IRC logs of #shogun for Sunday, 2016-07-17

--- Log opened Sun Jul 17 00:00:25 2016
-!- HeikoS [~heiko@host-92-0-162-192.as43234.net] has joined #shogun00:17
-!- mode/#shogun [+o HeikoS] by ChanServ00:17
-!- HeikoS [~heiko@host-92-0-162-192.as43234.net] has quit [Client Quit]00:22
shogun-buildbotbuild #32 of clang - thread analysis is complete: Failure [failed compile]  Build details are at http://buildbot.shogun-toolbox.org/builders/clang%20-%20thread%20analysis/builds/32  blamelist: Wu Lin <yorkerlin@users.noreply.github.com>, Wu Lin <yorker.lin@gmail.com>03:45
shogun-buildbotbuild #30 of clang - undefined behaviour analysis is complete: Failure [failed compile]  Build details are at http://buildbot.shogun-toolbox.org/builders/clang%20-%20undefined%20behaviour%20analysis/builds/30  blamelist: Wu Lin <yorkerlin@users.noreply.github.com>, Wu Lin <yorker.lin@gmail.com>03:49
shogun-buildbotbuild #46 of memleak - valgrind is complete: Failure [failed memory check]  Build details are at http://buildbot.shogun-toolbox.org/builders/memleak%20-%20valgrind/builds/46  blamelist: Wu Lin <yorkerlin@users.noreply.github.com>, Wu Lin <yorker.lin@gmail.com>05:42
shogun-buildbotbuild #1045 of nightly_none is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/nightly_none/builds/1045  blamelist: Wu Lin <yorkerlin@users.noreply.github.com>, Wu Lin <yorker.lin@gmail.com>05:56
-!- mizari [~mizari@95-174-213-100.nts.su] has joined #shogun06:29
shogun-buildbotbuild #1174 of nightly_default is complete: Failure [failed test notebooks]  Build details are at http://buildbot.shogun-toolbox.org/builders/nightly_default/builds/1174  blamelist: Wu Lin <yorkerlin@users.noreply.github.com>, Wu Lin <yorker.lin@gmail.com>06:37
-!- mizari [~mizari@95-174-213-100.nts.su] has quit [Ping timeout: 258 seconds]09:28
-!- mizari [~mizari@95-174-213-100.nts.su] has joined #shogun09:53
-!- sanuj [~sanuj@117.204.250.149] has joined #shogun10:06
-!- sanuj [~sanuj@117.204.250.149] has quit [Read error: Connection reset by peer]10:13
-!- sanuj [~sanuj@117.203.13.246] has joined #shogun10:29
-!- sanuj [~sanuj@117.203.13.246] has quit [Ping timeout: 246 seconds]10:59
-!- sanuj [~sanuj@61.0.30.224] has joined #shogun11:12
-!- sanuj [~sanuj@61.0.30.224] has quit [Ping timeout: 264 seconds]12:00
-!- sanuj [~sanuj@117.220.54.186] has joined #shogun12:12
-!- mizari [~mizari@95-174-213-100.nts.su] has quit [Read error: Connection reset by peer]12:23
-!- HeikoS [~heiko@host-92-0-162-192.as43234.net] has joined #shogun12:37
-!- mode/#shogun [+o HeikoS] by ChanServ12:37
-!- sanuj [~sanuj@117.220.54.186] has quit [Ping timeout: 272 seconds]12:50
-!- sanuj [~sanuj@117.203.21.85] has joined #shogun13:03
sanujHeikoS, yo13:05
@HeikoSsanuj: jojo13:05
sanujwhat's up13:05
@HeikoSsanuj: not too much :)13:06
@HeikoScatching up with things13:06
@HeikoSand you?13:06
sanujrecovering :)13:06
sanujHeikoS, can you see this https://github.com/shogun-toolbox/shogun/issues/332813:06
@HeikoSrecovering from?13:07
sanujcold and viral fever13:07
@HeikoSah I see13:07
@HeikoSget well soon!13:07
sanujyeah13:07
sanuji'm almost well13:07
sanuj:D13:07
@HeikoSoh13:08
@HeikoSI realised I forgot to put them in13:08
@HeikoSlet me push13:08
sanujHeikoS, okay!13:10
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun13:11
