Open in new window / Try shogun cloud
--- Log opened Thu Sep 19 00:00:05 2013
-!- FSCV [~FSCV@66.187.72.219] has quit [Quit: Leaving]00:08
-!- travis-ci [~travis-ci@ec2-54-211-149-66.compute-1.amazonaws.com] has joined #shogun00:20
travis-ci[travis-ci] it's Soeren Sonnenburg'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/1152421700:20
-!- travis-ci [~travis-ci@ec2-54-211-149-66.compute-1.amazonaws.com] has left #shogun []00:20
-!- zxtx [~zv@ip-64-134-197-118.public.wayport.net] has joined #shogun00:28
shogun-buildbotbuild #14 of clang34 - static analysis is complete: Failure [failed analyse]  Build details are at http://buildbot.shogun-toolbox.org/builders/clang34%20-%20static%20analysis/builds/14  blamelist: Soeren Sonnenburg <sonne@debian.org>00:31
-!- Heiko__ [05453f70@gateway/web/freenode/ip.5.69.63.112] has quit [Quit: Page closed]00:44
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout]01:08
-!- zxtx [~zv@ip-64-134-197-118.public.wayport.net] has quit [Ping timeout: 260 seconds]01:12
-!- zxtx [~zv@ip-64-134-197-230.public.wayport.net] has joined #shogun01:45
-!- van51 [~van51@athedsl-409794.home.otenet.gr] has quit [Quit: Leaving.]02:48
-!- sonne|osx_ [~sonne@f053037175.adsl.alicedsl.de] has joined #shogun03:05
shogun-buildbotbuild #471 of nightly_none is complete: Failure [failed compile]  Build details are at http://buildbot.shogun-toolbox.org/builders/nightly_none/builds/47103:06
-!- sonne|osx [~sonne@f053045233.adsl.alicedsl.de] has quit [Ping timeout: 240 seconds]03:07
-!- sonne|osx_ is now known as sonne|osx03:07
shogun-buildbotbuild #459 of nightly_all is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/nightly_all/builds/45903:32
-!- zxtx [~zv@ip-64-134-197-230.public.wayport.net] has quit [Ping timeout: 248 seconds]04:53
-!- zxtx [~zv@ip-64-134-197-230.public.wayport.net] has joined #shogun04:58
-!- zxtx [~zv@ip-64-134-197-230.public.wayport.net] has quit [Ping timeout: 248 seconds]05:29
-!- zxtx [~zv@ip-64-134-197-230.public.wayport.net] has joined #shogun05:44
-!- zxtx [~zv@ip-64-134-197-230.public.wayport.net] has quit [Ping timeout: 245 seconds]06:01
-!- zxtx [~zv@ip-64-134-197-230.public.wayport.net] has joined #shogun06:04
-!- foulwall [~zhengyang@114.255.40.22] has joined #shogun06:27
-!- sonne|osx [~sonne@f053037175.adsl.alicedsl.de] has quit [Quit: sonne|osx]06:27
-!- zxtx [~zv@ip-64-134-197-230.public.wayport.net] has quit [Ping timeout: 248 seconds]07:11
-!- zxtx [~zv@149.165.244.2] has joined #shogun07:23
-!- besser82 [~besser82@fedora/besser82] has joined #shogun07:38
-!- lisitsyn [~lisitsyn@fb2-lo1.global63.net] has quit [Quit: Leaving.]08:07
-!- hushell [~hushell@c-98-232-178-161.hsd1.or.comcast.net] has joined #shogun08:13
-!- sonne|osx [~sonne@82.113.98.161] has joined #shogun08:17
sonne|osxshogun-buildbot: force build --branch=develop "rpm1 - libshogun"08:21
shogun-buildbotbuild #1214 forced08:21
shogun-buildbotI'll give a shout when the build finishes08:21
sonne|osxshogun-buildbot: force build --branch=develop "precise - libshogun"08:22
shogun-buildbotbuild forced [ETA 22m07s]08:22
shogun-buildbotI'll give a shout when the build finishes08:22
-!- sonne|osx [~sonne@82.113.98.161] has quit [Quit: sonne|osx]08:35
shogun-buildbotbuild #100 of precise - libshogun is complete: Success [build successful]  Build details are at http://buildbot.shogun-toolbox.org/builders/precise%20-%20libshogun/builds/10008:38
wikingmmm we have a lot of problems with the 2 nightly targets :S08:38
shogun-buildbotbuild #1214 of rpm1 - libshogun is complete: Success [build successful]  Build details are at http://buildbot.shogun-toolbox.org/builders/rpm1%20-%20libshogun/builds/121408:53
sonne|workwiking:08:57
sonne|work(09:35:38 PM) sonney2k: wiking, you might consider turning of integration tests on travis IMHO it is sufficient that we run this stuff on the buildbots08:57
sonne|work(09:36:21 PM) sonney2k: wiking, I mean travis is good to get a rough picture if sth works and IMHO that is all we want at this stage leaving detailed analysis to our own BB08:57
sonne|work^ any comments?08:58
sonne|workwiking: /home/buildslave/nightly_none/build/tests/unit/multiclass/BaggingMachine_unittest.cc:25:45: error: 'class testing::NiceMock<shogun::MockCFeatures>' has no member named 'ref'08:59
sonne|work on nightly_none08:59
wikingyesh i know08:59
wikingsonne|work: we can disable integration on travis08:59
wikingthat's no problem08:59
wikingcurrently i'm fixing leaking errors09:01
wikingi think parametermap or such is leaking...09:15
-!- iglesiasg [~iglesias@2001:6b0:1:1da0:247a:a4fc:e52c:8133] has joined #shogun09:53
-!- mode/#shogun [+o iglesiasg] by ChanServ09:53
@iglesiasggood morning!09:54
-!- lisitsyn [~lisitsin@mxs.kg.ru] has joined #shogun09:59
wikingiglesiasg: yo! need help10:06
wikingiglesiasg: StructuredLabels unit test is indicating a lot of leak :(10:06
@iglesiasgwiking, shit, that's not good10:06
@iglesiasgwiking, do you have the trace?10:06
wikingiglesiasg: i can copy paste10:06
wikingjust a sec10:06
wikingiglesiasg: http://pastebin.com/38xthspY10:08
@iglesiasgwiking, there are also leaks coming from init_shogun_with_defaults10:09
@iglesiasgwiking, that looks weird to me10:09
wikingiglesiasg: dont bother about that now10:09
wikingiglesiasg: the lot of parameterMap leak is just beecause there's an objec that is not being freed10:10
@iglesiasgwiking, it leaks because the test is wrong I think10:12
@iglesiasgwiking, https://github.com/shogun-toolbox/shogun/blob/develop/tests/unit/labels/StructuredLabels_unittest.cc#L4910:12
@iglesiasgget_label returns an object, but there is no SG_UNREF for them10:12
wikingiglesiasg: fixit plz10:13
wikingif u can10:13
@iglesiasgthoralf, ^10:13
@iglesiasgwiking, yep. I can do that right away10:13
wikingcool thanx heaps10:13
@iglesiasgwiking, will need to compile in debug to double check so will take a few minutes..10:16
wikingyeps... thnx10:16
wikingi'm fixing some other shit in the meanwhile..10:16
@iglesiasgwiking, it should be ready, https://github.com/shogun-toolbox/shogun/pull/162710:29
wikingok i'll pull it10:30
wikingas soon as i have my fixes ready as well10:30
@iglesiasgall right10:30
wikingthnx a lot10:30
@iglesiasgno fixes were left here10:30
@iglesiasgso actually the ones in init_shogun_with_defaults were also a consequence of these ones10:31
@iglesiasgsorry10:31
@iglesiasgno leaks were left10:31
sonne|workwiking: I have a hard time debugging cyg1 and other crashers. Could we when a test fails run it with the gdb script in .src again to get a proper bt?10:41
wikingsonne|work: i'll try to see if i can do something about that10:42
@iglesiasgspeaking of which, cygwin is still the only way to install Shogun in Windows, right?10:47
@iglesiasgwiking, sonne|work, ^10:47
sonne|workiglesiasg: yes10:48
@iglesiasgjust wondering since, IIRC, something was mentioned about native in Windows when CMake came10:48
sonne|workiglesiasg: yeah sth to work on for later10:48
@iglesiasgI see10:48
wikingiglesiasg: i've applied some changes10:49
wikingiglesiasg: but more changes are required10:49
wikingas msvc is not the most standard compiler10:49
@iglesiasgaham I understand10:50
wikingok finally i've found a valgrind error that is actually not a unit test leak but an actual leak in the library10:50
@iglesiasgwiking, StructuredLabels still?10:53
wikingiglesiasg: nono10:53
wikingwhen's the meeting today?10:56
@iglesiasgin one hour approx10:56
lisitsyniglesiasg: we need some massive thing about __declspec(dllexport)10:58
@iglesiasglisitsyn, for windows stuff?10:59
lisitsyniglesiasg: yeah10:59
@iglesiasglisitsyn, I see10:59
lisitsyniglesiasg: otherwise all the classes are not visible from .dll10:59
@iglesiasgI am not really interested in it myself (windows sucks :P) but I know a guy who might be interested10:59
lisitsynI don't know if there is other way10:59
@iglesiasgalthough it is true that this could help us reach more people11:00
lisitsynno we definitely need it11:00
lisitsynI don't want to hear about windows sucks ;)11:00
@iglesiasglisitsyn, lol why not?11:00
lisitsynjust like most of the people are idiots but we have to live with that11:01
@iglesiasgit is the truth!!11:01
@iglesiasghahaha11:01
lisitsynthe majority is windows so we have to live with that too11:01
@iglesiasgI like your pragmatism11:01
wikinglisitsyn: MALSAR_unittest is leaking like hell :)11:03
@iglesiasglisitsyn, anyway, our users are developers as well. Do you still think windows is majority in that too?11:03
@iglesiasglisitsyn, I would say people using mac os or linux for development are more11:04
lisitsyniglesiasg: well remember that detection thing I develop11:04
@iglesiasgyep11:04
@iglesiasgwindows people the ones who ask you to do that?11:04
lisitsyniglesiasg: it has to be windows11:05
@iglesiasgbut run on windows or actually developed in windows?11:05
lisitsynwiking: I am deeply saddened ;)11:05
lisitsyniglesiasg: well run on windows11:05
@iglesiasglisitsyn, that's another thing :)11:05
lisitsyniglesiasg: of course I can train on linux11:05
wikinglisitsyn: plz fix11:05
lisitsynwiking: yeah let me try11:05
wikingok here's the first batch of leakfixes11:10
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun11:11
shogun-notifier-shogun: Viktor Gal :develop * c6f4ca7 / tests/unit/ (8 files): https://github.com/shogun-toolbox/shogun/commit/c6f4ca70c5cc278e15218f326355f8bf5d264b8b11:11
shogun-notifier-shogun: Fix a bunch of leaking unit tests11:11
shogun-notifier-shogun: and disable ref calls when reference counting is disabled11:11
wikingshogun-buildbot: force build --branch=develop 'nightly_none'11:12
shogun-buildbotbuild #472 forced11:12
shogun-buildbotI'll give a shout when the build finishes11:12
wikingi'd be great to have an xslt for this: http://buildbot.shogun-toolbox.org/memcheck/20130918-2140/DynamicAnalysis.xml11:14
-!- foulwall [~zhengyang@114.255.40.22] has quit [Quit: leaving]11:14
shogun-buildbotbuild #472 of nightly_none is complete: Failure [failed compile]  Build details are at http://buildbot.shogun-toolbox.org/builders/nightly_none/builds/47211:15
-!- foulwall [~AndChat63@219.141.159.37] has joined #shogun11:16
thoralfiglesiasg: Thanks for fixing.11:20
shogun-notifier-shogun: Fernando Iglesias :develop * d952d8d / tests/unit/labels/StructuredLabels_unittest.cc: https://github.com/shogun-toolbox/shogun/commit/d952d8de4e0901f974a0c7ec3f3e84107cffe85311:20
shogun-notifier-shogun: Fix memory leaks and indentation in structured labels unit test11:20
shogun-notifier-shogun: Viktor Gal :develop * 8567050 / tests/unit/labels/StructuredLabels_unittest.cc: https://github.com/shogun-toolbox/shogun/commit/85670508cde6cae94502602773e1d0415818753c11:20
shogun-notifier-shogun: Merge pull request #1627 from iglesias/fix/leak/structured_labels_unittest11:20
shogun-notifier-shogun:11:20
shogun-notifier-shogun: Fix memory leaks and indentation in structured labels unit test11:20
lisitsynwiking: https://github.com/lisitsyn/tapkee/blob/master/test/valgrind_tests_transformer.xslt I had sth like that11:20
thoralfiglesiasg: After having a working regression test, I "forgot" to care about SG_UNREF ;)11:20
@iglesiasgthoralf, no problem11:20
@iglesiasghehe11:20
-!- foulwall [~AndChat63@219.141.159.37] has quit [Read error: Connection reset by peer]11:23
shogun-buildbotbuild #2213 of deb1 - libshogun is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb1%20-%20libshogun/builds/2213  blamelist: Viktor Gal <vigsterkr@gmail.com>, Fernando Iglesias <fernando.iglesiasg@gmail.com>11:24
lisitsynwiking:11:27
lisitsyn-- Looking for clapack_dpotrf in ATLAS_LAPACK-NOTFOUND11:27
lisitsynCMake Error: The following variables are used in this project, but they are set to NOTFOUND.11:27
lisitsynPlease set them or make sure they are set and tested correctly in the CMake files:11:27
lisitsynATLAS_LAPACK11:27
lisitsyn    linked by target "cmTryCompileExec2966321832" in directory /home/blackburn/shogun/build/CMakeFiles/CMakeTmp11:27
lisitsynwiking: have you seen that before?11:27
wikingno not really11:36
wikingshogun-buildbot: force build --branch=develop 'deb1 - libshogun'11:36
shogun-buildbotbuild #2214 forced11:36
shogun-buildbotI'll give a shout when the build finishes11:36
shogun-buildbotbuild #2214 of deb1 - libshogun is complete: Success [build successful]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb1%20-%20libshogun/builds/221411:38
-!- az_de [82954e22@gateway/web/freenode/ip.130.149.78.34] has joined #shogun11:38
-!- foulwall [~AndChat63@219.141.159.55] has joined #shogun11:39
-!- thoralf [~thoralf@enki.zib.de] has quit [Quit: Konversation terminated!]11:46
lisitsynhey12:00
wikingyo12:00
lisitsynanybody? :D12:01
