2012-11-10T00:04:30 *** amstan has joined #aichallenge 2012-11-10T00:04:30 *** ChanServ sets mode: +o amstan 2012-11-10T02:11:54 *** antimatroidl has joined #aichallenge 2012-11-10T02:26:29 *** Simas_J has joined #aichallenge 2012-11-10T02:30:23 *** antimatroidl has quit IRC (Quit: Leaving.) 2012-11-10T02:48:16 *** epicmonkey has joined #aichallenge 2012-11-10T02:54:25 *** phildunstan has joined #aichallenge 2012-11-10T03:11:15 *** bqf has quit IRC (*.net *.split) 2012-11-10T03:13:34 *** bqf has joined #aichallenge 2012-11-10T03:19:54 *** foRei has joined #aichallenge 2012-11-10T03:30:02 *** thestinger has quit IRC (Quit: WeeChat 0.3.9) 2012-11-10T03:50:01 *** amstan has quit IRC (Quit: Konversation terminated!) 2012-11-10T03:50:27 *** mleise has joined #aichallenge 2012-11-10T03:53:17 *** pairofdice has joined #aichallenge 2012-11-10T04:03:17 *** ecidforiap has joined #aichallenge 2012-11-10T04:05:06 *** pairofdice has quit IRC (Ping timeout: 265 seconds) 2012-11-10T04:18:40 *** epicmonkey has quit IRC (Ping timeout: 246 seconds) 2012-11-10T04:26:06 *** Simas_J has quit IRC (Ping timeout: 268 seconds) 2012-11-10T04:39:54 *** epicmonkey has joined #aichallenge 2012-11-10T04:42:49 *** pairofdice has joined #aichallenge 2012-11-10T05:11:14 *** sigh has joined #aichallenge 2012-11-10T05:24:34 *** antimatroidl has joined #aichallenge 2012-11-10T05:40:58 *** smiley1983 has quit IRC (Ping timeout: 252 seconds) 2012-11-10T05:42:44 *** smiley1983 has joined #aichallenge 2012-11-10T05:55:17 *** smiley1983 has quit IRC (Ping timeout: 255 seconds) 2012-11-10T05:57:10 *** smiley1983 has joined #aichallenge 2012-11-10T06:03:08 *** epicmonkey has quit IRC (Ping timeout: 245 seconds) 2012-11-10T06:32:31 *** pairofdice has quit IRC (Ping timeout: 265 seconds) 2012-11-10T06:44:24 *** tsc has quit IRC (Ping timeout: 245 seconds) 2012-11-10T06:53:32 *** antimatroidl1 has joined #aichallenge 2012-11-10T06:54:10 *** antimatroidl has quit IRC (Ping timeout: 260 seconds) 2012-11-10T07:10:42 *** phildunstan has quit IRC () 2012-11-10T07:28:41 *** g0llum has joined #aichallenge 2012-11-10T07:29:17 *** g0llum has quit IRC (Read error: Connection reset by peer) 2012-11-10T07:29:31 *** g0llum has joined #aichallenge 2012-11-10T07:34:25 oh, nice, another tronbot ;) 2012-11-10T07:36:23 g0llum: mine :) 2012-11-10T07:37:01 It's got some problems - outright bugs in multiplayer mode, but seems to be non-suicidal in two-player games 2012-11-10T07:37:03 cool xD a winner, too! 2012-11-10T07:37:42 Do you think I should make smaller maps? Tonight is my last chance to change anything for the next few days 2012-11-10T07:39:42 i think, they're ok. bit more water, maybe .. 2012-11-10T07:42:48 *** pairofdice has joined #aichallenge 2012-11-10T07:45:06 Maps replaced with ones with more water - I'll check back in an hour or two. Bots will need to be reconnected (had to stop script to load new maps) 2012-11-10T07:45:34 *** ramn[] has quit IRC (Read error: Connection reset by peer) 2012-11-10T07:45:54 yea, i see ;) nice! 2012-11-10T07:50:28 At least, they were supposed to have more water... I'll look into it later 2012-11-10T07:50:41 *** epicmonkey has joined #aichallenge 2012-11-10T07:51:23 nvm, it's ok 2012-11-10T07:52:45 *** Scooper has joined #aichallenge 2012-11-10T07:53:55 what#s your bot doing ? does not seem very aggressive 2012-11-10T07:54:14 monte carlo, and probably wrongly :) 2012-11-10T07:55:18 stopped and restarted tron server; worked out what I did wrong, the maps should hopefully have more water now :) 2012-11-10T07:55:26 ohh monte carlo what ? 