--- Log opened Sat Jun 29 00:00:09 2013 | ||
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 00:21 | |
-!- mode/#shogun [+o iglesiasg] by ChanServ | 00:21 | |
-!- van51 [~van51@athedsl-408350.home.otenet.gr] has left #shogun ["QUIT :Leaving."] | 00:59 | |
-!- lambday [67157d37@gateway/web/freenode/ip.103.21.125.55] has joined #shogun | 01:35 | |
-!- nube [~rho@49.244.126.232] has joined #shogun | 03:12 | |
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving] | 03:34 | |
-!- pickle27 [~Kevin@d67-193-243-174.home3.cgocable.net] has quit [Quit: Leaving] | 03:40 | |
-!- zxtx [~zv@cpe-66-68-190-37.austin.res.rr.com] has quit [Ping timeout: 268 seconds] | 04:00 | |
-!- zxtx [~zv@cpe-66-68-190-37.austin.res.rr.com] has joined #shogun | 04:15 | |
-!- foulwall [~user@2001:da8:215:6100:c16f:bb5c:eea9:bb6e] has joined #shogun | 04:54 | |
-!- gsomix [~Miranda@109.169.186.163] has joined #shogun | 04:56 | |
gsomix | sonney2k: hello! I'm working, but will be available in IRC little later. Do not worry. | 04:58 |
---|---|---|
gsomix | cu | 05:01 |
-!- gsomix [~Miranda@109.169.186.163] has quit [Client Quit] | 05:01 | |
foulwall | sonney2k: the gp demo needs GaussianLikelyhood, which needs eigen3. and I wanna install eigen3 dependency and recompile shogun. May I do that? | 05:15 |
-!- foulwall [~user@2001:da8:215:6100:c16f:bb5c:eea9:bb6e] has quit [Ping timeout: 245 seconds] | 05:30 | |
-!- lambday [67157d37@gateway/web/freenode/ip.103.21.125.55] has quit [] | 06:09 | |
-!- nube [~rho@49.244.126.232] has quit [Ping timeout: 252 seconds] | 06:14 | |
-!- nube [~rho@49.244.44.30] has joined #shogun | 06:27 | |
-!- foulwall [~user@2001:da8:215:6100:786a:b4ef:bdd9:a060] has joined #shogun | 07:16 | |
-!- hushell [~hushell@c-24-21-141-32.hsd1.or.comcast.net] has joined #shogun | 07:19 | |
-!- foulwall [~user@2001:da8:215:6100:786a:b4ef:bdd9:a060] has quit [Ping timeout: 264 seconds] | 07:51 | |
-!- foulwall [~user@2001:da8:215:6100:d0b2:adb:70f2:f835] has joined #shogun | 08:02 | |
-!- foulwall [~user@2001:da8:215:6100:d0b2:adb:70f2:f835] has quit [Ping timeout: 245 seconds] | 08:59 | |
-!- foulwall` [~user@2001:da8:215:6100:6c02:37e1:2d7:e591] has joined #shogun | 08:59 | |
@wiking | sonney2k: i cannot fix json installs on fbsd ... and especially not on github | 09:33 |
-!- foulwall` [~user@2001:da8:215:6100:6c02:37e1:2d7:e591] has quit [Remote host closed the connection] | 09:44 | |
-!- gsomix [~gsomix@83.234.169.104] has joined #shogun | 09:53 | |
* gsomix sings: "I'm going off the rails on a crazy train..." | 09:54 | |
-!- gsomix [~gsomix@83.234.169.104] has quit [Ping timeout: 256 seconds] | 10:00 | |
-!- foulwall [~user@2001:da8:215:6100:6c02:37e1:2d7:e591] has joined #shogun | 10:10 | |
-!- gsomix [~gsomix@185.6.239.242] has joined #shogun | 10:12 | |
-!- foulwall [~user@2001:da8:215:6100:6c02:37e1:2d7:e591] has quit [Remote host closed the connection] | 10:15 | |
-!- foulwall [~user@li379-21.members.linode.com] has joined #shogun | 10:16 | |
-!- foulwall [~user@li379-21.members.linode.com] has quit [Remote host closed the connection] | 10:17 | |
-!- foulwall [~user@li379-21.members.linode.com] has joined #shogun | 10:21 | |
-!- foulwall [~user@li379-21.members.linode.com] has quit [Remote host closed the connection] | 10:24 | |
-!- foulwall [~user@li379-21.members.linode.com] has joined #shogun | 10:25 | |
-!- foulwall [~user@li379-21.members.linode.com] has quit [Remote host closed the connection] | 10:32 | |