lisitsynsonne|work: around?12:01
hushellHello everyone12:01
wikingmeeting time?12:01
lisitsynyeah12:01
wikingok12:01
wikingwhat's the agenda for today?12:02
@iglesiasghello hello12:02
lisitsynokay so agenda would be gsoc + release12:02
@iglesiasgit sounds good12:02
lisitsynaz_de: around?12:02
* az_de yes, hi all12:03
lisitsyngood12:03
lisitsynit is not good time for kevin so good you are here12:03
lisitsyngsomix: foulwall: Soeren guys around?12:03
sonne|worklisitsyn: hi there12:03
lisitsynoh sorry I forgot to send a reminder12:03
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has joined #shogun12:04
lisitsynokay nice votjakovr here12:04
-!- thoralf [~thoralf@enki.zib.de] has joined #shogun12:04
lisitsynlets wait for a bit, Heiko should be here iirc12:04
-!- Cheng [73406f11@gateway/web/freenode/ip.115.64.111.17] has joined #shogun12:04
lisitsynhey Cheng thanks for joining us12:04
ChengHi all12:04
-!- HeikoS [05453f70@gateway/web/freenode/ip.5.69.63.112] has joined #shogun12:05
@iglesiasghi Cheng!12:05
HeikoShello! sorry for the delay12:05
foulwallhi12:05
lisitsynHeikoS: no problem we didn't start anyway12:05
lisitsyn:)12:05
votjakovrhi everyone!12:05
lisitsynso who's missing? Georg said he will be a bit late12:05
lisitsynand Kevin probably can't make it12:05
lisitsynHeikoS: lambday is out too, right?12:06
HeikoSlisitsyn: I dont know, we did not say anything12:06
HeikoSlet me check my mail12:06
sonne|workapologies but I will be offline for ~10 mins - please proceed without me12:06
lisitsynsonne|work: alright12:06
lisitsyngsomix: foulwall: 2nd ping ;)12:07
gsomixlisitsyn, pong.12:07
lisitsyngood12:07
foulwallpong12:07
lisitsynokay then lets start12:07
gsomixforgot about the meeting.12:07
lisitsynI didn't prepare detailed agenda actually12:07
lisitsynbut you all know that we have pencil down approacing12:08
-!- lambday [67157e37@gateway/web/freenode/ip.103.21.126.55] has joined #shogun12:08
lisitsynand actually soft pencil down is here12:08
lisitsynoh lambday good time to join us12:08
lisitsyn;)12:08
lambdaylisitsyn: sorry I am late12:08
HeikoSlambday: hi!12:08
lambdayHeikoS: hey :)12:08
lisitsynnp lambday thanks for joining us12:08
lisitsynlets just talk about each project one by one12:09
lisitsynand discuss what have been done12:09
lisitsynand what's you think is still left12:09
lisitsynare you ok with that?12:09
HeikoSsure12:09
lisitsynand may be anybody would like to start?12:09
HeikoSwe can12:10
HeikoSlambday: ?12:10
lisitsynjust one note12:10
lisitsynplease put an url with ipython notebook12:10
lambdayall the parts of my project has been implemented...12:10
shogun-buildbotbuild #1497 of cyg1 - libshogun is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/cyg1%20-%20libshogun/builds/1497  blamelist: Viktor Gal <viktor.gal@maeth.com>12:10
HeikoSok let me go ahead12:10
@iglesiasg(for some reason, I have very large  lag, so I might take time to answer -- sorry for that)12:11
HeikoSso Rahul started with the notebook, the current draft is here12:11
HeikoShttp://nbviewer.ipython.org/659199412:11
HeikoSStill some things to improve and polish, but its getting there12:11
HeikoScurrently there is an issue wie complex typemaps in swig, which he is at12:11
HeikoSpython examples not there yet but are easy since its all subparts from the notebook12:11
HeikoSlambday: what about the code-review? did you have a look on that?12:12
hushellHeikoS: I just sent an email to lambday about the peer review12:12
lambdayHeikoS: hushell mailed me regarding this... I am thinking of doing it in the weekends12:12
hushelllambday: hi!12:12
thoralfWill something in the dev meeting be non-gsoc related?12:12
lambdayhushell: hey man! just replied to your mail.. :)12:13
wikingthoralf: on the end12:13
lisitsynthoralf: yeah lets talk about it in the end12:13
wikingthoralf: so take a coffee ;)12:13
hushelllambday: good! :)12:13
-!- georg_zeller [~zeller@embln.embl.de] has joined #shogun12:13
HeikoSok so code review will be done last minute then, guys, please start this before Saturday and take some time12:13
lisitsynhey georg_zeller12:13
georg_zellerhi and sorry for being late12:13
thoralfOkay, please ping me.  My IRC client will wake me up then. :)12:13
lisitsynHeikoS: do we have review pairs anywhere around?12:14
lisitsynlambday is reviewing ..?12:14
lambdayHeikoS: alright...12:14
HeikoSpeople should know, if not I will find them12:14
lambdaylisitsyn: hushell's12:14
lisitsyngood12:14
-!- sonne|osx [~sonne@89.204.135.239] has joined #shogun12:14
@iglesiasgfoulwall, you are my couple :P12:14
lisitsynokay so lambday12:14
HeikoSvan51 - gsomix hushell - lambday pickle27 - votjakovr iglesias - foulwall12:14
lisitsynwhat you might need to focus is polish ipython notebook12:15
lisitsynI don't think we need python examples right now12:15
wikinglambday: and ping me on the end of the meeting plz12:15
HeikoSehm the split is not the minus but the " "12:15
lambdaylisitsyn: yes alright12:15
lambdaywiking: sure :)12:15
HeikoSyeah notebook is more important than python examples12:15
lisitsynnotebook seems more important12:15
wikingnotebook++12:15
HeikoSbut writing examples takes just one day both libshogun and notebook,12:15
lambdayplus python examples won't take more than an hour once the notebook is complete12:16
lisitsynwell12:16
lisitsynyou need to resolve that issue12:16
HeikoS@all In general, notebooks should explain more things, keep in mind that people reading them might not know what you are doing there :)12:16
lisitsynand code-review12:16
HeikoSyep agreed!12:16
lambdaylisitsyn: yeah swig is giving me real pain :'(12:16
wikinglambday: just ask here12:17
HeikoSlambday: if this takes too long, maybe postpone the complex stuff and finish the rest first12:17
wikinglambday: we've fixed the other day the bug we had with r modular...12:17
HeikoSlambday: its just a small subset of your work12:17
lambdaywiking: alright let's talk about the swig thing in the end..12:17
lisitsynokay so lambday's project is ok and we have no real issues12:17
lambdayHeikoS: yeah, so kept that part incomplete12:17
lisitsynHeikoS: want to continue with votjakovr?12:17
lambdaywiking: I saw, it was renaming afterall :)12:17
sonne|osxlambday: just tell us what you need / where you have issues12:17
HeikoSlisitsyn: okay12:18
sonne|osxlambday: we will get it to work.12:18
wikinglambday: indeed... but lets get back to this on the end... anyhow great work!12:18
lambdaysonne|osx: I need to make our complex64_t (std::complex<double>) with numpy complex6412:18
lambdaywiking: thanks man :)12:18
HeikoSvotjakovr: wanna give us the update?12:18
sonne|osxlambday: should be easy to do :D12:18
sonne|osxlambday: but later12:19
lambdayalright12:19
lambday:)12:19
lisitsynoh okay so while votjakovr is afk12:19
lisitsynmay be anyone else?12:19
HeikoSvotjakovr: sleeping? :)12:19
HeikoSok so let me do this12:19
votjakovrHeikoS: oops :)12:19
-!- travis-ci [~travis-ci@ec2-54-211-149-66.compute-1.amazonaws.com] has joined #shogun12:19
travis-ci[travis-ci] it's Viktor Gal'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/1154312112:19
-!- travis-ci [~travis-ci@ec2-54-211-149-66.compute-1.amazonaws.com] has left #shogun []12:19
HeikoSah12:19
lisitsynokay we have some issues here right?12:20
votjakovrHeikoS: unfortunately my project is not ready yet :(12:20
votjakovrmodel selection part is missing12:20
sonne|osxvotjakovr: and apart from that?12:20
lisitsynwhat can we do about it?12:21
HeikoSvotjakovr: if you cannot finish this today or latest tomorrow, we will have to postpone it, it is very important that the notebook and some examples are there12:21
votjakovri'll finish it today12:21
HeikoSI mean, already you can write about a lot of things12:21
sonne|osxHeikoS, votjakovr but the rest is 'done' ?12:21
HeikoSsonne|osx: yeah if this is merged, the we can polish12:21
lisitsynlets just decide what is missing to consider the project done12:22
HeikoSsonne|osx: wanted to do multiclass originally, but that is unrealistic12:22
HeikoSlisitsyn: model-selection, notebook, examples12:22
lisitsyncan we postpone examples already?12:22
HeikoSlisitsyn: they are last in the list, but I think they are important12:22
HeikoSbut the notebook is more important12:23
sonne|osxHeikoS: can you write some GP notebooks? IIRC you already have them?12:23
lisitsynI mean we are in difficult timing12:23
HeikoSvotjakovr: in fact I cal help you a bit with the notebook if that helps12:23
sonne|osxthen votjakovr can do the MS12:23
HeikoSsonne|osx: yeah, I will start one, good idea12:23
HeikoSvotjakovr: do you agree? then you can focus on the modelselection12:23
HeikoSvotjakovr: but please hurry up a lot!12:23
lisitsynand code-review12:23
sonne|osxvotjakovr: but the agreement would be that you finish it then12:24
-!- georg_zeller [~zeller@embln.embl.de] has quit [Quit: georg_zeller]12:24
votjakovractually i've started both of them (model-selection, notebook, examples)12:24
-!- georg_zeller [~zeller@embln.embl.de] has joined #shogun12:24
HeikoSvotjakovr: pls send your current state then and I will push them a bit ok?12:24
lisitsynvotjakovr: lets just assure both HeikoS and you mean the same12:24
sonne|osxvotjakovr: so what do you think? When would you finish?12:24
votjakovrHeikoS: i'll try to finish model-selection today12:25
lisitsynwhat's the worst estimate of that part?12:25
HeikoSvotjakovr: that would be good!12:25
votjakovrexamples toaday/tomorrow12:25
votjakovrHeikoS: notebook takes some time12:25
sonne|osxHeikoS: I would suggest to see how it goes. If MS is done by today or tomorrow everything will work out nicely.12:25
HeikoSvotjakovr: notebook first! Then examples12:26
HeikoSexamples are just demonstration of API12:26
sonne|osxvotjakovr: so do notebooks a little later and make sure to get some help from HeikoS12:26
HeikoSnotebook is about what is going on behind the scences and what you have done12:26
HeikoSyep, votjakovr pls send the current state or even send a PR with it, then I can update12:26
HeikoSas soon as you have *something* for the notebook, send it12:27
votjakovrHeikoS: ok, i'll send little bit later12:27
lisitsynalright so I guess that's the plan?12:27
sonne|osxlooks like12:27
lisitsynvotjakovr: please keep HeikoS updated12:27
votjakovrlisitsyn: ok12:27
lisitsynso if you are stucked or whatever just let him know12:28
sonne|osxlisitsyn: next?12:28
@iglesiasgI am ready!12:28
lisitsyniglesiasg: georg_zeller please continue then12:28
sonne|osx(it is already 12:28)12:28
@iglesiasgThere is this notebook http://nbviewer.ipython.org/6576096 that currently contains a first section explaining what LMNN is, and a second section applying LMNN to a real data set illustrating feature selection.12:28
@iglesiasgI have to update this second part to use Shogun's cross validation, instead of how it is done at the moment.12:28
HeikoSapologies, but I have to be AFK for a while, checking back later for non-gsoc stuff12:28
lisitsynsonne|osx: yeah lets speed up12:28
@iglesiasgThen, I will add another data set in the notebook where I am comparing kNN and LMNN in a multiclass classification task. I will normalize and whiten the data for this task as we have discussed in the mailing list.12:28
lisitsynyour notebook looks quite nice12:29
@iglesiasglisitsyn, thanks! :) I think it is pretty standard actually12:29
lisitsyngeorg_zeller: iglesiasg: so what do you think is left here?12:29
@iglesiasgoh! And I forgot, I have to update the first part too, showing some neighbourhood graphs12:29
georg_zellerwe have discussed a few updates on the notebook but those shouldn't take long12:30
@iglesiasgso basically updating the notebook with these things12:30
lisitsynah yes12:30
lisitsynapart from that?12:30
georg_zellermaybe 2-3 days I'd guess?12:30
@iglesiasgyep12:30
@iglesiasggeorg_zeller, yes, I think so12:30
@iglesiasgthen I have to do the code review12:30
lisitsynso nothing apart from that right?12:30
georg_zellercoding is done as far as I'm concerned12:30
lisitsynyeah and the code review12:30
@iglesiasgbut I hope to get that done in the weekend12:30
georg_zellerwhat about examples?12:30
sonne|osxiglesiasg: I wouldn't mind if you could help us get the release polished12:30
lisitsynokay it looks healthy12:30
@iglesiasggeorg_zeller, there are libshogun, octave and python examples already12:31
sonne|osxiglesiasg: you know website, bug and warning fixes etc12:31
@iglesiasgjust simple stuff12:31
sonne|osxhelping out others...12:31
georg_zellerthen fine from my side12:31
wikingsonne|osx: none gsoc related stuff on the end of this meeting :)12:31
@iglesiasgsonne|osx, sure, once GSoC period is over I get to this12:31
lisitsynokay thanks a lot georg_zeller iglesiasg your project is totally ok :)12:31