shogun-notifier-shogun: Heiko Strathmann :feature/meta_vector_matrix * 1f5dbc8 / examples/meta/generator/targets/cpp.json: https://github.com/shogun-toolbox/shogun/commit/1f5dbc843f070244585911552bbd0d33d8d7ef9913:11
shogun-notifier-shogun: add matrix and vector types for cpp13:11
shogun-notifier-shogun: Heiko Strathmann :feature/meta_vector_matrix * ef3ba8c / examples/meta/generator/targets/cpp.json: https://github.com/shogun-toolbox/shogun/commit/ef3ba8c0f3cee8942fd718ea84b1ace08622473413:11
shogun-notifier-shogun: added long real vector and matrix13:11
shogun-notifier-shogun: Heiko Strathmann :feature/meta_vector_matrix * 1829cf4 / examples/meta/src/tests/matrix_types.sg,examples/meta/src/tests/vector_types.sg: https://github.com/shogun-toolbox/shogun/commit/1829cf4ddb84d17901bed21c7a7e41d9b320971a13:11
shogun-notifier-shogun: meta example tests for vector and matrix types13:11
shogun-notifier-shogun: Heiko Strathmann :feature/meta_vector_matrix * 738612f / examples/meta/src/tests/matrix_types.sg,examples/meta/src/tests/vector_types.sg: https://github.com/shogun-toolbox/shogun/commit/738612febb82e88cc372352e448ef6e9a0e2d0a513:11
shogun-notifier-shogun: not test long real vector for now, travis doesn't like it13:11
sanujHeikoS, btw13:12
sanuji got an intern13:12
sanujhttp://camp.lcsr.jhu.edu/intern/13:12
@HeikoSinternship you mean?13:14
@HeikoSUS13:14
@HeikoScool!13:14
@HeikoSthatll be cool13:14
sanujyeah internship13:14
sanujlet's see13:15
sanuji hope i get visa13:15
sanujHeikoS, it's for 9 months13:15
sanujstarting in september13:16
@HeikoSgreat13:16
@HeikoSis it paid?13:16
sanujHeikoS, they said 1000 usd13:16
sanujapprox13:16
sanuja friend told me not enough to survive there13:17
sanujhaha13:17
@HeikoSpcm?13:17
@HeikoSyeah careful with that13:17
sanujHeikoS, pcm?13:17
@HeikoSper calendar month13:18
sanujoh i see13:18
sanujyes i think13:19
sanujHeikoS, this is also a cool lab mila.umontreal.ca13:19
sanujfor deep learning13:19
sanujyoshua bengio13:19
sanujHeikoS, talking about tags13:20
sanuji think we should go full throttle with the conversion of old classes to  use tags on a feature branch13:20
@HeikoSyeah thats a cool lab13:20
@HeikoSconversion as in?13:20
sanujHeikoS, to use tags parameter framework13:21
sanujinstead of SG_ADD13:21
sanujuse register_param13:21
@HeikoSI think we might want to change SG_ADD?13:21
@HeikoSno?13:21
@HeikoSlisitsyn: ^13:21
sanujlisitsyn, is away13:21
sanujHeikoS, we can remove parameter.h and SG_ADD13:23
@HeikoSsanuj: ah yes13:23
@HeikoSsanuj: but then we have to replace every SG_ADD in shogun13:23
@HeikoSthat is kind of a lot of work ;)13:23
sanuj:D13:23
@HeikoSoh and btw13:23
@HeikoSif we remove Parameter.h13:23
@HeikoSthen our serialization doesnt work anymore13:23
sanujoh13:23
@HeikoSso what I would do for now, is just adding things13:24
@HeikoSand once everything works, we can remove the old stuff13:24
@HeikoSbut if SG_ADD would register parameter to both your and the old framework, that might be a good first step13:24
@HeikoSdid you already try to replace a couple of SG_ADD instances?13:24
@HeikoSdid that work?13:24
@HeikoSI think it might be a good idea to get a feeling for how the build behaves13:24
sanujHeikoS, i didn't try that13:24
sanujyes13:25
@HeikoSso you could send a PR where you replace a few SG_ADDs13:25
@HeikoSlike different categories of SG_ADD we have13:25