-!- foulwall [~user@li379-21.members.linode.com] has joined #shogun | 10:32 | |
-!- foulwall [~user@li379-21.members.linode.com] has quit [Remote host closed the connection] | 10:35 | |
-!- foulwall [~user@li379-21.members.linode.com] has joined #shogun | 10:35 | |
-!- foulwall [~user@li379-21.members.linode.com] has quit [Remote host closed the connection] | 10:46 | |
-!- nube [~rho@49.244.44.30] has quit [Quit: Leaving.] | 11:08 | |
-!- gsomix [~gsomix@185.6.239.242] has quit [Quit: Leaving] | 11:29 | |
-!- foulwall [~user@2001:da8:215:6100:89ff:9169:bc0b:b5fe] has joined #shogun | 11:59 | |
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun | 12:15 | |
shogun-notifier- | shogun: iglesias :develop * 7f8700e / src/shogun/machine/StructuredOutputMachine.cpp: https://github.com/shogun-toolbox/shogun/commit/7f8700ed8896404d112a344f85aad9859fcee99d | 12:15 |
shogun-notifier- | shogun: Fix possible memory leak in StructuredOutputMachine | 12:15 |
shogun-notifier- | shogun: Fernando Iglesias :develop * 0ef3c6e / src/shogun/machine/StructuredOutputMachine.cpp: https://github.com/shogun-toolbox/shogun/commit/0ef3c6e15eeb83489bb83a07846846a6a9eee028 | 12:15 |
shogun-notifier- | shogun: Merge pull request #1197 from iglesias/feature/structure_fix_leak | 12:15 |
shogun-notifier- | shogun: | 12:15 |
shogun-notifier- | shogun: Fix possible memory leak in StructuredOutputMachine | 12:15 |
shogun-buildbot | build #993 of cyg1 - libshogun is complete: Failure [failed configure] Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/993 blamelist: iglesias <fernando.iglesiasg@gmail.com> | 12:19 |
-!- vgorbati [~vgorbati@95.111.237.216] has joined #shogun | 12:25 | |
shogun-buildbot | build #1161 of bsd1 - libshogun is complete: Failure [failed test_1] Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/1161 blamelist: iglesias <fernando.iglesiasg@gmail.com> | 12:26 |
foulwall | ls | 12:37 |
-!- travis-ci [~travis-ci@ec2-107-21-65-148.compute-1.amazonaws.com] has joined #shogun | 12:37 | |
travis-ci | [travis-ci] it's Fernando Iglesias's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: http://travis-ci.org/shogun-toolbox/shogun/builds/8565857 | 12:37 |
-!- travis-ci [~travis-ci@ec2-107-21-65-148.compute-1.amazonaws.com] has left #shogun [] | 12:37 | |
shogun-buildbot | build #1162 of bsd1 - libshogun is complete: Failure [failed test_1] Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/1162 blamelist: Fernando Iglesias <fernando.iglesiasg@gmail.com> | 12:38 |
shogun-buildbot | build #994 of cyg1 - libshogun is complete: Failure [failed compile] Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/994 blamelist: Fernando Iglesias <fernando.iglesiasg@gmail.com> | 12:41 |
@sonney2k | wiking, why not? | 12:42 |
@sonney2k | wiking, just install json-c manually from https://github.com/json-c/json-c and all good | 12:43 |
foulwall | hi sonney2k, can I install eigen3 and recompile shogun on the virtual box? | 12:44 |
foulwall | I'm now merging dvalcarce's demo | 12:44 |
@sonney2k | foulwall, sure but I can do that as well (tonight) | 12:46 |
@sonney2k | foulwall, go go :) | 12:46 |
foulwall | sonney2k: :) | 12:47 |
-!- foulwall [~user@2001:da8:215:6100:89ff:9169:bc0b:b5fe] has quit [Ping timeout: 245 seconds] | 13:25 | |
-!- vgorbati [~vgorbati@95.111.237.216] has quit [Ping timeout: 256 seconds] | 14:09 | |
-!- vgorbati [~vgorbati@95.111.237.216] has joined #shogun | 14:31 | |
-!- vgorbati [~vgorbati@95.111.237.216] has quit [Ping timeout: 246 seconds] | 14:40 | |