wikingsonne|osx: unfortunately lisitsyn mixed up monthly dev meeting with gsoc meeting...12:31
lisitsynnext is sonne|work foulwall Cheng please12:31
lisitsynhaha12:31
georg_zellerwe talked about a few interesting applications but better for the time after GSoC12:32
lisitsynwell12:32
lisitsyngeorg_zeller: yeah it would be nice if you continue12:32
sonne|osxwiking: no no that is not the issue12:32
georg_zellerwould be great for me too12:32
lisitsynwiking: lets say I combined meetings12:32
sonne|osxwiking: we need help to get the release done so anyone who thinks his project looks OK please help us!12:32
lisitsynotherwise I seem stupid hahah12:33
wikinglisitsyn: no u fucked it up ;)12:33
lisitsynwiking: tss12:33
wikingokey continue12:33
lisitsynsonne|osx: foulwall bitte12:33
wikingCheng foulwall sonne|osx12:33
wikingspeed it up12:33
wikingthe demo site just looks great12:33
sonne|osxfoulwall: ?12:33
lisitsyngsomix?12:33
gsomixyep12:33
lisitsynsonne|osx: gsomix bitte12:34
wikinggsomix: let's talk12:34
gsomixOK. I'm finishing protobuf support for shogun right now. For now we have simple IO system (buffered read, simple parsers), csv and libsvm readers.12:34
lisitsynschnelle12:34
lisitsyn:D12:34
gsomixAfter protobuf related works I'll start writing ipython notebook and nice examples.12:34
gsomixUnfortunately support of matlab m-files and some `exotic' formats is out of schedule (*bad guy). But I'll stay with shogun for this work.12:34
sonne|osxSo whoever has a nice idea for a demo ask foulwall12:34
gsomixMaybe Soeren want to say sth?12:34
lisitsynwhat's left with protobuf?12:34
sonne|osxe.g. van51's language detection demo is now in applications12:34
wikinggsomix: yeah those io stuff are great you've implemented... looking forward for the protobuf stuff12:34
sonne|osxhttp://foulwall.com:8000/application/language_detect/12:35
sonne|osxif you can think of sth like this for your project and need help from foulwall now would be the time12:35
lisitsynsonne|osx: what's left?12:35
sonne|osxwell the splice site recognizer/arts detector12:35
sonne|osxbut it really is us that need to do sth here IMHO12:36
lisitsynthat's out, right?12:36
gsomixlisitsyn, today I'll send PR with skeleton. And then we need and protobuf support for many data types...12:36
gsomix*need to add12:36
sonne|osxgsomix: ok fine!12:36
sonne|osxI will try to do the CR quickly12:36
* gsomix *sigh*12:36
sonne|osxI guess everyone here hopes that you can continue with .mat files12:36
foulwallHi, all . sorry for the network  latency12:36
sonne|osxeven psot gsoc12:37
sonne|osxpost12:37
gsomixsonne|osx, of course.12:37
wikinggsomix: cvmatio... have you seen that project?12:37
gsomixwiking, nope. we plan use some code from octave.12:38
wikinggsomix: it's binding opencv stuff to .mat files: github.com/hbristow/cvmatio12:38
gsomixbut OK, I'll see. thanks!12:38
wikinggsomix: it's pretty small and easy to port stuff that you need from there12:38
sonne|osxgsomix: that would work as well. but it is matlab 5 only not the new 7.3 format12:38
-!- foulwall [~AndChat63@219.141.159.55] has quit [Quit: Bye]12:38
sonne|osxwiking: makes sense to use it!12:38
wikingsonne|osx: works fine with newer matlab stuff12:38
gsomixcool!12:38
wikingsonne|osx: i mean the one i've sent just now12:38
sonne|osxwiking: most certainly not!12:39
-!- foulwall [~AndChat63@219.141.159.55] has joined #shogun12:39
wikingsonne|osx: ?12:39
wikingsonne|osx: i'm using matlab 2012 and i can read those .mat files with this :)12:39
sonne|osxwiking: ohh you are right12:39
sonne|osxman I had an old link12:39
sonne|osxhttps://github.com/hbristow/cvmatio12:40
wikingsonne|osx: CVMATIO implements a Matlab v7 MAT file parser in C++, giving users the12:40
sonne|osxgsomix: this will be of great help then12:40
wikingability to interact with binary MAT files in their own projects.12:40
wiking;)12:40
foulwallHi all, sorry  for the  disconnect.12:40
wikingso go ahead and have fun with it...12:40
sonne|osxgsomix: so it would be quite easy to do12:40
gsomixsonney2k, wiking is super hero for me now. :)12:40
sonne|osxalright then gsomix finish protobuf and then you can try this out :)12:40
wikingfoulwall: we are listening12:40
wiking:>12:40
wikingfoulwall: great work with the demo site!12:41
sonne|osxgsomix: alright then that's about it12:41
sonne|osxnext ?12:41
sonne|osxor foulwall ?12:41
foulwallHi , I'm vsry happy to implement your ideas12:41
lisitsynsorry back12:41
lisitsynwhat's left in foulwall's project?12:42
sonne|osxwiking: I still think it cannot do v7.3 (or4) formats though. But matlab 5 to 7.2 will work12:43
sonne|osxgsomix: https://github.com/hbristow/cvmatio/blob/master/src/MatlabIO.cpp - really very little code12:43
foulwallI'll make a more detailed docs, and make a python interface for the mldata hdf5 downloader12:43
wikingsonne|osx: it has some extra stuff in the .hpp... but yeah it's about 2000 LoC12:43
wikingtops12:43
foulwallI'll make them before  the hard deadline12:43
wikingsonne|osx: well it's better than nothing... :P12:44
sonne|osxfoulwall: I think the only missing stuff is arts/asp ?12:44
sonne|osxHeikoS: do we have any example which we want to work interactively still?12:45
foulwallOh. forget to merge that, arts and asp pr will  send today12:45
wikingsonne|osx: HeikoS is away atm12:45
sonne|osxfoulwall: ahh and the tree clustering but you have the tree already there so all that is needed is to call the hierarchical clustering example12:46
sonne|osxwiking: we should put foulwalls demos live on fatbot12:46
sonne|osxwiking: can we use your image for that?12:46
wikingsonne|osx: why on fatbot?12:46
wikingsonne|osx: but of course we can use that image...12:46
wikingsonne|osx: i'll setup up the startup script for that image today... along with the ipython notebook server12:47
lisitsynokay so12:47
wikingand then i can add as well the demo site12:47
sonne|osxwiking: fast/big machine!12:47
wikingsonne|osx: okok no worries12:47
lisitsyncan we proceed?12:47
wikinglisitsyn: yes who's next?12:47
lisitsynaz_de: pickle27 and me12:47
foulwallokey12:47
az_demaybe I can continue12:48
lisitsynaz_de: around12:48
lisitsyn?12:48
az_deok,  I just repeat Kevins e-mail, that12:48
lisitsynaz_de: yeah thanks12:48
az_de"all of his code for GSoC has been pushed to GitHub and merged! "12:48
lisitsynaz_de: we don't have anything left in the project, right?12:48
sonne|osxfoulwall: send a PR with arts/asp and then I show you how it works with hierarchical clustering (tomorrow?)12:48
wikingaz_de: that says everything great work from kevin! :)12:48
wikinglike his ipython notebook etc12:49
az_deand he is now waiting for feedback ...12:49
sonne|osxaz_de: in this case can you involve him to help us with the release?12:49
foulwallok sonne|work12:49
az_deyeah, The ipy-notebooks he created are really cool :-)12:49
lisitsynno he is not available I am afraid12:49
wikingsonne|osx: i think he doesn't have much time (see that preproc email he sent)12:49
lisitsynhe wanted to go travelling once gsoc is done12:49
sonne|osxaz_de: it will take time to get feedback from users. not before 3.0 I think12:49
sonne|osxahh too bad12:49
az_deok, I see12:49
wikingbut anyhow kevin did a great job12:50
wikinghe's gsoc project is totally ok12:50
wikingimo12:50
az_deAll in all, we have a very successful GSoC project!12:50
wikingyes, great work from all of you!12:50
az_deThere was a discussion to have some explicit whitening function in shogun, which is currently only implicitly done in some ICA methods.12:50
wikingaz_de: somebody will take over...12:50
az_deKevin has no time to do this now, but it should be straightforward .12:50
wikingthanx a lot again!12:50
wikingnext?12:50
sonne|osxyeah would be a nice preprocesor12:50
wikinglisitsyn: who's next?12:51
sonne|osxso thanks az_de for doing a great job with mentoring :)12:51
az_de:-)12:51
wikingaz_de: yeah really super job !12:51
HeikoSre12:51
wikingHeikoS: great12:51
lisitsynwiking: thats only hushell left I think12:51
wikinghushell: heheeey12:51
wikinghow's factor graph going along?12:51
wikingi see a lot of discussion about the notebook12:52
wikingwhat are the remaining issues?12:52
hushellok, let me start it12:52
hushellnot much to say, I have finished basic coding, the ipynb is here: http://nbviewer.ipython.org/657947212:52
shogun-buildbotbuild #15 of clang34 - static analysis is complete: Failure [failed analyse]  Build details are at http://buildbot.shogun-toolbox.org/builders/clang34%20-%20static%20analysis/builds/15  blamelist: Viktor Gal <viktor.gal@maeth.com>12:53
wikinghushell: ok cool how's the state of the notebook12:54
hushellwhat left is some comparison between SOSVM solvers, I decided to postpone the BCFW solver after gsoc, before the hard deadline, I'll compare the primalMosek, dualBMRM and the coming subgradient descent one12:54
wikingwhen is it going to be finalized?12:54
sonne|osxhushell: incredibly cool!12:54
@iglesiasghushell, btw, did you implement SGD finally?12:54
wikinghushell: u should try various solvers of dualBMRM12:54
wikinghushell: as there are  4 of them12:54
hushelliglesiasg: I have finished the code of SGD12:54
@iglesiasghushell, I thought you mentioned something about it in one of the discussions in github, but maybe I missed the PR12:54
wikinghushell: it's easy to do that you just give the right enum for that...12:55
hushellwiking: yeah, good idea, then there are more things to compare12:55
sonne|osxhushell: so very well done12:55
wikinghushell: go with the dualBMRM variants first... as mosek for example is not for free..12:55
wikinghushell: and as well we can see how the various dualBMRM behaves with a real world problem...12:55
sonne|osxyeah all non-free code is not so useful :/12:56
sonne|osxand we cannot automatically test it12:56
hushellI found mosek stuff are not in documentation12:56
sonne|osxso it will die out anyway12:56
wikinghushell: yeah as it's non-free.... as said go with the other SOSVM solvers...12:56
HeikoSsonne|osx: what was that with the interactive stuff? (later)12:56
hushellwiking: okay, then compare all BRRM solvers and the SGD12:57
wikinghushell: which is at the moment DUALBMRM... CCSOSVM needs more love from svmlight ;P12:57
wikinghushell: yeah that's good12:57
hushellI'll done this by Friday12:57
wikingok any more issues?12:57
wikingor stuff you need help with?12:57
hushellyeah, one question about the dataset12:57
sonne|osxHeikoS: anything we still want to run interactively as demo that would need extra work from foulwall?12:57
wikinghushell: shoot go ahead12:58
-!- travis-ci [~travis-ci@ec2-54-211-149-66.compute-1.amazonaws.com] has joined #shogun12:58
travis-ci[travis-ci] it's Viktor Gal'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/1154337412:58
-!- travis-ci [~travis-ci@ec2-54-211-149-66.compute-1.amazonaws.com] has left #shogun []12:58
HeikoSsonne|osx: ah ok, not really for GSoC, but I hope it is easy to do these things myself and its well documented12:58
lisitsynokay so gsoc-wise12:58
@iglesiasghushell, if you can use matlab, basically load the .mat file and then use the function csvwrite12:58
hushellso if I have a cell array, is it possible to write it into csv, and read it from python?12:58
HeikoSsonne|osx: wanted to pimp the GP stuff a bit mote (regression heatmaps etc) but no time for this now12:58
wikinghushell: csv u can read from python w/o any problems...12:58
sonne|osxHeikoS: well you should do it now - otherwise you will never realize if there are issues12:58
@iglesiasghushell, what do you have in this cell array?12:59
lisitsynjust a general thing - don't bother with code submission for now12:59
shogun-buildbotbuild #8 of debian wheezy - memcheck is complete: Failure [failed memory check]  Build details are at http://buildbot.shogun-toolbox.org/builders/debian%20wheezy%20-%20memcheck/builds/8  blamelist: Viktor Gal <viktor.gal@maeth.com>12:59
hushelliglesiasg: just many matrix12:59
sonne|osxHeikoS: I mean realize when it is too late12:59
HeikoSsonne|osx: the model-selection is not yet working, but it would be nice to have a button for that12:59
sonne|osxhushell: I sse12:59
sonne|osx* see12:59
HeikoSsonne|osx: thats why I say that there should be very good documentation of the framework12:59
hushelleach cell is a matrix12:59
sonne|osxhushell: maybe we just use a .mat for now12:59
HeikoSsonne|osx: if no-one can edit it after foulwall left, that is of no use12:59
wikingsonne|osx: hushell go with .mat12:59
hushellso with scipy.io is more convenient13:00
wikingsonne|osx: and i'll help out gsomix with .mat reader13:00
sonne|osxI don't think it is easy to convert that to a .csv13:00
sonne|osxit needs a custom format13:00