2012-11-10T07:55:30 ok 2012-11-10T07:55:54 yes 2012-11-10T07:56:12 it just plays out random non-suicidal moves, recording the result back to the first move in the sequence 2012-11-10T07:56:35 then chooses the move which "led" to the greatest number of wins 2012-11-10T07:58:03 you're doing some bfs there to evaluate ? 2012-11-10T07:58:18 none at all :) 2012-11-10T07:58:25 OOOOhhh 2012-11-10T07:58:31 wow 2012-11-10T07:58:50 the only "search" is the thousands of random playouts 2012-11-10T08:02:52 seems, i got to try harder ;) 2012-11-10T08:13:53 read a1k0ns postmortem, the voronoi diagram was blinking at me all the time, so i tried something like that 1st. 2012-11-10T08:21:45 *** dici has quit IRC (Ping timeout: 264 seconds) 2012-11-10T08:24:02 *** epicmonkey has quit IRC (Ping timeout: 246 seconds) 2012-11-10T08:44:33 *** phildunstan has joined #aichallenge 2012-11-10T08:51:37 *** ramn[mba] has joined #aichallenge 2012-11-10T08:53:34 *** ramn[home] has joined #aichallenge 2012-11-10T09:32:10 *** ramn[mba] has quit IRC (Quit: ramn[mba]) 2012-11-10T09:55:43 g0llum: yes, I like the voronoi diagrams, that idea is awesome; I'm definitely interested in writing a tree-searching + heuristics bot 2012-11-10T09:56:06 however, there are some other things I want to do first, and paid work in the way 2012-11-10T10:00:14 *** pairofdice has quit IRC (Quit: In girum imus nocte et consumimur igni.) 2012-11-10T10:00:26 *** mcstar has joined #aichallenge 2012-11-10T10:11:30 *** pairofdice has joined #aichallenge 2012-11-10T10:16:27 ;) yea, take your time.. 2012-11-10T10:21:25 *** moises has joined #aichallenge 2012-11-10T10:41:08 before, I added "num_players" and "agents_per_player" to the starting info; it could be deduced before, but now it tells you 2012-11-10T10:42:59 *** sigh has quit IRC (Remote host closed the connection) 2012-11-10T10:47:13 smiley1983: are you rewriting the engine? 2012-11-10T11:08:17 *** moises has quit IRC (Remote host closed the connection) 2012-11-10T11:08:39 *** moises has joined #aichallenge 2012-11-10T11:50:05 *** moises has quit IRC (Remote host closed the connection) 2012-11-10T12:32:05 http://tcp-score.appspot.com/?user=ocmc_002&port=2086 2012-11-10T12:32:52 ^^ recycled the cheesy score graph from the ants contest 2012-11-10T12:51:55 *** phildunstan has quit IRC (Ping timeout: 245 seconds) 2012-11-10T13:23:50 *** thestinger has joined #aichallenge 2012-11-10T13:58:59 *** Accoun has quit IRC () 2012-11-10T14:18:43 *** Accoun has joined #aichallenge 2012-11-10T14:25:22 antimatroidl1: how much have you been thinking about RPS? 2012-11-10T14:26:16 mcstar: \o 2012-11-10T14:26:24 o/ 2012-11-10T14:26:31 mcstar: http://rosalind.info/problems/as-table/ puzzles I bookmarked for later 2012-11-10T14:26:33 :P 2012-11-10T14:27:34 Codecademy online Python course 2012-11-10T14:27:36 wtf 2012-11-10T14:28:10 so, now we want to learn bioinformatics? 2012-11-10T14:28:16 sounds challenging 2012-11-10T14:29:09 thestinger: i switched to a non-antialiased font for all monospace fonts 2012-11-10T14:29:31 I didn't actually look at any yet 2012-11-10T14:29:45 gohufont is nice 2012-11-10T14:29:52 quite readable 2012-11-10T14:30:11 wait, fonts or bioinformatics problems? 