-!- vgorbati [~vgorbati@95.111.237.216] has joined #shogun | 14:41 | |
-!- gsomix [~androirc@85.26.234.240] has joined #shogun | 14:45 | |
gsomix | sonney2k, lisitsyn I'm in train. CU tomorrow. | 14:47 |
-!- gsomix [~androirc@85.26.234.240] has quit [Quit: AndroIRC - Android IRC Client ( http://www.androirc.com )] | 14:53 | |
-!- vgorbati [~vgorbati@95.111.237.216] has quit [Ping timeout: 260 seconds] | 15:04 | |
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout] | 15:15 | |
-!- rasmi [~rasmi@117.197.235.55] has joined #shogun | 15:44 | |
-!- rasmi [~rasmi@117.197.235.55] has quit [Ping timeout: 240 seconds] | 15:50 | |
-!- lisitsyn [~lisitsyn@5-28-21-24.clients.tlt.100megabit.ru] has quit [Quit: Leaving.] | 16:00 | |
-!- lisitsyn [~lisitsyn@5-28-21-24.clients.tlt.100megabit.ru] has joined #shogun | 16:02 | |
-!- vgorbati [~vgorbati@91.216.173.29] has joined #shogun | 16:07 | |
-!- foulwall [~user@2001:da8:215:c252:5ce7:6005:ba6c:41c8] has joined #shogun | 17:30 | |
-!- foulwall` [~user@2001:da8:215:c252:186b:938e:c7c0:7f7] has joined #shogun | 17:37 | |
-!- foulwall [~user@2001:da8:215:c252:5ce7:6005:ba6c:41c8] has quit [Ping timeout: 245 seconds] | 17:39 | |
-!- zxtx [~zv@cpe-66-68-190-37.austin.res.rr.com] has quit [Ping timeout: 246 seconds] | 17:53 | |
-!- van51 [~van51@athedsl-408350.home.otenet.gr] has joined #shogun | 18:06 | |
-!- foulwall` [~user@2001:da8:215:c252:186b:938e:c7c0:7f7] has quit [Ping timeout: 245 seconds] | 18:08 | |
@sonney2k | wiking, ohh crasher in random | 19:59 |
@sonney2k | wiking, http://dpaste.com/1275425/ | 19:59 |
van51 | sonney2k: hello | 20:06 |
van51 | sonney2k: I've sent a PR for my extension of the dotfeatures class | 20:08 |
van51 | sonney2k: it's a WIP | 20:08 |
van51 | sonney2k: so when you have time and feel like it you can start reviewing it | 20:08 |
van51 | I'll be back later | 20:09 |
van51 | bye for now! | 20:09 |
-!- van51 [~van51@athedsl-408350.home.otenet.gr] has quit [Quit: Leaving.] | 20:09 | |
-!- vgorbati [~vgorbati@91.216.173.29] has quit [Read error: Operation timed out] | 20:13 | |
-!- pickle27 [~Kevin@d67-193-243-174.home3.cgocable.net] has joined #shogun | 20:18 | |
@sonney2k | lisitsyn, http://www.wired.com/wiredscience/2013/06/could-superman-punch-someone-into-space/ | 20:41 |
lisitsyn | sonney2k: science! | 20:58 |
@sonney2k | I like the force to superman-b's head being >4000 aircraft carriers | 21:03 |
lisitsyn | sonney2k: I wonder if this is enough to burn someone | 21:05 |
lisitsyn | sounds like a lot of energy at the same time at the same place | 21:06 |
lisitsyn | pickle27: hey! | 21:06 |
pickle27 | lisitsyn: hey there! | 21:06 |
lisitsyn | pickle27: howdy? | 21:06 |
pickle27 | good I finished up ffdiag yesterday | 21:07 |
pickle27 | Im going to code the ffsep algorithm next which uses ffdiag to do BSS | 21:07 |
lisitsyn | pickle27: ffdiag is made of your glue code and that C code right? | 21:09 |
pickle27 | the one I showed you was jadiag, ffdiag is another port from the R package but there was more R and less C | 21:10 |
pickle27 | so its more of a direct port | 21:10 |
lisitsyn | ah sorry they are all **diag :D | 21:10 |
pickle27 | yeah its confusing | 21:10 |
pickle27 | jadiag is part of JADE | 21:10 |
pickle27 | but I think I am going to do ffsep first because the next steps are more clear | 21:11 |
lisitsyn | oh another thing called JADE | 21:11 |
lisitsyn | :D | 21:11 |