HeikoSsonne|osx: I will have a look later, its not possible before GSoC end since I want to help my guys a bit13:00
wikinghushell: don't waste time on this...13:00
hushellsonne|osx: yeah, because I don't want to do the train-test splitting in python13:00
-!- thoralf [~thoralf@enki.zib.de] has quit [Ping timeout: 260 seconds]13:00
hushellokay13:00
hushellI'll try to improve the dataset later13:00
sonne|osxhushell: that is the easiest. anyway just do it13:00
foulwallHeikoS , I'll be very careful writting the docs of the framework13:01
sonne|osxwith .mat13:01
wikingjust go ahead with .mat and forget the rest...13:01
wikingcool13:01
wikinghushell: thanks a lot for the work13:01
wikinggreat job!13:01
sonne|osxso we are done with the summaries?13:01
wikingany of the gsoc students... if u get stuck plz mail shout here on irc for help13:01
wikingdont waste your time on things alone..13:01
hushellThanks for help! :)13:02
wiking*all* ^13:02
wikingcool13:02
wikingsonne|osx: yeah it's done13:02
sonne|osxSo Sept 16 is suggested pencils down13:02
sonne|osxerr 2313:02
lisitsynno13:02
HeikoSno 23 is hard13:02
lisitsynsept 16 is suggested13:02
lisitsyn23 is hard13:02
sonne|osx(well the hard pencils down)13:02
wikingok13:02
sonne|osxwe will then again have to do evaluations13:02
sonne|osxand we should do them by Sept 2513:02
sonne|osxstudents bug your mentors!13:03
HeikoSotherwise no money ;)13:03
georg_zelleralright date's saved13:03
sonne|osxyeah and HeikoS no summit :D13:03
shogun-buildbotbuild #3 of clang34 - undefined behaviour analysis is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/clang34%20-%20undefined%20behaviour%20analysis/builds/3  blamelist: Viktor Gal <viktor.gal@maeth.com>13:03
@iglesiasghaha13:03
@iglesiasgHeikoS, bug mentors you too!!13:03
@iglesiasg:D13:03
HeikoShaha13:04
HeikoScan do :)13:04
sonne|osxAlright - this cannot be repeated often enough.13:04
sonne|osxWe are very happy with the work you have all done.13:04
sonne|osxAnd it would be very cool if you would stay post-gsoc and do cool stuff with shogun.13:04
georg_zellergreat work from everyone indeed! Impressive!13:05
HeikoSYep, incredible results this GSoC. I am sure this will get a lot of attention13:05
sonne|osxAs you know we plan to do a shogun 3.0.0 release just once gsoc is over.13:05
wikingyeah plz hang around13:05
sonne|osxSo all your work will be included and it would be *great* if you could help us fixing bugs/warnings etc.13:05
hushellvery exciting about the 3.0.0!13:05
wikingookeeeeey13:06
HeikoSWe probably will need a bugfix/cleanup release 3.1 later this year where we fix all problem that people found in 3.0 :)13:06
@iglesiasgI was in a similar situation last year, I cannot do other thing but strongly suggest to keep hanging around and collaborate13:06
wiking3.0.0 time!13:06
@iglesiasgit cannot be of course at the same pace as during GSoC, we all have other stuff to do13:06
wikingagenda for that13:06
sonne|osxIndeed 3.0 means we doubled shoguns code base since the last release13:06
wiking1) fix fix fix13:06
wiking2) fix fix fix13:06
wiking:>13:06
lisitsynhaa this means13:06
wikingokey so i need some help here13:06
lisitsynwith 5.0 we will get unmaintanable software13:06
lisitsyn:D13:06
wikingwe are having a lot of leaks atm13:06
wikingand a lot of undefined behaviour and a lot of static analyzer errors13:07
wikingso13:07
sonne|osxlisitsyn: you forget wikings botnet13:07
wikingnow comes the prio list13:07
HeikoSwiking: maybe we should have a  priority list for the release? checking all issues for 3.013:07
sonne|osxyeah as wiking said13:07
lisitsynsonne|osx: well it could help to ddos NSA13:07
lisitsynbut not to maintain ;)13:07
wikingso HeikoS it's just coming now13:07
sonne|osxhttp://buildbot.shogun-toolbox.org/waterfall13:07
wikingso please pleaaaase13:07
wikingbefore doing anything try to do this first13:07
sonne|osxfix bugs / failed tests13:07
sonne|osxwarnings13:07
wikingHeikoS: for you the list:13:07
sonne|osxfor example these http://buildbot.shogun-toolbox.org/static_analysis/2013-09-19-1/13:08
HeikoSok so first goal is to make all builds work I guess?13:08
-!- lambday [67157e37@gateway/web/freenode/ip.103.21.126.55] has quit [Ping timeout: 250 seconds]13:09
sonne|osxyeah we are trying this for a long time already...13:09
sonne|osxbut we are close (only cygwin fails basically)13:09
wikingHeikoS: https://github.com/shogun-toolbox/shogun/issues/1574 evaluation_cross_validation_multiclass base_load_file_parameters /base_load_all_file_parameters  aaadn all the errors at static analyzer for  /base /Parameter.cpp (http://buildbot.shogun-toolbox.org/static_analysis/2013-09-19-1/)13:09
wikingHeikoS: and there are some stuff about *MMD13:09
sonne|osxand the rest is bugs from static analyzers13:09
wikingsonne|osx: yeah not always...13:10
HeikoSwiking, alright I will do those things first and then get back13:10
wikingsonne|osx: most of the time actually not but13:10
wikingeven if it is13:10
wikingone can silence it with an assert or so13:10
sonne|osxwell stuff we should fix but it is not prime importance13:10
hushelliglesiasg: then I'll have another PR for the SGD on Friday, will you have time to have a short review?13:10
wikingHeikoS: especially start with https://github.com/shogun-toolbox/shogun/issues/157413:10
@iglesiasghushell, sure, thank you!13:10
wikingi need this fixed asap...13:10
sonne|osxwe need debian packages, besser82 is doing fedora packages13:10
HeikoSkk13:10
sonne|osxand we need osx packages13:11
sonne|osxetc13:11
wikingHeikoS: because it silences other warnings from other modules...13:11
wiking*all* of you13:11
HeikoSah thats annoying13:11
wikingwho are doing release: http://buildbot.shogun-toolbox.org/static_analysis/2013-09-19-1/13:11
besser82sonne|osx: aren't you the DD? ;)13:11
wikingtry to at least silence it13:11
hushelliglesiasg: as Patrick suggested, I should check the objective values, slack variables, should be the same as his matlab code, this will take a lot of time13:11
wikinglisitsyn: i gave you your malsar leaks...13:11
wikinglisitsyn: and fix the dead assigments here: http://buildbot.shogun-toolbox.org/static_analysis/2013-09-19-1/13:12
@iglesiasghushell, I understand13:12
lisitsynwiking: yeah ok13:12
wikingfor all the tapkee related stuff13:12
hushelliglesiasg: but I'll try to fix this by Friday, we have code review in the weekend13:12
sonne|osxbesser82: sure sure but I need to sleep too :D13:12
hushelliglesiasg: thanks in advance! :)13:12
@iglesiasghushell, but it will be so cool to have SGD working good13:12
besser82sonne|osx: sleep? Who needs that? :-P13:12
wikingcurrent leak xml: http://buildbot.shogun-toolbox.org/memcheck/20130919-1012/DynamicAnalysis.xml13:12
sonne|osxiglesiasg: I still hope you do the website so we have the notebooks displayed nicely13:13
wikingbasically there's still tons of them there13:13
hushelliglesiasg: I hope it will be faster then cutting plane methods ;)13:13
@iglesiasgsonne|osx, yes13:13
sonne|osxand HeikoS you wanted to change the README to be a .md and to have links between the READMEs and INSTALLs13:13
hushellokay, I'll go back to sleep, thank you guys!13:13
HeikoSsonne|osx: yes I still want to13:13
sonne|osxhushell: vojtech's stuff is *very* fast so I am not so convinced13:13
@iglesiasgsonne|osx, after what wiking has done with the cloud notebooks thing it should be easy, right?13:14
sonne|osxhushell: thanks everyone for attending13:14
wikingiglesiasg: unrelated13:14
sonne|osxhushell: and good night13:14
HeikoSgoood night :)13:14
wikingiglesiasg: we need to statically show the notebooks somewhere...13:14
wikinghushell: gnite13:14
sonne|osxiglesiasg: ?13:14
hushellhave a great day! guys!13:14
@iglesiasgok, then I misunderstood the cloud notebooks thing13:14
wikingiglesiasg: indeed13:14
sonne|osxwiking: ?13:14
wikingiglesiasg: cloud is for editing playing13:14
@iglesiasgwiking, oh yeah sure13:15
sonne|osxwell that is totally unrelated13:15
wikingiglesiasg: but we still need to be able to statically show somewhere the notebooks13:15
@iglesiasgfuck I am stupid13:15
sonne|osxiglesiasg: we need to display examples on the website13:15
wikingiglesiasg: and it should be able to handle shogun versions13:15
sonne|osxinteractive demoing is sth else13:15
sonne|osxand we have 2 parts here13:15
@iglesiasgthe cloud thing is to offer like an ipython session with shogun installed13:15
sonne|osx1) is foulwalls demos / apps13:15
wikingiglesiasg: say "notebooks generated" with shogun 3.0, and then latest and shogun 2.013:15
wikingand such...13:15
@iglesiasgwiking, aham I see13:15
sonne|osx2) is the notebooks ...13:15
@iglesiasgso there are a couple of parts here13:16
@iglesiasgone is to generate the notebooks periodically13:16
wikingyes13:16
wikingiglesiasg: that's done13:16
@iglesiasgsave them13:16
sonne|osxguys I have to get back to work13:16
@iglesiasgand put a link to them in the webpage13:16
sonne|osxcu all13:16
wikingiglesiasg: it needs integration now to the webpage13:16
-!- sonne|osx [~sonne@89.204.135.239] has quit [Quit: sonne|osx]13:16
@iglesiasgsonne|osx, bye!13:16
wikingiglesiasg: #148213:16
wiking#148113:16
wiking#148313:16
wikingthat's all the issues here13:17
@iglesiasgwiking, mm ok. Is it basically about putting links in the webpage then?13:17
wikingthe notebook generation (nightly) is done13:17
shogun-buildbotbuild #8 of clang34 - thread analysis is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/clang34%20-%20thread%20analysis/builds/8  blamelist: Viktor Gal <viktor.gal@maeth.com>13:17
@iglesiasgthat sounds tooe asy13:17
wikingiglesiasg: well not only that... somehow we need to be able to switch between versions13:17
@iglesiasgok13:17
wikingso like shogun 3.013:17
wikingand latest13:17
wikingfor the same notebook13:17
@iglesiasgbut first things first probably13:17
-!- lambday [67157c37@gateway/web/freenode/ip.103.21.124.55] has joined #shogun13:17
@iglesiasgfirst I will aim at showing them generated for one version, and then extend to several versions13:18
@iglesiasgwiking, what do you think?13:18
wikinglambday: i've told you about your mem leak.... please fix that before doing anything else...13:18
lambdaywiking: working on that13:18
wikingiglesiasg: i'm fine with that13:18
wikinglambday: great thnx...13:18
lambdaysorry I got dc.. power cut :(13:18
wikingHeikoS: i think you should just define -1 enums and then do a default: SG_ERROR(...) in the switch branch :P13:18
@iglesiasgwiking, where are they currently generated? the notebooks13:19
wikingiglesiasg: jsut a sec...13:19
@iglesiasgthe idea is that they are in doc/notebooks/* without the output saved13:19
-!- Cheng [73406f11@gateway/web/freenode/ip.115.64.111.17] has quit [Quit: Page closed]13:19
@iglesiasgso that must be executed somewhere and saved somewhere else with the output13:20
@iglesiasgwiking, sure13:20
wikingiglesiasg: the nightly_default does that13:20
wikingi mean execution + saving13:20
@iglesiasgok13:20
@iglesiasgwiking, one question about the visualization13:20
wikingiglesiasg: http://buildbot.shogun-toolbox.org/builders/nightly_default/builds/528/steps/notebooks/logs/stdio13:21
wikingi'm just checking where's the output13:21
@iglesiasgwiking, are we aiming at basically providing links to nbviewer.ipython.org/****13:21
wikingiglesiasg: no... we need to save it locally13:21
wikingi mean the full thing13:21
@iglesiasgor have it rendered somewhere else?13:21
wikingit should be fully rendered on our webpage imho13:21
@iglesiasgwiking, yep sure, the full thing saved locally13:21
@iglesiasgmmm I see rendered on our webpage13:22
@iglesiasgit is indeed possible because basically the notebook runs on the browser always13:22
wikingiglesiasg: on shogun website server the notebooks are here13:23
wiking/home/shogun/static/notebook13:23
wikingi dont know if this helps you at allk13:23
wikingiglesiasg: indeed it does: http://shogun-toolbox.org/static/notebook/ecg_sep.ipynb13:23
@iglesiasgcool!13:24
@iglesiasgthat helps me for sure13:24
wikingiglesiasg: so all the notebooks are saved there13:24
wikingbut these are the nightlies13:24
@iglesiasgwiking, but there is no output in there I think13:25
wikingiglesiasg: http://shogun-toolbox.org/static/notebook/gp_regression.ipynb13:25
wikingthis one does for sure13:25
wikingif u think one is missing something13:25
wikinglet me know13:25
wikingand i'll fix it13:25
@iglesiasgwiking, yep13:25
@iglesiasgthat one has the output13:25
@iglesiasgso first I have to investigate how to render the notebook from those urls13:26
wikinglcoally13:26
wikingif u need anything for that let me know13:26
wikingi think we'll just do something like13:26
wikingstatic/notebook/{RELEASE_VERSION}/13:26