2012-11-10T14:31:11 those bioinformatics ones 2012-11-10T14:34:08 thestinger: how secure do you think this openid stuff is? 2012-11-10T14:34:43 https://en.wikipedia.org/wiki/OpenID#Security not secure? :P 2012-11-10T14:35:11 i imagine, i can put the openid logo on my site 2012-11-10T14:35:16 redirect the client for google 2012-11-10T14:35:28 but run a js script that logs the keypresses 2012-11-10T14:35:30 and sends them to me 2012-11-10T14:38:18 doubt it 2012-11-10T14:38:21 they log in our google 2012-11-10T14:38:24 on google* 2012-11-10T14:38:26 not your site 2012-11-10T14:39:16 yeah, but i can run the google login site in a different frame or something 2012-11-10T14:39:22 and my code could still be loaded 2012-11-10T14:39:43 im certain, you can do this somehow, im just stupid 2012-11-10T14:40:20 first of all, you have to trust the site that provides openid authentication 2012-11-10T14:40:34 you also have to trust the openid protocol, which isn't secure 2012-11-10T14:40:37 i wouldnt use it to authenticate with russian sites, for example 2012-11-10T14:41:03 anyway screw openid 2012-11-10T14:41:10 it just centralizes trust even more 2012-11-10T14:44:09 i think im gonna quickly write up an rpscontest bot 2012-11-10T14:50:17 *** jacob_strauss has joined #aichallenge 2012-11-10T14:52:36 SSPPRSPSPRPRPPSRRSRPPSRSRSSSSSSPPPPSRPRSRPSPRPSPRR 2012-11-10T14:54:29 thanks! the winning sequence i was looking for 2012-11-10T14:54:39 no, wait, i need 1000 rounds 2012-11-10T14:54:54 yea, answer is S in that case 2012-11-10T14:55:50 mcstar: just solve it the spammer way 2012-11-10T14:55:56 markov chains and more markov chains 2012-11-10T14:55:58 pretty much related to that rosalind / bioinformatics stuff, isn#'t it ? 2012-11-10T14:56:52 so nice, when you cant kill a python program with ^C 2012-11-10T14:57:13 thestinger: why dont you write a bot too? 2012-11-10T14:57:30 don't feel like :P 2012-11-10T14:57:39 need to finish this crappy web thing 2012-11-10T15:10:13 *** phildunstan has joined #aichallenge 2012-11-10T15:40:08 *** amstan has joined #aichallenge 2012-11-10T15:40:09 *** ChanServ sets mode: +o amstan 2012-11-10T16:03:43 *** Areks|2 has joined #aichallenge 2012-11-10T16:07:45 *** Areks has quit IRC (Ping timeout: 272 seconds) 2012-11-10T16:09:56 *** djr_ has joined #aichallenge 2012-11-10T16:16:56 *** jacob_strauss has quit IRC (Quit: jacob_strauss) 2012-11-10T16:23:46 thestinger: there is no fst and snd in python >.< 2012-11-10T16:23:55 mcstar: fst, snd = (1, 2) 2012-11-10T16:24:02 thats useless 2012-11-10T16:24:07 why :P 2012-11-10T16:24:14 because those are not functions 2012-11-10T16:24:30 nvm, ill use a lambda 2012-11-10T16:24:36 itemgetter(0) 2012-11-10T16:24:54 hm 2012-11-10T16:24:59 mcstar: http://docs.python.org/3.3/library/operator.html#module-operator 2012-11-10T16:25:01 what is that a prt of? 2012-11-10T16:25:13 oh, >=3? 2012-11-10T16:26:02 >>> from operator import itemgetter 2012-11-10T16:26:04 >>> itemgetter(0)([4,3,2]) 2012-11-10T16:26:06 4 2012-11-10T16:26:24 great 2012-11-10T16:26:30 workarounds for everything 2012-11-10T16:26:35 :) 2012-11-10T16:26:50 why is that a workaround? :P 2012-11-10T16:27:00 they're the actual functions python uses 2012-11-10T16:27:21 well, if you could partially apply [] ... 2012-11-10T16:27:42 whats the name of the function that [] is translated to? 2012-11-10T16:28:31 nothing? :P 2012-11-10T16:28:40 oh 2012-11-10T16:28:52 1 0 BUILD_LIST 0 2012-11-10T16:28:54 i thought it is translated to a .