sanuji see13:25
@HeikoSthe model selection framework is another issue13:25
sanujbecause we need to start the conversion13:25
@HeikoSwhy you cannot just remove the old thing13:25
@HeikoSyes definitely13:25
@HeikoSwhat about you try to hack SG_ADD a bit=?13:25
sanujhaha13:25
sanujwe can have SG_ADD13:26
@HeikoSor just add a few register_params13:26
sanujand register_param separately13:26
@HeikoSunder the corresponding SG_AFF13:26
@HeikoSADD13:26
sanujyeah13:26
@HeikoSyeah that is a good way to start testing it13:26
sanujthat would be easy to test if things work13:26
@HeikoSBUT13:26
@HeikoSyou want to change SG_ADD to do that13:26
@HeikoSeventually13:26
sanujHeikoS, but at the end if we only have register_param then it doesn't make sense to use SG_ADD13:27
@HeikoSsanuj: SG_ADD can just be a wrapper for that13:27
sanujit would be an alias13:27
@HeikoSyes13:28
sanujbut why do we need an alias13:28
@HeikoSthis way we dont need to touch all these files13:28
sanujohkay13:28
@HeikoSbecause we are lazy13:28
sanuji see13:28
sanujhaha13:28
@HeikoSbecause we want to get our effects with minimal changes13:28
@HeikoSevery file you change opens some possibility for doing it wrong13:28
sanujyeah13:28
@HeikoSso I rather just touch one file13:28
@HeikoSi.e. changing SG_ADD13:28
@HeikoSin one place, and then all classes use tags13:29
sanujlisitsyn was saying it might be tricky to hack SG_ADD13:29
sanujwill have to see for myself13:29
-!- sanuj [~sanuj@117.203.21.85] has quit [Ping timeout: 246 seconds]13:34
shogun-notifier-shogun-data: Saurabh7 :master * a042da0 / testsuite/meta/multiclass_classifier/multiclass_ecoc_random.dat: https://github.com/shogun-toolbox/shogun-data/commit/a042da0da3c13bd3148d1ff60ad05f79bc53931813:47
shogun-notifier-shogun-data: update data for liblinear13:47
shogun-notifier-shogun-data: Heiko Strathmann :master * 20317d4 / testsuite/meta/multiclass_classifier/multiclass_ecoc_random.dat: https://github.com/shogun-toolbox/shogun-data/commit/20317d458e71657fd494892735b59e8740a0048e13:47
shogun-notifier-shogun-data: Merge pull request #117 from Saurabh7/newlibl13:47
shogun-notifier-shogun-data:13:47
shogun-notifier-shogun-data: update data for liblinear13:47
-!- sanuj [~sanuj@61.0.30.72] has joined #shogun13:47
@HeikoSsanuj:  see email13:50
sanujyeah13:50
sanujHeikoS, which classes shall i touch?13:50
-!- travis-ci [~travis-ci@ec2-54-211-114-135.compute-1.amazonaws.com] has joined #shogun13:56
travis-ciit's Heiko Strathmann's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/14533440213:56
-!- travis-ci [~travis-ci@ec2-54-211-114-135.compute-1.amazonaws.com] has left #shogun []13:56
@HeikoSsanuj: choose any13:57
@HeikoSdo one add per category type13:57
@HeikoSthat is: one for scalar13:57
@HeikoSfloat, int13:57
@HeikoSone for vector13:57
@HeikoSone for matrix13:57
@HeikoSetc13:57
@HeikoSand then we see what travis says13:58
@HeikoSsanuj: btw what about the plugin pr?13:59
@HeikoSwhere did that code come from?13:59
sanujohky14:00
sanujHeikoS, from aer14:00
@HeikoSI see14:00
sanujit needs some modifications though14:00
sanujwill update it14:00
sanujHeikoS, for plugins14:01
@HeikoSokok14:01
sanujusers will use createX()14:01
@HeikoSI see14:01
sanujnot sure how to implement it14:01
@HeikoSsanuj: I hope we can push this to some nice point during GSoC still14:01