wikingfor the various verions of the notebooks13:26
@iglesiasgmake sense13:26
wikingand then we'll do somehow the switching from the website...13:27
@iglesiasgsorry, what did you mean with locally just a minute ago?13:27
wikingiglesiasg: basically what we need is somehow list the availble notebooks on the website13:27
wikingand the somehow to be able to render it13:27
@iglesiasgagree13:27
wikinglist is easy from django i guess13:27
wikingls directory13:28
wiking:P13:28
wikingok i have to take a break now13:28
wikingppl r waiting for me on lunch13:28
@iglesiasghehe yeah13:28
wikingbbl13:28
wikinglet me know if u need something from me13:28
@iglesiasgI will work on the rendering thing13:28
wikingiglesiasg: afaik there's a python script for this ;)13:28
wikingiglesiasg: within ipython notebook13:28
@iglesiasgit makes total sense13:28
@iglesiasgI was going to say that13:28
sonne|workwiking: what is with videos?13:28
@iglesiasgthat I bet there is something already to do it :)13:29
sonne|workworkshop videos I mean?13:29
wikingiglesiasg: even more... i know that there is an option of ipython to generate a html13:29
@iglesiasgyeeah videoooooz13:29
wikingiglesiasg: please find out the command and i'll make sure that you get the htmls generated13:29
wikingsonne|work: will be ready by 3.013:29
wiking:D13:29
wikingbbl ... ppl r really waiting for me..13:29
sonne|workwiking: ohoh13:29
wikingwhat?13:30
wikingplz be fast13:30
@iglesiasgwiking, ok13:30
@iglesiasggo and have lunck now!13:30
@iglesiasg:D13:30
wikingiglesiasg: "These notebooks are normal files that can be shared with colleagues, converted to other formats such as HTML or PDF, etc"13:30
wikingso we can even put the pdf somewhere...13:30
wikinganyhow as u can see it's possible13:30
wikingit's just one ipython notebook command13:30
@iglesiasgyep13:31
@iglesiasgso the idea is to the nightly doing that as well, right?13:31
@iglesiasggenerating the HTML13:31
-!- thoralf [~thoralf@enki.zib.de] has joined #shogun13:34
thoralfHey.13:34
@iglesiasglunch time for me too, see you later13:35
-!- foulwall [~AndChat63@219.141.159.55] has quit [Remote host closed the connection]13:35
-!- georg_zeller [~zeller@embln.embl.de] has left #shogun []13:37
-!- az_de [82954e22@gateway/web/freenode/ip.130.149.78.34] has quit []13:37
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has quit [Quit: Good bye!]13:53
wikingback13:56
@iglesiasgwiking, found it out already13:57
@iglesiasgI just tested it13:57
@iglesiasgwiking, ipython nbconvert notebook.ipynb13:57
wikingiglesiasg: what's the command?13:57
@iglesiasgyou get a nice notebook.html13:57
wikingiglesiasg: it'll be the same name?13:57
@iglesiasgpandoc, among maybe other stuff, is required13:57
@iglesiasgyes13:57
wikingiglesiasg: i guess it's not a selfcontained html13:58
wikingbut there are like images saved next to it13:58
wikingor?13:58
@iglesiasgwhy not?13:58
@iglesiasgI just opened it in my browser13:58
@iglesiasgimages and everything13:58
@iglesiasgyes, self contained13:58
wikingiglesiasg: well i thought that the images are not within the html13:58
wikingso they are in the html encoded base64?13:58
@iglesiasgwiking, they are encoded in the html13:59
@iglesiasgno idea what base13:59
wikinglol13:59
wikingok lemme test it13:59
wikinghehehe14:00
wikingPandocMissing: The command 'pandoc -f markdown -t html --mathjax' returned an error: [Errno 2] No such file or directory.14:00
@iglesiasgtold ya14:00
@iglesiasgwiking, http://pastebin.com/yVdJ7mb314:00
@iglesiasgwiking, tell me the base :P14:00
wikingokey14:00
wikingbase6414:00
wiking<img src="data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAARwAAAETCAYAAADkork1AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz14:00
wiking...14:00
@iglesiasgyeah14:00
@iglesiasgI just realized...14:00
wikingokeey so we need this pandoc installed on 7nn14:01
wikingand then we are ready to go14:01
@iglesiasgdidn't I mention before that I am stupid :P14:01
wikingi'll add this into the ipython notebook generator script14:01
@iglesiasgwiking, cool14:01
wikingand then basically in the same directory14:01
wikingwhere the notebooks are14:01
wikingthe html is going to reside as well14:01
wikingsonne|work: do i have sudo on 7nn? :D14:01
wikingsonne|work: seems not... :(14:02
@iglesiasgno hurry14:02
wikingiglesiasg: ok we'll have to wait till sonne|work installs pandoc on that machine. but anyhow we can assume that there's going to be the html files as well14:02
@iglesiasgI have to finish off my other LMNN stuff first anyway14:02
@iglesiasgwiking, sure14:02
wikingiglesiasg: then basically what we need is to list the *.html files in that directory14:03
wikingmake a reference to them14:03
@iglesiasgyes14:03
wikingand we are set14:03
wikingi'll setup later the demo site on fatbot14:03
wikingand then we'll figure it out how to integrate that into the website14:04
@iglesiasgokok14:04
lambdaywiking: seems like I fixed it.. for allocating the list of SGSparseVectors and SGSparseVectorEntry, if I use SG_CALLOC instead of SG_MALLOC, that fixes it14:04
wikinglambday: lol?!14:05
lambdayso SG_CALLOC calls constructor for each one in the list while SG_MALLOC doesn't14:05
wikinglambday: how is that possible...14:05
wikingokeeey14:05
lambdaysonne|work: is that right?14:05
lambdayargh something went wrong again14:06
wikinglambday: calloc and mallco should be the same14:07
wikingapart from how the memory is initialized14:07
wikingi mean in case of calloc it's full 0 in case of malloc it's random14:07
sonne|worklambday: no SG_MALLOC should do that too14:07
lambdaychecking more14:09
lambdaylol14:09
lambdayif I do this --->14:09
lambdaySGSparseMatrix<type> sm(size, size);14:10
lambdayand then sm.sparse_matrix=blah;14:10
lambdayit leaks14:10
lambdaybut14:10
lambdayif I do SGSparseMatrix<type> sm;14:10
lambdaysm.sparse_matrix=...., sm.num_vectors=... sm.num_features=... its alright14:10
lambday:-/14:10
lambdaysonne|work: is there a reason for this?? :(14:12
lambdaywiking: fixed it though..14:12
lambdaywiking: I was checking that particular test-case explicitely... if I want to run valgrind for all of the tests, what should I use? (sorry irc pms are not saved I completely forgot what you said :( )14:14
sonne|worklambday: that is weird14:16
wikinglambday:14:16
lambdaysonne|work: yes! :-/14:16
wikingvalgrind "-q" "--tool=memcheck" "--leak-check=yes" "--show-reachable=yes" "--workaround-gcc296-bugs=yes" "--num-callers=50" ./tests/unit/shogun-unit-test "--gtest_filter=SparseMatrixOperator.*"14:16
lambdaysonne|work: wait I'll show you a gist14:16
wikingshould show u if there's an error14:16
lambdaywiking: thanks a lot14:16
lambdaywiking: wait.. so I should use cmake using just -DENABLE_TESTING=ON and then run valgrind separately?14:17
wikingyes14:18
lambdaywiking: thanks :)14:18
wikingok this is amaazing14:18
wikingin multiclass machine refcount suddenly jumps up to 3 from 114:18
wiking:)14:18
wikingsonne|work: do you know by heart how to setup a watchdog?14:18
wikingin gdb14:18
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout]14:20
sonne|workwiking: why watchdog?14:21
sonne|workwhat for?14:21
-!- van51 [~van51@athedsl-409794.home.otenet.gr] has joined #shogun14:22
wikingno not watchdog14:22
wikingbut a watch14:22
wiking..14:22
sonne|workwiking: ahh no14:23
lambdayHeikoS: hi14:23
lambdayHeikoS: do you think the PR containing 2 python modular examples are okay? I made the changes that you suggested14:24
lisitsynwiking: e.g. catch throw14:24
lisitsyncatch File.hpp:17514:24
lisitsyniirc14:24
HeikoSlambday:  did you already send?14:24
lambdayHeikoS: yes I updated earlier :)14:25
lambdayHeikoS: I want to send another PR with the sparse matrix leak thing14:25
lambdayI don't want it to get inside that PR :(14:25
lambdaywiking: running valgrind "-q" "--tool=memcheck" "--leak-check=yes" "--show-reachable=yes" "--workaround-gcc296-bugs=yes" "--num-callers=50" ./tests/unit/shogun-unit-test "--gtest_filter=SparseMatrixOperator.*" reports no leak, just normal gtest style output14:26
HeikoSlambday:  sorry I totally forgot about that one, since travis failed back then14:26
HeikoSlambday: let me check14:26
wikinglisitsyn: nono i want to specify  a memory address where and if there's a change in it i want it to break :)14:26
lambdayHeikoS: np ! :)14:27
lambdayHeikoS: I also need some help with ozone mcmc14:27
HeikoSlambday: problem is that the integration testts fail with this one14:28
HeikoSI will merge, but pls fix them asap before doing anything else14:28
HeikoShave a look at travis output14:28
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun14:28
shogun-notifier-shogun: lambday :develop * ad00437 / examples/undocumented/python_modular/ (2 files): https://github.com/shogun-toolbox/shogun/commit/ad0043712ce6f6e2d27be2457a2b249c877f68fd14:28
shogun-notifier-shogun: Added python modular examples (log-det)14:28
shogun-notifier-shogun: Heiko Strathmann :develop * a9509ce / examples/undocumented/python_modular/ (2 files): https://github.com/shogun-toolbox/shogun/commit/a9509cea22729876f57517a98c26838b4862c91d14:28
shogun-notifier-shogun: Merge pull request #1602 from lambday/feature/log_determinant14:28
shogun-notifier-shogun:14:28
shogun-notifier-shogun: Added python modular examples (log-det)14:28
HeikoSlambday: ozone mcmc?14:29
HeikoSlambday: ok can do, let me just finish some stuff here for wiking14:29
lisitsynwiking: ohh didn't try it14:29
lisitsynwiking: please tell me how to do that in case of success14:29
lambdayHeikoS: integration fails?14:29
lambdayHeikoS: checking the travis14:29
HeikoSlambday: yep on your examples14:29
HeikoScheck the output14:30
wikinglisitsyn: works like a charm14:30
wikinglisitsyn: watch -l *0x2a795c014:30
HeikoSlambday: run the tester.py locally14:30
wikinglisitsyn: watch -l *<memory location>14:30
HeikoSbefore comitting python examples you might have to update data14:30
lambdayHeikoS: alright14:30
lambdaychecking14:30
lisitsynwiking: once changed it stops executing, right?14:31
wikinglisitsyn: yes it's like a breakpoint14:32
wikinglisitsyn: and it'll as well show the new and old value14:32
wikingand of course you have the whole backtrace14:32
lisitsynwiking: awesome14:33
wikingyeah it's basic14:33
wiking;)14:33
wikingi just keep forgeting the syntax14:33
wikingand i just just really start using lldb14:33
wikingit's really fucking great in some aspects14:33
shogun-buildbotbuild #2216 of deb1 - libshogun is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb1%20-%20libshogun/builds/2216  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>14:34
thoralfwiking: Using lldb?  How?  Tell me. :)14:38
wikingthoralf: http://lldb.llvm.org/lldb-gdb.html14:39
thoralfwiking: All the great things you found out.14:39
shogun-notifier-shogun: Viktor Gal :develop * 12d3062 / src/shogun/machine/LinearMulticlassMachine.h: https://github.com/shogun-toolbox/shogun/commit/12d3062d2d7e9ba3ad334f74f952ce62b31c031b14:39
wikingok here's another leakfix14:39
shogun-notifier-shogun: Fix memory leak in LinearMulticlassMachine issue #162914:39
thoralfOh no.  No read-the-fine-manual again. ;)14:39
shogun-buildbotbuild #2217 of deb1 - libshogun is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb1%20-%20libshogun/builds/2217  blamelist: Viktor Gal <viktor.gal@maeth.com>14:44
wikingoh fucking libshogun-library_mldatahdf514:44
shogun-buildbotbuild #2215 of deb1 - libshogun is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb1%20-%20libshogun/builds/2215  blamelist: lambday <heavensdevil6909@gmail.com>14:47
wikingmmm14:48
wikingthis seems to be an old PR14:48
wikingshogun-buildbot: force build --branch=develop 'deb1 - libshogun'14:48
shogun-buildbotbuild #2218 forced14:48
shogun-buildbotI'll give a shout when the build finishes14:48
shogun-buildbotbuild #2218 of deb1 - libshogun is complete: Success [build successful]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb1%20-%20libshogun/builds/221814:49
wikingHeikoS: awesome for that PR14:49
wikingHeikoS: can we merge?14:50
HeikoSwiking: I cannot run tests locally here since computer from 200614:50
HeikoSor even older ;)14:50
wikingoookeey14:50
HeikoSso have a look at travis and feel free to merge if it passes, which I think it should14:50
wikingi'll see how the first 2 jobs are runnning on travis14:50
wikingif they pass i'll merge14:50
HeikoSlambday: your merged PR does not cause trouble on travis, its other people's stuff14:55
lambdayHeikoS: yes... I guess earlier issue was because shogun-data PR was not merged till then14:56
HeikoSlambday: I see, good!15:02
lambdayHeikoS: sending the PR for memory leak then15:02
wikinglambday: i'll check it now15:03
lambdaywiking: thanks :)15:05