lisitsyn | that's java agent development framework for me | 21:11 |
lisitsyn | :D | 21:11 |
lisitsyn | pickle27: ok I am going to finally really check out references | 21:11 |
pickle27 | yeah I think jadiag is the diagonalizer for the JADE BSS alg | 21:11 |
lisitsyn | pickle27: how do you design it now? | 21:12 |
lisitsyn | I mean is it a set of functions? | 21:12 |
lisitsyn | we've got to get to some PR faster I think! | 21:13 |
pickle27 | at the moment yes | 21:13 |
pickle27 | yeah I was thinking I should put something together so people can see it | 21:13 |
lisitsyn | pickle27: I don't mind if you put these functions to shogun/lib or so | 21:14 |
-!- zxtx [~zv@12.231.120.253] has joined #shogun | 21:14 | |
lisitsyn | other way you may put them to the class directly | 21:14 |
lisitsyn | pickle27: what is the class I am talking about btw :D | 21:15 |
lisitsyn | pickle27: if you have time now lets think about its interface | 21:15 |
pickle27 | shogun/lib looks like an alright spot for now | 21:15 |
pickle27 | sure | 21:15 |
lisitsyn | the input is indeed simple | 21:16 |
lisitsyn | that's DotFeatures I guess | 21:16 |
pickle27 | yeah | 21:16 |
pickle27 | and I think that eventually we'll want the diagonalization techniques to be private | 21:16 |
lisitsyn | but the output? | 21:16 |
pickle27 | well the algorithm would separate the sources | 21:17 |
pickle27 | hmm | 21:17 |
pickle27 | labels doesn't make sense | 21:17 |
lisitsyn | pickle27: no not labels for sure | 21:17 |
pickle27 | maybe it returns DotFeatures | 21:17 |
lisitsyn | I guess that's one feature in multiple features out? | 21:17 |
lisitsyn | feature like features instance | 21:17 |
lisitsyn | in shogun sense | 21:17 |
pickle27 | yeah that actually makes sense | 21:18 |
lisitsyn | ok then we need new base class | 21:19 |
pickle27 | yeah | 21:20 |
pickle27 | do we want these to live in Shogun/lib or somewhere else | 21:20 |
lisitsyn | pickle27: what? base class? | 21:20 |
pickle27 | yeah I was envisioning a director that had the base class and then all the methods | 21:21 |
pickle27 | and by methods I mean derrived classes | 21:21 |
@sonney2k | lisitsyn, why do we have a metric directory for just one class? | 21:21 |
lisitsyn | sonney2k: already merged? | 21:21 |
lisitsyn | sonney2k: that's indeed confusing | 21:22 |
lisitsyn | pickle27: we could put them to converter actually | 21:23 |
lisitsyn | returning combined features | 21:23 |
lisitsyn | that's just 'vertically' stacked features | 21:23 |
lisitsyn | one might obtain 'subfeatures' quite easily | 21:23 |
@sonney2k | I have no idea where this is going but *I* would have put LMNN in multiclass and not added an additional distance | 21:23 |
lisitsyn | sonney2k: we have the same problem for transfer | 21:23 |
@sonney2k | lisitsyn, which problem? | 21:24 |
lisitsyn | sonney2k: too light folder | 21:24 |
@sonney2k | $ find shogun/transfer/ | wc -l | 21:25 |
@sonney2k | 52 | 21:25 |
@sonney2k | > 1 | 21:25 |
lisitsyn | sonney2k: ha didn't know | 21:25 |
lisitsyn | :D | 21:25 |
lisitsyn | sonney2k: I agree on LMNN -> multiclass | 21:25 |
pickle27 | lisitsyn: converter might make sense | 21:27 |
pickle27 | or preproccesor | 21:27 |
pickle27 | brb one sec | 21:30 |
-!- pickle27 [~Kevin@d67-193-243-174.home3.cgocable.net] has quit [Quit: Leaving] | 21:30 | |
@sonney2k | lisitsyn, bah I killed the cygwin machine while trying to upgrade cygwin on it | 21:31 |
lisitsyn | sonney2k: I find cygwin disgusting | 21:32 |