__method__() 2012-11-10T16:28:59 it becomes a bytecode 2012-11-10T16:29:10 there's even a bytecode for list comprehensions (which is why they're fast in cpython) 2012-11-10T16:29:20 they are? 2012-11-10T16:29:26 compared to what? 2012-11-10T16:29:32 XD 2012-11-10T16:29:37 for x in xs: a.append(x) 2012-11-10T16:29:42 [x for x in xs] 2012-11-10T16:29:45 the 2nd is like 20x faster 2012-11-10T16:29:48 now, thats cheating 2012-11-10T16:29:49 in pypy they're the same 2012-11-10T16:29:52 *** dici has joined #aichallenge 2012-11-10T16:30:01 .append must copy the whole vector 2012-11-10T16:30:11 no 2012-11-10T16:30:13 (maybe it doesnt) 2012-11-10T16:30:18 it does the same thing 2012-11-10T16:30:22 but .append is a method lookupo 2012-11-10T16:30:24 lookup * 2012-11-10T16:30:25 probably uses more storage then it needs then 2012-11-10T16:30:36 lists have O(1) append 2012-11-10T16:30:39 like std::vector 2012-11-10T16:30:40 amortized 2012-11-10T16:30:52 what? 2012-11-10T16:31:07 when it's full, it multiplies the capacity by 1.5x 2012-11-10T16:31:21 which means it's O(1) amortized to append to a dynamic array that's implemented sanely 2012-11-10T16:31:26 yeah, so how could you calculate amortized complexities on that?? 2012-11-10T16:31:38 because the memory allocation algorithms are known 2012-11-10T16:31:42 did you average over all vector lenghts? 2012-11-10T16:32:02 a = [1,2,3] 2012-11-10T16:32:03 anyway, it has to compy at some point 2012-11-10T16:32:07 copy*( 2012-11-10T16:32:15 yes, it does 2012-11-10T16:32:35 also, it is not a big problem 2012-11-10T16:32:36 and it's known how often it will reallocate 2012-11-10T16:32:38 *** pairofdice has quit IRC (Quit: In girum imus nocte et consumimur igni.) 2012-11-10T16:32:44 because you know how much it multiplies the capacity 2012-11-10T16:32:45 since i guess python keeps every object on the heap 2012-11-10T16:32:52 and the vectors only store a reference to them 2012-11-10T16:33:00 yes 2012-11-10T16:33:08 array.array stores them packed 2012-11-10T16:33:55 *** g0llum has quit IRC (Read error: Connection reset by peer) 2012-11-10T16:34:13 mcstar: https://en.wikipedia.org/wiki/Dynamic_array#Performance look, a nice table :P 2012-11-10T16:34:32 I'm not sure how you can compare a balanced tree to them though 2012-11-10T16:34:36 maybe not a nice table 2012-11-10T16:35:01 that is a doubly linked list 2012-11-10T16:35:05 anyway when you say insert into a std::map is O(logn) - that's amortized too 2012-11-10T16:35:10 mcstar: [] is a dynamic array 2012-11-10T16:35:15 not a linked list 2012-11-10T16:35:20 no, the list in the table 2012-11-10T16:35:23 not a 'list' 2012-11-10T16:35:28 but a dll 2012-11-10T16:36:01 mcstar: http://wiki.python.org/moin/TimeComplexity 2012-11-10T16:36:14 why should i read that? 2012-11-10T16:36:27 because you don't believe me :P 2012-11-10T16:36:39 i didnt say that 2012-11-10T16:56:26 i had an idea for the longest path problem in tron 2012-11-10T16:56:57 http://www.reddit.com/r/math/comments/12z5b8/why_is_longest_path_npcomplete/ from that post it suggests you can solve the longest path problem on a DAG 2012-11-10T16:57:14 of course you're not going to have a DAG most of the time with the remaining space once separated 2012-11-10T16:57:31 but you could take samples of DAG's on what's left and take the best max path found on those 2012-11-10T17:03:10 tron is mostly boring 2012-11-10T17:16:31 you're mostly boring 2012-11-10T17:16:36 :P 2012-11-10T17:16:56 you clearly haven't written a tron bot 2012-11-10T17:17:04 seeing your little guy take the kill is awesome 2012-11-10T17:17:17 it's precise and brutal 2012-11-10T17:17:46 especially when it teaches you a thing or two about your own test examples 2012-11-10T17:19:19 antimatroidl1: i said boring, because there is not much action 2012-11-10T17:19:30 the 'battle' is fairly limited 2012-11-10T17:19:44 you are just trying to cut the other guy off 2012-11-10T17:20:04 who fucks up first loses 2012-11-10T17:20:51 it's not always easy to see what is or isn't a fuck up 2012-11-10T17:21:01 especially for more complicated maps 2012-11-10T17:25:56 antimatroidl1: how about, adding a limited number of lifes? 