shogun-notifier-shogun: lambday :develop * 72cbe0e / src/shogun/mathematics/linalg/linop/SparseMatrixOperator.h: https://github.com/shogun-toolbox/shogun/commit/72cbe0e60b1dc580c9f1103883b6dde5e0fa056215:05
shogun-notifier-shogun: SparseMatrixOperator memory leak fixed15:05
shogun-notifier-shogun: Heiko Strathmann :develop * 6e8ab05 / src/shogun/mathematics/linalg/linop/SparseMatrixOperator.h: https://github.com/shogun-toolbox/shogun/commit/6e8ab05a25626208cb7aa53a9066082e2bc7379515:05
shogun-notifier-shogun: Merge pull request #1632 from lambday/feature/log_determinant15:05
shogun-notifier-shogun:15:05
shogun-notifier-shogun: SparseMatrixOperator memory leak fixed15:05
lambdayHeikoS: the fix was indeed weird15:06
HeikoSyeah, thanks for figuring that out! :)15:06
lambdaystill no idea why it was :-/15:06
lambdayyeah num_feature/num_vector thing was there but I was testing with a square matrix :-/15:07
lambdayit didn't cause anything15:07
wikinglambday: ok great thnx for the fix!15:09
wikingand the fast response15:09
lambdaywiking: np :)15:09
lambdaywiking: now you have to help me out with swig :(15:10
wikingah cool15:10
wikinglemme check what could be done about that one15:10
wikingbasically u need a numpy complex64 <-> complex64_t mapping?15:10
wikingright?15:10
lambdaywiking: yes... I want to use ComplexSparseMatrixOperator, which is CSparseMatrixOperator<complex64_t>15:11
lambdayand SGVector<complex64_t>15:11
wikingokeeey15:11
lambdaywiking: I didn't add this template to modular interface yet, tried with a few things but didn't work15:11
wikingi'll check how it can be done15:11
lambdaywiking: thanks man15:12
wikingwhere's complex64_t defined?15:16
wikingok i see15:17
-!- van51 [~van51@athedsl-409794.home.otenet.gr] has quit [Quit: Leaving.]15:18
wikinglambday: ok so you need numpy mapping for SGVector<complex64_t>15:20
wikingand what else?15:21
lambdaywiking: and numpy mapping for csc_matrix(complex64) for SGSparseMatrix<complex64_t>15:22
wikingok so u need SGSparseMatrix<complex64_t> as well15:23
lambdayyes15:23
wikingmaybe SGMatrix<complex64_t> ?15:23
lambdaywiking: good to have, not urgent15:23
lambdayfor my examples I'll just be using sparse ones15:24
wikingok i'm testing now15:26
HeikoSwiking: unit tests segfault. I have to look into this a bit later today15:33
HeikoSgoing now,see you!15:33
lambdaywiking: thanks15:36
lambdayHeikoS: I'll update notebook once complex things work.. see you15:37
-!- HeikoS [05453f70@gateway/web/freenode/ip.5.69.63.112] has quit [Ping timeout: 250 seconds]15:38
-!- travis-ci [~travis-ci@ec2-54-234-21-192.compute-1.amazonaws.com] has joined #shogun15:51
travis-ci[travis-ci] it's Heiko Strathmann'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/1154904315:51
-!- travis-ci [~travis-ci@ec2-54-234-21-192.compute-1.amazonaws.com] has left #shogun []15:51
shogun-buildbotbuild #1498 of cyg1 - libshogun is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/cyg1%20-%20libshogun/builds/1498  blamelist: lambday <heavensdevil6909@gmail.com>15:55
-!- lambday_ [67157d37@gateway/web/freenode/ip.103.21.125.55] has joined #shogun15:56
-!- lambday [67157c37@gateway/web/freenode/ip.103.21.124.55] has quit [Ping timeout: 250 seconds]15:59
wikingsonne|work: is there a reason why we dont use https://github.com/numpy/numpy/blob/master/doc/swig/numpy.i16:00
wiking?16:00
sonne|workwiking: yes it does not support all the data types shogun has - however we started with this one16:01
wiking :(16:01
wikingsonne|work: i've just realised that we are actually copying back and forth values of an array16:01
wikingi.e. sgvector -> numpy16:02
wikingthat's not a mapping but a copy16:02
wikingor am i reading it wrong?16:02
sonne|workwiking: yes we copy16:02
sonne|workin all interfaces!16:02
wikingsonne|work: isn't there a way to actually just map the memory?16:02
wikingor the memory is totally differnt16:03
wikinghence we cannot do it16:03
wikingi mean the structure16:03
sonne|workwiking: shogun objects destroy memory currently16:03
sonne|workfree I mean16:03
wikingsonne|work: mmm you mean reference is not being hold on the modular interface side16:03
wikingright?16:03
sonne|workyes16:05
sonne|workand we free memory too16:06
wikingmmmm16:06
wikinghow can a create a RealVector from numpy array?16:06
sonne|workyou cannot16:06
sonne|workohh no I think you can16:06
sonne|workwiking: x=RealVector()16:07
sonne|workx.set(numpy.array([1.,2,3]))16:07
wikingok16:08
wikingdoh16:08
wikingSGVector<complex64_t> a;16:09
wikinga[0] should work16:09
wikingright?16:09
sonne|workwiking: no idea16:10
wikinglambday_: ?16:10
sonne|workwiking: with float64_t for sure16:10
wikingsonne|work:  yeah i know16:11
wikingIn [34]: x = ComplexVector()16:11
wikingIn [36]: a = np.array([1, 2, 3], dtype=complex)16:11
wikingIn [37]: x.set(a)16:11
wikingthis works16:11
sonne|workbut Idon't see why not16:11
wikingbut then i try16:11
wikingx[0]16:11
wikingIn [38]: x[0]16:11
wiking---------------------------------------------------------------------------16:11
wikingTypeError                                 Traceback (most recent call last)16:11
wiking<ipython-input-38-1ae75c28907a> in <module>()16:11
wiking----> 1 x[0]16:11
wikingTypeError: 'ComplexVector' object does not support indexing16:12
sonne|workahh from python16:12
wikingyeah16:12
sonne|workno I guess it needs the inplace stuff gsomix was doing16:12
wikingbut there's a lot of things like this in SGVector.cpp16:12
wikingdefine COMPLEX64_ERROR_NOARG(function) \16:12
wikingSG_SERROR("SGVector::%s():: Not supported for complex64_t\n",\16:12
wiking...16:12
sonne|workwiking: yeah sure. some ops cannot be sanely defined with complex16:13
wikingok i think i need more work on this16:13
wikingbut then again16:13
thoralfwiking: min/max/sort won't work with complex16:13
wikingthoralf: i've seen that16:13
wiking [40]: x.vector16:13
wikingOut[40]: <Swig Object of type 'std::complex< double > *' at 0x113039500>16:13
wikingIn [41]: x.vlen16:13
wikingOut[41]: 316:13
wikingso it seems it is set...16:14
wikingi need extra stuff in the typemapping16:14
wikingi guess..16:14
wikingbut basic stuff is already ready :16:14
wiking:P16:14
lambday_wiking: hey16:15
wikinglambday_: so you were working with SGVector<complex64_t> right?16:16
lambday_wiking: didn't try with vector, was trying with sparse matrix16:16
lambday_but I need vector too16:16
-!- travis-ci [~travis-ci@ec2-23-22-84-25.compute-1.amazonaws.com] has joined #shogun16:36
travis-ci[travis-ci] it's Viktor Gal'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/1154937516:36
-!- travis-ci [~travis-ci@ec2-23-22-84-25.compute-1.amazonaws.com] has left #shogun []16:36
shogun-buildbotbuild #1499 of cyg1 - libshogun is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/cyg1%20-%20libshogun/builds/1499  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>16:40
shogun-buildbotbuild #16 of clang34 - static analysis is complete: Failure [failed analyse]  Build details are at http://buildbot.shogun-toolbox.org/builders/clang34%20-%20static%20analysis/builds/16  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>16:44
-!- gsomix [~gsomix@188.168.3.89] has quit [Quit: Leaving]16:52
shogun-buildbotbuild #4 of clang34 - undefined behaviour analysis is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/clang34%20-%20undefined%20behaviour%20analysis/builds/4  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>, lambday <heavensdevil6909@gmail.com>16:55
@iglesiasglambday_, 6909 haha16:55
shogun-buildbotbuild #9 of debian wheezy - memcheck is complete: Success [build successful]  Build details are at http://buildbot.shogun-toolbox.org/builders/debian%20wheezy%20-%20memcheck/builds/916:56
@iglesiasglambday_, I had noticed heavensdevil, but didn't the numbers :P16:56
-!- lambday_ [67157d37@gateway/web/freenode/ip.103.21.125.55] has quit [Ping timeout: 250 seconds]17:00
-!- lambday [67157c37@gateway/web/freenode/ip.103.21.124.55] has joined #shogun17:01
lambdayiglesiasg: :D17:01
lambdayI am going for dinner.. will come back in a while17:01
lambdaysee you :)17:01
-!- lambday [67157c37@gateway/web/freenode/ip.103.21.124.55] has quit [Client Quit]17:01
-!- van51 [~van51@athedsl-409794.home.otenet.gr] has joined #shogun17:02
-!- HeikoS [~heiko@nat-163-62.internal.eduroam.ucl.ac.uk] has joined #shogun17:04
-!- mode/#shogun [+o HeikoS] by ChanServ17:04
-!- lisitsyn [~lisitsin@mxs.kg.ru] has quit [Quit: Leaving.]17:07
@HeikoSwiking: wow by changes screw up a lot of things ;)17:09
shogun-buildbotbuild #9 of clang34 - thread analysis is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/clang34%20-%20thread%20analysis/builds/9  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>, lambday <heavensdevil6909@gmail.com>17:10
@HeikoSwiking: fixed17:29
@HeikoStests pass locally17:29
@HeikoSiglesiasg: your lmnn modular integration tests fail on my machine still17:36
@iglesiasgHeikoS, :O17:36
@iglesiasgHeikoS, latest develop?17:37
@HeikoSiglesiasg: yes, have you tested?17:37
@iglesiasgyes, I did test yesterday when I pushed it17:37
shogun-notifier-shogun: Heiko Strathmann :develop * c5dbcb4 / src/shogun/ (5 files): https://github.com/shogun-toolbox/shogun/commit/c5dbcb49d370180f24441d2e2367d0dccd74b9b917:37
shogun-notifier-shogun: -added UNDEFINED cases in data type enums to avoid unpredictable behaviour.17:37
shogun-notifier-shogun: -Many switch statements did not handle default, fixed that17:37
shogun-notifier-shogun: Heiko Strathmann :develop * 26f4434 / src/shogun/base/ParameterMap.cpp: https://github.com/shogun-toolbox/shogun/commit/26f443420c27a8b8a3d0d57e19ed7a5d3553e5d917:37
shogun-notifier-shogun: -added enums to is_empty check of SGParamInfo17:37
shogun-notifier-shogun: -to_string method now doesnt throw an error when data types are undefined17:37
shogun-notifier-shogun: Heiko Strathmann :develop * 1983e95 / src/shogun/lib/DataType.cpp: https://github.com/shogun-toolbox/shogun/commit/1983e95f954c8e2475fd60eaef6e12692d44b9b217:37
shogun-notifier-shogun: same as before17:37
shogun-notifier-shogun: Heiko Strathmann :develop * 5a300a2 / src/shogun/ (5 files): https://github.com/shogun-toolbox/shogun/commit/5a300a210492f1f11152abb50e3158a0e5afe3c017:37
@iglesiasgit was fine both here and in travis17:37
shogun-notifier-shogun: Merge pull request #1631 from karlnapf/develop17:37
shogun-notifier-shogun:17:37
shogun-notifier-shogun: added undefined cases for data type enums17:37
@HeikoSwiking: ^17:37
@HeikoSiglesiasg: mmm17:37
@HeikoSiglesiasg: maybe my data is out of date, let me check17:37
@iglesiasgHeikoS,  ok, let me know please17:41
@HeikoSiglesiasg: nope, passes, forry!17:42
@iglesiasgHeikoS, :)17:42
-!- zxtx [~zv@149.165.244.2] has quit [Ping timeout: 268 seconds]17:50
shogun-buildbotbuild #17 of clang34 - static analysis is complete: Failure [failed analyse]  Build details are at http://buildbot.shogun-toolbox.org/builders/clang34%20-%20static%20analysis/builds/17  blamelist: lambday <heavensdevil6909@gmail.com>17:54
besser82HeikoS: you got my messages from yesterday?17:55
@HeikoSbesser82: nope sorry, computer crashed in the evening,17:55
@HeikoSbesser82: ah I read the cmake thing17:55
@HeikoSbesser82: you are going to refactor it?17:55
@HeikoSwhats the motivation?17:55
-!- FSCV [~FSCV@66.187.72.219] has joined #shogun17:56
besser82HeikoS: some quirks on different dists with different modules17:56
@HeikoSI see17:56
@HeikoSbesser82: cool!17:56
besser82HeikoS: your ATLAS-problem btw is caused by a mad atlas-lib your admin brewed :(17:56
@HeikoSI like that  alot17:56
@HeikoSbesser82: I got around that finally17:56
@HeikoSbesser82: using your classic thing to compile shogun17:56
besser82HeikoS: allright17:56
@HeikoSand virtual-env'ed a custom python17:57
@HeikoSwith all modules17:57
besser82HeikoS: :D17:57
@HeikoSI need17:57
@HeikoSannoying, but ok, at least I am independent now17:57
@HeikoSbesser82: btw stripping libshogun did not make eigen3's sparse module faster17:59
@HeikoSThey actually know that they are half as fast as choldmo18:00
@HeikoSd18:00
@HeikoSBut installing cholmod on this cluster seems impossible, tried many things and failed. The admins also dont get it working18:00
@HeikoSso I guess i have to rely on eigen318:00
@HeikoSgood thing is: it now finally works! :LD18:00
besser82HeikoS: Sry, then. but was worth a try :)18:00
shogun-buildbotbuild #2222 of deb1 - libshogun is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb1%20-%20libshogun/builds/2222  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>18:01
shogun-buildbotbuild #2221 of deb1 - libshogun is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb1%20-%20libshogun/builds/2221  blamelist: Heiko Strathmann <heiko@Computer.(none)>18:02