-!- pickle27 [~Kevin@d67-193-243-174.home3.cgocable.net] has joined #shogun | 21:37 | |
pickle27 | lisitsyn: hey Im back | 21:37 |
pickle27 | what do you think between converters and preprocessor | 21:38 |
lisitsyn | pickle27: preprocessor is something that can be initialized | 21:38 |
lisitsyn | then used | 21:38 |
@sonney2k | lisitsyn, we should have this for shogun bug #1000 https://www.youtube.com/watch?v=oAiVsbXVP6k | 21:38 |
lisitsyn | sonney2k: haha nice | 21:39 |
lisitsyn | sonney2k: the guy enjoys the joke a lot haha | 21:40 |
pickle27 | okay yeah converter makes more sense | 21:40 |
lisitsyn | pickle27: I am unsure is there something that works online in BSS? | 21:40 |
pickle27 | what do you mean | 21:41 |
lisitsyn | pickle27: like you first learn | 21:42 |
lisitsyn | then work out of sample | 21:42 |
lisitsyn | pickle27: like PCA | 21:42 |
pickle27 | I don't really think so (Im still pretty new to BSS) | 21:42 |
lisitsyn | yeah me too | 21:42 |
pickle27 | It looks like your learn a demix matrix which applies to the given input | 21:42 |
lisitsyn | lets ask andreas | 21:42 |
pickle27 | so it wouldn't really be useful to new data | 21:43 |
lisitsyn | pickle27: could you please write a mail with that question? | 21:43 |
lisitsyn | we may not oversee that, if there are some out of sample extensions we should take a look | 21:43 |
pickle27 | okay I'll put something out on the mailing list | 21:44 |
@sonney2k | wiking, and buildbot ubu1 - libshogun ist offline | 21:44 |
pickle27 | I'll also cc Andreas as Im not sure if hes on the mailing list or not | 21:44 |
lisitsyn | pickle27: so just to clarify | 21:45 |
lisitsyn | the question is whether any of BSS techniques is applicable to out of sample data | 21:45 |
pickle27 | yes | 21:45 |
pickle27 | and my tasks for now and next week are to finish ffsep and get what I've got now into shogun so I can make a PR | 21:47 |
lisitsyn | pickle27: sounds good | 21:47 |
pickle27 | also bit unrelated | 21:47 |
pickle27 | Im still considering coming to the workshop | 21:47 |
pickle27 | trying to decide if I have the time | 21:47 |
lisitsyn | pickle27: are you done with your masters already btw? | 21:48 |
pickle27 | lisitsyn: I sent my supervisor my first draft last week so I am waiting on his edits | 21:48 |
pickle27 | so Im pretty close :) | 21:48 |
lisitsyn | pickle27: if you come I think I'll better give you my time to present your project ;) | 21:49 |
pickle27 | GSoC or Masters? | 21:49 |
lisitsyn | pickle27: gsoc | 21:49 |
pickle27 | cool | 21:49 |
pickle27 | does anyone know how to go about flying cheap last minute lol? | 21:51 |
lisitsyn | no idea | 21:52 |
lisitsyn | :) | 21:52 |
pickle27 | anyways I'm still on the fence I'll probably decide next week | 21:53 |
pickle27 | I think it would be fun though! | 21:53 |
@sonney2k | pickle27, where are you flyiing from? | 22:07 |
pickle27 | Toronto | 22:07 |
@sonney2k | woah! expensive then | 22:07 |
pickle27 | yeah a bit eh! | 22:07 |
@sonney2k | like 1k$ | 22:07 |
pickle27 | yup | 22:08 |
-!- zxtx [~zv@12.231.120.253] has quit [Ping timeout: 276 seconds] | 22:42 | |
-!- zxtx [~zv@12.231.120.253] has joined #shogun | 23:00 | |
-!- zxtx [~zv@12.231.120.253] has quit [Ping timeout: 256 seconds] | 23:05 | |
-!- zxtx [~zv@12.231.120.253] has joined #shogun | 23:18 | |
-!- zxtx [~zv@12.231.120.253] has quit [Ping timeout: 264 seconds] | 23:52 | |
--- Log closed Sun Jun 30 00:00:10 2013 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!