2012-11-10T17:26:07 you could cut through the trail that many times 2012-11-10T17:26:39 meh, i like the simple rules 2012-11-10T17:28:30 great 2012-11-10T17:28:43 my bot has a 10% success rate against the hardest rps bots 2012-11-10T17:29:00 (no, not great, really) 2012-11-10T17:40:11 *** thestinger has quit IRC (Ping timeout: 246 seconds) 2012-11-10T17:42:50 *** sigh has joined #aichallenge 2012-11-10T17:42:59 mcstar: I don't know why people play games where the optimal bot is random-bot 2012-11-10T17:43:17 it itsnt 2012-11-10T17:43:19 isnt 2012-11-10T17:43:35 which i find quite funny 2012-11-10T17:43:43 since it shows an aspect of real life 2012-11-10T17:44:09 first of all, equilibrium doesnt apply 2012-11-10T17:44:20 since you are not playing for an infinite amount of time 2012-11-10T17:44:32 neglectible 2012-11-10T17:44:37 no 2012-11-10T17:44:59 and players dont necessarily follow 'best' strategies 2012-11-10T17:44:59 *negligible 2012-11-10T17:45:14 since they are trying to outsmart each other 2012-11-10T17:45:37 if you turn RPS into an opponent modeling problem, it is your fault that you end up with 10% :) 2012-11-10T17:45:43 if you have opponents, that dont play 'rationally' in the mathematics sense, you can exploint them 2012-11-10T17:45:50 just use random bot and win 50% of the time! 2012-11-10T17:46:07 mleise: but that doesnt take me to the top of the leaderbord 2012-11-10T17:46:09 board* 2012-11-10T17:46:36 so you want to exploit other people's weaknesses? ^^ 2012-11-10T17:46:47 thats whats the game is about, yes 2012-11-10T17:46:56 predictibility and decision correlation 2012-11-10T17:47:53 have you already employed the basic strategies that are listed on the website of CSC? 2012-11-10T17:47:54 *** thestinger has joined #aichallenge 2012-11-10T17:48:01 no 2012-11-10T17:48:04 whats csc? 2012-11-10T17:48:08 http://csclub.uwaterloo.ca/~j3camero/contest/fall2009/sample_code.php 2012-11-10T17:48:26 well, i might read it 2012-11-10T17:48:28 idk 2012-11-10T17:48:32 i want to do my own thing 2012-11-10T17:48:57 im just doing some simple statistics, and wondering about normalization now 2012-11-10T18:07:58 *** antimatroidl1 has quit IRC (Ping timeout: 264 seconds) 2012-11-10T18:08:21 why don't you just switch to random play if the oponent seems to be stronger after a few rounds? 2012-11-10T18:09:23 great idea! 2012-11-10T18:09:41 really? 2012-11-10T18:09:45 no 2012-11-10T18:10:17 *** antimatroidl has joined #aichallenge 2012-11-10T18:24:14 if you want some nice ideas for rps read the iocaine powder summary http://www.ofb.net/~egnor/iocaine.html from the The First International RoShamBo Programming Competition (1999) http://webdocs.cs.ualberta.ca/~darse/rsb-results1.html 2012-11-10T18:41:58 *** dici has quit IRC (Read error: Connection reset by peer) 2012-11-10T18:47:53 *** foRei has quit IRC (Quit: Bye) 2012-11-10T18:48:35 *** phildunstan has quit IRC (Quit: fallen asleep at his desk) 2012-11-10T18:53:20 gn 2012-11-10T18:53:22 *** mcstar has quit IRC (Quit: mcstar) 2012-11-10T19:34:44 *** Scooper has quit IRC (Quit: Leaving) 2012-11-10T21:14:19 *** scribble has joined #aichallenge 2012-11-10T21:39:57 *** sigh has quit IRC (Remote host closed the connection) 2012-11-10T23:12:52 *** antimatroidl has quit IRC (Quit: Leaving.) 2012-11-10T23:15:11 *** sigh has joined #aichallenge