-!- FSCV [~FSCV@66.187.72.219] has quit [Ping timeout: 264 seconds]18:09
-!- foulwall [~zhengyang@114.255.40.22] has joined #shogun18:17
foulwallping iglesiasg18:17
@iglesiasgfoulwall, hey18:17
@iglesiasgfoulwall, how are you doing?18:17
foulwalliglesiasg: :)18:17
van51foulwall: hey!18:19
foulwallHey van51 :)18:20
van51foulwall: could you take a look at : https://github.com/shogun-toolbox/shogun-demo/pull/2418:20
van51foulwall: I'd like to know if you have a problem with disabling autoescaping18:20
@iglesiasgfoulwall, so would it be good for you if I put all my code in another repository?18:20
@iglesiasgfoulwall, I think it is probably the most comfortable option18:20
@iglesiasgfoulwall, and you can just write comments in github18:20
foulwallok van5118:21
@iglesiasgfoulwall, so I am checking how comments work in github18:22
foulwallI have read hushell's suggestion, and it's ok without creating another repo.18:22
@iglesiasgfoulwall, it seems it is not possible to comment in files directly but one can comment in commits18:22
foulwallAh, I forgot the comment,18:23
@iglesiasgfoulwall, so I can do several commits18:23
@iglesiasgfoulwall, one for the method itself, another one for some examples, etc18:23
@iglesiasgfoulwall, and you can put particular stuff associated with a line or a method, etc in there18:23
@iglesiasgfoulwall, and more general stuff in an issue18:23
foulwallThank you:)18:23
foulwallThat's pretty convenient18:24
@iglesiasgfoulwall, awesome then18:24
@iglesiasgfoulwall, are you ok doing the same for me to review your code?18:24
foulwallI'll have a look at your repo. and for me, just shogun-demo? Maybe tommorrow after I update the asp/arts demo and hierarchical clustering example?18:25
@iglesiasgfoulwall, aaah true you worked in another repo18:26
@iglesiasgfoulwall, so basically all your work is in shogun-demo?18:26
@iglesiasgor should I also look at something in shogun-web?18:26
foulwalliglesiasg: All my work are wiped out from shogun-web18:27
@iglesiasgfoulwall, mmm ok. I understand I just have to look at shogun-demo then18:28
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has joined #shogun18:28
@iglesiasgfoulwall, I will prepare the repository then either tomorrow or on Saturday morning18:30
@iglesiasgfoulwall, so that is like between 1 day or 1.5 days from now18:30
@iglesiasgfoulwall, is that good for you?18:30
-!- foulwall` [~user@114.255.40.14] has joined #shogun18:31
-!- thoralf [~thoralf@enki.zib.de] has quit [Quit: Konversation terminated!]18:32
-!- foulwall [~zhengyang@114.255.40.22] has quit [Ping timeout: 245 seconds]18:34
foulwall`iglesiasg: okey:) and I've done a simple mldata.org downloader. That's quite a toy stuff. Here is a commit: https://github.com/shogun-toolbox/shogun/commit/cd9c28e592a8b660d48f8120f3ce6aa2456e2dcf18:35
foulwall`sorry for the network crashed again.18:35
@iglesiasgok18:36
@iglesiasgno problem18:36
-!- foulwall` [~user@114.255.40.14] has quit [Remote host closed the connection]18:37
-!- foulwall` [~user@greymonkey.im] has joined #shogun18:37
votjakovrHeikoS: hi! build_parameter_dictionary() method of SGObject now add to dictionary only exact parameters (not SGObject). GradientEvaluation class builds dictionary of current differentiable function lazy. Is it ok?18:37
@HeikoSvotjakovr: hi!18:37
@HeikoSvotjakovr: is that in your fork?18:38
@HeikoSvotjakovr: can I have a look?18:38
@HeikoSvotjakovr: seems ok to me, but I am not sure whether I completely get it18:38
votjakovrHeikoS: sorry not yet18:38
@HeikoSso let me understand that18:38
@HeikoSbefore, the parameter dictionary contained SGOjects?18:39
@HeikoSand now?18:39
@HeikoSlazy dictionary building is OK18:39
votjakovrHeikoS: contains only exact paramters18:39
@HeikoSwhat does that mean?18:39
@HeikoSvotjakovr: are there any problems that this might cause?18:40
votjakovrHeikoS: for example "sigma" of GaussianLikelihood18:40
@HeikoSah18:40
@HeikoSyeah thats ok for now18:40
@HeikoSvotjakovr: I mean if it causes problems we have to think about it more carefully later18:40
@HeikoSbut now the focus is on finishing18:40
@HeikoSso just go ahead with this18:40
@HeikoShow much longer do you think it will take?18:40
votjakovrHeikoS: yeah, i'll try to send PR tonight, late tomorrow morning18:41
@HeikoSvotjakovr: ok18:41
@HeikoSvotjakovr: what about your notebook18:41
foulwall`van51: :) Thanks for the pr. Sorry for I forget considering escape:( And I'll comment to ask sonney2k to merge.18:41
@HeikoSvotjakovr: I will be away on saturday/sunday, so tomorrow is the last chance to get help from me18:41
@HeikoSI can still review though18:42
@HeikoSbut not edit myself18:42
votjakovrHeikoS: ok, thank you very much :)18:42
@HeikoSvotjakovr: so when will you send your draft?18:42
van51foulwall`: ok, thanks! I'm glad that you approve :)18:43
foulwall`van51: Seen your notes, I'll add a section of the class/author link.18:43
votjakovrHeikoS: i'll try to do it tomorrow18:43
van51foulwall`: cool! although it doesn't have to be something separate, it can be linked directly from where they are mentioned in the about18:45
foulwall`van51: I'll to add some style on that, so plain there18:46
foulwall`iglesiasg: okey:)18:47
van51foulwall`: ok then, as you wish :)18:47
@iglesiasgfoulwall`, so it is set now, I hope it is not a pain to review my code :P18:47
foulwall`haha iglesiasg18:49
-!- foulwall` [~user@greymonkey.im] has quit [Remote host closed the connection]18:50
@HeikoSvotjakovr: okay, thanks. Pls keep in mind that it is fine to send a PR with an incomplete draft. I will just extend it during the day tomorrow.18:50
wikingHeikoS:yes18:57
wikingHeikoS: did u fix it?18:57
@HeikoSwiking: ehm yes what?18:57
@HeikoSwiking: yes all fine now18:57
wikingHeikoS: ok cool18:57
@HeikoSwiking: waiting for travis, but I checked everything locally and merged so that you can go on18:57
@HeikoSah travis says green18:57
@HeikoSfor libshogun18:57
@HeikoSmodular red18:57
wikinglet's see18:58
@HeikoSbut these are other people errors18:58
wikingshogun-buildbot: force build --branch=develop 'deb1 - libshogun'18:58
shogun-buildbotbuild #2223 forced18:58
shogun-buildbotI'll give a shout when the build finishes18:58
@HeikoSiglesiasg: lmnn failed, but I think thats fine since he fixed it18:58
@HeikoSwiking: quite tricky to change the stuff18:58
@HeikoSas I did some very nasty things with these uninitialised values18:58
@iglesiasgHeikoS, what's up with LMNN? Why does it fail still?18:58
@HeikoSbut now better, just had to handle them in switch18:59
@HeikoSiglesiasg: I dont know, it passed locally here18:59
wikingHeikoS: now i'll be interested in the undefined behaviour bot's output18:59
shogun-buildbotbuild #2223 of deb1 - libshogun is complete: Success [build successful]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb1%20-%20libshogun/builds/222318:59
wikingshogun-buildbot: force build --branch=develop 'clang34 - undefined behaviour analysis'19:00
shogun-buildbotbuild #5 forced19:00
shogun-buildbotI'll give a shout when the build finishes19:00
wikinggsomix ?19:00
-!- travis-ci [~travis-ci@ec2-54-234-21-192.compute-1.amazonaws.com] has joined #shogun19:07
travis-ci[travis-ci] it's Heiko Strathmann'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/1155038719:07
-!- travis-ci [~travis-ci@ec2-54-234-21-192.compute-1.amazonaws.com] has left #shogun []19:07
shogun-buildbotbuild #5 of clang34 - undefined behaviour analysis is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/clang34%20-%20undefined%20behaviour%20analysis/builds/519:09
wikingok i think i have ComplexVector working :)19:10
wikingHeikoS: do u know what else do we need?19:10
wikingiglesiasg: you have some weird stuff going on in lmnn19:10
wikingalthough that's more like an eigen problem as well19:11
wikingbut worth to investigate19:11
@iglesiasgwiking, mmm19:11
@iglesiasgwiking, there is nothing random AFAIK19:11
@iglesiasgwiking, what does it happen with the test? It fails sometimes and sometimes not?19:11
wikingnono19:11
wikingiglesiasg: http://buildbot.shogun-toolbox.org/builders/clang34%20-%20undefined%20behaviour%20analysis/builds/5/steps/test/logs/stdio19:11
wikingsearch for LMNNImpl.find_impostors_exact19:12
wikingand you'll see19:12
wikingHeikoS: around?19:12
@iglesiasgwiking, wow19:18
@iglesiasgwiking, wtf is that19:18
wikingiglesiasg: hehehe ... heavy shit19:18
wikingiglesiasg: you are not the only one who's making this mistake19:18
wikingbut it'd be great to find out how we could fix it19:19
@iglesiasgwiking, but the test does not fail anyway?19:19
@iglesiasgfunny19:19
wikingiglesiasg: well it's an alignment issue19:19
wikingso most of the time i think it's fine19:19
wikingbut i guess there's a machine + os + compiler combo on the world19:19
@iglesiasgwiking, I am writing an issue for me so I don't forget to have a look19:19
wikingwhere this would result in a segfault19:19
wikingiglesiasg: yeah go ahead... i'm just checking now the other issues19:20
@iglesiasgbut I guess I will have a look in some time19:20
wiking/home/buildslave/clang34_-_undefined_behaviour_analysis/build/build/src/shogun/libshogun.so.14:0x47c3706: runtime error: value -inf is outside the range of representable values of type 'int'19:20
wiking:)19:20
wikingso there's some weird shit going on19:20
votjakovrHeikoS: May i temporally move to MS_NOT_AVALIABLE m_compact parameter of GaussianKernel and combined_kernel_weight of Kernel?19:24
votjakovrHeikoS: there is no code to compute derivatives wrt these parameters19:26
votjakovrHeikoS: and i'd like to temporally disable them for model selection19:27
-!- iglesiasg [~iglesias@2001:6b0:1:1da0:247a:a4fc:e52c:8133] has quit [Ping timeout: 259 seconds]19:37
@HeikoSvotjakovr: no please dont, can you do another solution?19:40
@HeikoSvotjakovr: like some return value that tells you that there are no gradients available?19:40
-!- iglesiasg [~iglesias@2001:6b0:1:1da0:247a:a4fc:e52c:8133] has joined #shogun19:41
-!- mode/#shogun [+o iglesiasg] by ChanServ19:41
votjakovrHeikoS: ok19:41
wikingHeikoS: so as i was saying i'm preparing now the complex64_t mapping for lambday. do you know what he exactly would need?19:42
wikingHeikoS: i have now sgvector<complex> sgmatrix<complex> sgsparsevector<complex> and sgsparsematrix<complex>19:42
@HeikoSwiking: not really, but basically he needs shogun's complex64 to numpy complex19:42
@HeikoSand this for SGVector, SGSparseMatrix19:42
wikingok19:43
wikinglet's hope that this is what he needs19:43
wikingand that it works19:43
@HeikoSsparse vector is a good idea, but I dont htink he uses that19:43
@HeikoSwiking: yeah :)19:43
wikingi've tested sgvector19:43
wikingthat works nicely19:43
wikingfor the others i need some examples to test19:44
votjakovrHeikoS: i think it's not really good19:44
@HeikoSvotjakovr: but disabling parameters just for GPs is neither19:45
@HeikoSpeople wont be able to do grid-search for them19:45
@HeikoSvotjakovr: also, these are discrete parameters so no derivative will ever be able to be computed19:45
@HeikoSvotjakovr: and for this, we cannot kick them out19:45
votjakovrHeikoS: yeah unfortunately19:46
@HeikoSvotjakovr: so maybe a call by reference paramter which is a boolean flag which tells you whether its possible to compute the derivate?19:47
votjakovrHeikoS: but we could add something like GRID_SEARCH_AVALIABLE ?19:47
@HeikoSvotjakovr: and its false by default, only true if there is an implementation19:47
@HeikoSyeah something like this19:47
@HeikoSbut I would rather go for GRADIENT available19:48
@HeikoSsince grid-search is more general19:48
votjakovrHeikoS: oops yeah19:48
votjakovrHeikoS: so they should be registered like: SG_ADD(..., GRADIENT_SEARCH_AVALIABLE), right?19:49
@HeikoSvotjakovr: mmmh19:50
@HeikoSvotjakovr: I dont like explicit things19:50
@HeikoSsince then people have to register them with gradient AND implement gradient function19:50
@HeikoSrather have a gradient function that tells you NO, and then devs might overload it to say YES, HERE IS YOUR GRADIENT19:51
votjakovrHeikoS: ok19:54
shogun-notifier-shogun: Viktor Gal :develop * 5592220 / / (4 files): https://github.com/shogun-toolbox/shogun/commit/5592220bd062571ddfb02817bf3023aada143e8619:55
shogun-notifier-shogun: WIP: adding complex64_t <-> numpy complex mapping (issue #1493)19:55
votjakovrHeikoS: this should be a member of TParameter, right?20:00
@iglesiasgall right guys, see you later20:10
-!- iglesiasg [~iglesias@2001:6b0:1:1da0:247a:a4fc:e52c:8133] has quit [Quit: Ex-Chat]20:10
@HeikoSvotjakovr: no, this should be in the class that offers the derivative20:22
votjakovrHeikoS: mmmh, GaussianKernel offers derivative wrt width. I need to choose width from all of m_model_selection_parameters of GaussianKernel only width20:27