@HeikoSbut it will definitely needs some love even after14:01
@HeikoShope youre up for that :D14:01
sanujyeah i know14:01
sanujyes i'm up for that14:01
sanuj:D14:01
sanujHeikoS, i'll make a dir in shogun root called plugin14:02
sanujnew plugins will be added in that dir14:02
sanujand cmake will just build every dir in the plugins dir into a plugin.so14:02
@HeikoSsanuj: is that discussed with sergey?14:04
@HeikoSI think we also said something like every class will be a plugin?14:05
@HeikoSor are plugins defined via an "include" list of classes14:05
@HeikoS?14:05
-!- HeikoS [~heiko@host-92-0-162-192.as43234.net] has quit [Quit: Leaving.]14:09
-!- sanuj [~sanuj@61.0.30.72] has quit [Ping timeout: 272 seconds]14:21
-!- HeikoS [~heiko@host-92-0-162-192.as43234.net] has joined #shogun14:53
-!- mode/#shogun [+o HeikoS] by ChanServ14:53
shogun-notifier-shogun: OXPHOS :feature/linalg_refactor * 8aa71f2 / / (8 files): https://github.com/shogun-toolbox/shogun/commit/8aa71f249203497e3fd58167df904b67d3e5ad3015:08
shogun-notifier-shogun: linalg-sgmatrix-memory transfer15:08
shogun-notifier-shogun: OXPHOS :feature/linalg_refactor * c87084d / / (5 files): https://github.com/shogun-toolbox/shogun/commit/c87084d986c2ae642e85d15cb99de7235b621a6c15:08
shogun-notifier-shogun: fix style15:08
shogun-notifier-shogun: Heiko Strathmann :feature/linalg_refactor * d269956 / / (8 files): https://github.com/shogun-toolbox/shogun/commit/d26995630dbb401da914b49784cd0420fb4f863015:08
shogun-notifier-shogun: Merge pull request #3348 from OXPHOS/linalg_refactor_matrix15:08
shogun-notifier-shogun:15:08
shogun-notifier-shogun: LinalgRefactor - matrix15:08
-!- OXPHOS [c68f0c0c@gateway/web/freenode/ip.198.143.12.12] has joined #shogun15:16
-!- travis-ci [~travis-ci@ec2-54-211-114-135.compute-1.amazonaws.com] has joined #shogun15:53
travis-ciit's Heiko Strathmann's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/14534578115:53
-!- travis-ci [~travis-ci@ec2-54-211-114-135.compute-1.amazonaws.com] has left #shogun []15:53
shogun-notifier-shogun: Heiko Strathmann :feature/meta_vector_matrix * 6717a15 / examples/meta/src/tests/matrix_types.sg,examples/meta/src/tests/vector_types.sg: https://github.com/shogun-toolbox/shogun/commit/6717a153d0c967a9bcd377b0fa28a3dedfd76a1816:52
shogun-notifier-shogun: use correct syntax for booleans16:52
-!- HeikoS [~heiko@host-92-0-162-192.as43234.net] has quit [Quit: Leaving.]17:06
-!- sanuj [~sanuj@117.203.8.213] has joined #shogun17:35
-!- travis-ci [~travis-ci@ec2-54-211-114-135.compute-1.amazonaws.com] has joined #shogun17:44
travis-ciit's Heiko Strathmann's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/14535843517:44
-!- travis-ci [~travis-ci@ec2-54-211-114-135.compute-1.amazonaws.com] has left #shogun []17:44
-!- sanuj [~sanuj@117.203.8.213] has quit [Ping timeout: 240 seconds]17:56
-!- OXPHOS [c68f0c0c@gateway/web/freenode/ip.198.143.12.12] has quit [Quit: Page closed]18:20
-!- sanuj [~sanuj@117.203.8.213] has joined #shogun19:04
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout]19:52
-!- sanuj [~sanuj@117.203.8.213] has quit [Ping timeout: 240 seconds]20:21
-!- Netsplit *.net <-> *.split quits: sonney2k_20:56
-!- sonney2k [~shogun@7nn.de] has joined #shogun21:15
--- Log closed Mon Jul 18 00:00:27 2016

Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!