problems and do the judging during practice). 20:59:01 For once I'm ahead of the game. I only have one more problem to solve. 20:59:19 Gotcha. 20:59:31 How do you intend to do the bike one? 21:00:43 Start with an empty graph. 21:00:53 Add start and end points as verticies. 21:01:04 Do the same with all points on the building-defining rectaingles. 21:01:48 Add edges between all vertices whose respective points are connected by a line that doesn't intersect the interior of a building. 21:02:02 Let the weights of the edges be the distance between the respective points. 21:02:12 Find shortest path from start vertex to end vertex. 21:02:25 At least, that's the approach I'm taking. 21:03:16 Yep, that's how I did it. 21:03:23 I made a list of line segments... 21:03:58 Which are the line segments making up all rectangles, plus the diagonals of each rectangle (to prevent cutting through the rectangle, but I need to refine that a little). 22:50:29 * tmoertel goes to bed 22:55:58 --- quit: tmoertel () 23:59:59 --- log: ended haskell/02.02.06