@HeikoSvotjakovr: I dont get that, could you explain again?20:28
votjakovrHeikoS: for example i get GaussianKernel and i'd like to build parameter dictionary, which contains only the parameters which we can use, when we want to compute derivative ("width" parameter in this case). How is it better to do this?20:30
shogun-buildbotbuild #104 of precise - libshogun is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/precise%20-%20libshogun/builds/104  blamelist: Viktor Gal <viktor.gal@maeth.com>20:34
@HeikoSvotjakovr: I see20:35
@HeikoSvotjakovr: so there are two problems:20:35
@HeikoS1) is the parameter continous20:35
@HeikoS2) if yet, is the derivative implemented20:35
@HeikoSif 1) is false, you dont ask for derivative20:35
@HeikoSif you ask for derivative 2), then it might not be implemented20:36
@HeikoSvotjakovr: so what about saying that every float parameter can have derivatives20:36
@HeikoSvotjakovr: this is the only thing that makes sense, 1) is solved20:36
@HeikoSfor 2) we need some mechanism that a class knows which derivatives it offers20:36
@HeikoSbut I would like to not do this via the SG_ADD macro20:37
@HeikoSbut rather by a local map in CSGObject20:37
@HeikoSor even better20:37
@HeikoSsomething that automagically says YES if a method is implemented20:37
@HeikoSso that in order to add a derivate, you only have to implement/overload a method20:38
votjakovrHeikoS: we have just one method20:39
votjakovrHeikoS: which gets parameter as argument20:39
votjakovrHeikoS: and returns derivative wrt given parameter20:39
votjakovrif there is no derivatives wrt given parameter it shows an error20:40
@HeikoSvotjakovr: what about a flag in that method then?20:40
@HeikoSvotjakovr: std implementation returns "empty derivative" and set the flag to false20:40
@HeikoSif a dev overloads, it returns the real derivate and sets flag to true20:40
-!- travis-ci [~travis-ci@ec2-54-242-103-128.compute-1.amazonaws.com] has joined #shogun20:41
travis-ci[travis-ci] it's Heiko Strathmann'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/1155701320:41
-!- travis-ci [~travis-ci@ec2-54-242-103-128.compute-1.amazonaws.com] has left #shogun []20:41
@HeikoSthis way, you can just check whether the derivative exists via trying to compute it20:41
@HeikoSthe only downside is that you always have to compute the derivative if you just want to know whether thats possible. But that should not be a problem20:41
votjakovrHeikoS: i actually don't need to check20:41
votjakovrHeikoS: i need the parameters20:41
votjakovrHeikoS: to build a dictionary20:42
@HeikoSvotjakovr: ah so you just need to know the parameters that offer a derivative?20:42
votjakovrHeikoS: yeah20:42
@HeikoSvotjakovr: mmh, so this cannot be solver via overloading the method then20:42
@HeikoSvotjakovr: needs to be set by hand20:43
@HeikoSso a subset of the modelselection parameters would have also to be derivative parameters20:43
votjakovrHeikoS: yeah20:43
@HeikoSvotjakovr: what about adding another Parameter instance to CSGObject then?20:43
@HeikoSvotjakovr: or in fact, maybe even your flag in SG_ADD20:43
@HeikoSvotjakovr: would there be any other cases?20:44
votjakovrHeikoS: that what i suggested :)20:44
@HeikoSbecause we dont want to pollute the SG_ADD too much20:44
@HeikoSmost classes wont have derivatives20:44
votjakovrHeikoS: SG_ADD(..., GRADIENT_SEARCH_AVAILABLE)20:44
@HeikoSvotjakovr: I think another Parameter instance would be best20:44
@HeikoSvotjakovr: no, too much writing and most classes dont have this20:44
@HeikoSmmmh20:45
@HeikoSsonney2k: thoughts on this?20:45
@HeikoSvotjakovr: macros do not support default arguments right?20:45
@HeikoSsince I dont want t rouch all SG_ADD lines in shogun20:46
votjakovrHeikoS: unfortunately20:46
@HeikoSonly the ones where a derivative is there20:46
@HeikoSsonne|work: thoughts on this? ^20:46
@HeikoSvotjakovr: mmh tough decision20:47
@HeikoSvotjakovr: wait20:47
@HeikoSif you just want to know which parameters are continous, you are just interested in those which potentially can have derivatives right?20:48
@HeikoS*potentially*20:48
@HeikoSbut the impleentation still can throw an error20:48
@HeikoSyou just have to implement it if you want to do gradient model-selection20:48
@HeikoSso all float* parameters could be added,20:48
@HeikoSso maybe its possible to add a method to CSGObject that tells you the parameters that are continous?20:49
@HeikoSvotjakovr: and the othher possibility would be to actually let the user provide a parameter tree with the parameters20:49
@HeikoSvotjakovr: time is tight, maybe we should go with this solution for now, its ugly, but the other thing takes too long20:49
@HeikoSvotjakovr: your problem is solved if users explicitly provide parameters for gradient search right?20:49
@HeikoSyou can add the other thing later maybe?20:50
votjakovrHeikoS: i need to build dictionary anyway20:50
votjakovrHeikoS: since gradient result is using it20:51
@HeikoSvotjakovr: but cant you build it from a ModelSelectionParameter tree?20:52
votjakovrHeikoS: why not to add m_gradient_parameters to SGObject?20:52
@HeikoSvotjakovr: maybe you are right20:53
votjakovrHeikoS: add third option GRADIENT_SEARCH_AVAILABLE20:53
@HeikoSvotjakovr: but lets think about one thing first20:53
votjakovrHeikoS: and if it is enabled we add parameter to m_parameters m_model_selection_parameters and m_gradient_parameters20:53
@HeikoSvotjakovr: in fact, do it. Replace all SG_ADD calls in shogun with GRADIENT_SEARCH_AVAILABLE=false20:54
@HeikoSand then set it to true for the ones you want20:54
@HeikoSvotjakovr: though a default argument would be better20:54
@HeikoSvotjakovr: i.e. SG_ADD(bla, MS_NOT_AVAILABLE) and SG_ADD(bla, MS_NOT_AVAILABLE, GRADIENT_NOT_AVAILABLE) being equivalent20:54
@HeikoSso that one only has to explicitly say when one wants gradients20:55
@HeikoSagreed?20:55
votjakovrHeikoS: yep20:56
@HeikoSvotjakovr: I dont know if its possible to overload macros, but try, if not we have to add it to all SG_ADD for now20:56
@HeikoSand a third field in CSGObject20:56
votjakovrHeikoS: ok, i think this is temporal solution, right? I mean, will we do some improvement of parameter and modsel in future?20:59
@HeikoSvotjakovr: yeah, I still like to keep things as clean as possible since doing things later doesnt work most of the time ;)20:59
@HeikoSvotjakovr: but go ahead20:59
@HeikoSthis is a good solution in fact20:59
@HeikoSwe just should think about ways of overloading the macro21:00
votjakovrHeikoS: have a look: http://stackoverflow.com/questions/8886812/overload-c-macros21:02
@HeikoSvotjakovr: okay, if you can make this work, it would be great, but dont spend too mnuch time on macro magic ;)21:03
shogun-buildbotbuild #1500 of cyg1 - libshogun is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/cyg1%20-%20libshogun/builds/1500  blamelist: Viktor Gal <viktor.gal@maeth.com>21:05
@HeikoSvotjakovr: ok, sounds like a good plan, I will go home now, see you tomorrow! looking forward to the PR, this is really going to be good stuff  :)21:07
votjakovrHeikoS: see you ;)21:07
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has quit [Quit: Good bye!]21:21
-!- van51 [~van51@athedsl-409794.home.otenet.gr] has quit [Quit: Leaving.]21:26
-!- lisitsyn [~lisitsyn@fb2-lo1.global63.net] has joined #shogun21:28
-!- travis-ci [~travis-ci@ec2-54-234-21-192.compute-1.amazonaws.com] has joined #shogun21:35
travis-ci[travis-ci] it's Viktor Gal'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/1156225721:35
-!- travis-ci [~travis-ci@ec2-54-234-21-192.compute-1.amazonaws.com] has left #shogun []21:35
-!- lisitsyn1 [~lisitsyn@fb2-lo1.global63.net] has joined #shogun21:48
-!- lisitsyn [~lisitsyn@fb2-lo1.global63.net] has quit [Read error: Connection reset by peer]21:48
shogun-buildbotbuild #18 of clang34 - static analysis is complete: Failure [failed analyse]  Build details are at http://buildbot.shogun-toolbox.org/builders/clang34%20-%20static%20analysis/builds/18  blamelist: Viktor Gal <viktor.gal@maeth.com>21:52
-!- lisitsyn [~lisitsyn@fb2-lo1.global63.net] has joined #shogun21:56
-!- lisitsyn1 [~lisitsyn@fb2-lo1.global63.net] has quit [Read error: Connection reset by peer]21:56
wikinglool22:03
wikingwhy dont we ever try to integrate this: https://projects.coin-or.org/Osi22:03
wiking?22:03
-!- iglesiasg [~iglesias@s83-179-44-135.cust.tele2.se] has joined #shogun22:04
-!- mode/#shogun [+o iglesiasg] by ChanServ22:04
wikingit's like having a million of solvers22:04
shogun-buildbotbuild #6 of clang34 - undefined behaviour analysis is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/clang34%20-%20undefined%20behaviour%20analysis/builds/6  blamelist: Viktor Gal <viktor.gal@maeth.com>22:04
wikinginterfaced as one22:05
wikingiglesiasg: i have a question22:05
@iglesiasgI might have an answer22:05
@iglesiasgwiking, shoot22:05
wikingm_rescode = MSK_putavec(m_task, MSK_ACC_CON, con_idx, nnz+1,22:05
wikingasub.vector, aval.vector);22:05
wikingReplaces all elements in one row of A.?22:05
@iglesiasgok it is a exercise of memory22:05
* wiking is trying to update the interface for mosek22:06
@iglesiasgyes22:06
@iglesiasgit should do that22:06
@iglesiasgI am almost 90% confident it does that22:06
wikingok cool22:07
@iglesiasgwiking, new Mosek version?22:07
wikingiglesiasg: yeah 722:08
wikingi'm now trying to find an equivalent of MSK_putaveclist22:09
@iglesiasgI am so looking forward to have some time and substitute that for another solver22:09
-!- lisitsyn [~lisitsyn@fb2-lo1.global63.net] has quit [Read error: Connection reset by peer]22:09
-!- lisitsyn1 [~lisitsyn@fb2-lo1.global63.net] has joined #shogun22:09
@iglesiasgwhen I talked to Nico in the workshop22:09
@iglesiasghe said it should be possible to do subgradient descent and use it instead of the Mosek solver we are using right now22:09
@iglesiasgwiking, do you think it makes sense too?22:10
@iglesiasgand I mean, gradient descent it is not hard. Even if Mosek solver is faster, it probably pays off just to get rid of this shit22:10
wikingiglesiasg: Osi: https://projects.coin-or.org/Osi22:11
wikingiglesiasg: it is interfacing a lot of solvers22:11
wiking(mosek included)22:11
@iglesiasgwiking, LP22:11
wikingmmm fuck it only interfaces LP22:12
wikingno QP?22:12
@iglesiasgwiking, we are using mosek for a QP22:12
@iglesiasgthey only mention LP so far22:12
@iglesiasglet me check22:12
@iglesiasgI think so, yeah22:12
wikinghttps://projects.coin-or.org/OptiML22:13
@iglesiasgA'x=022:13
wikinghttp://www.coin-or.org/projects/22:14
@iglesiasgit is like libqp22:14
wikingi'm just checking there what they have there22:14
wikingseems like heaps of stuff22:14
@iglesiasgfor the SO-SVM we need A'x<=b22:14
@iglesiasgokok22:14
-!- lisitsyn1 [~lisitsyn@fb2-lo1.global63.net] has quit [Quit: Leaving.]22:14
@iglesiasgwiking, still. I think it might be good just to code up a solver for that22:16
-!- besser82 [~besser82@fedora/besser82] has quit [Quit: freedom, friends, features, first ---> fedoraproject.org]22:18
wiking This routine finds a KT point for the Quadratic LCP (QP) problem22:18
wiking1922:18
wiking20c       minimize    f(x) = ct.x + xt.G.x/222:18
wiking2122:18
wiking22c       subject to  l <= [I : A]t.x <= u22:18
wikingoh come oon man22:18
wikingthis guy if fucking crazy22:18
wikinghe wrote this solver in fortran22:18
shogun-buildbotbuild #10 of clang34 - thread analysis is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/clang34%20-%20thread%20analysis/builds/10  blamelist: Viktor Gal <viktor.gal@maeth.com>22:18
@iglesiasghaha22:19
@iglesiasgwiking,  I never did fortran. Is it more hardcore than C?22:19
@iglesiasgI guess not22:19
wikingit is fucking hardcore22:19
@iglesiasgI mean, it is not assembler, right?22:19
@iglesiasgwiking, have you seen our BMRM?22:19
@iglesiasgthat is hardcore I think hehe22:19
wikingiglesiasg: hehe no its cool22:20
wikingi've even fixed some shit in it22:20
wikingand then did some extension of it in libncbm22:20
wiking:P22:20
@iglesiasgnice22:21
wikingok so i wonder what could be the replacement of that MSK_putaveclist22:22
@iglesiasgcheck for the methods to add constraints22:22
@iglesiasgit is basically that22:23
@iglesiasglinear inequality constraint22:23
@iglesiasgthat's how they were called in Mosek's doc IIRC22:23
wikingChanges the bounds of a list of constraints.22:23
wiking?22:23
wikingcould it be this?22:23
@iglesiasgnot really22:23
@iglesiasgA x <= b22:23
@iglesiasgthat one you mention is to change the b, I think22:24
wikingyes22:24
wikingand u change with aveclist?22:24
wikingx?22:24
wikingi mean A's22:24
@iglesiasgno no22:24
wikingrow22:24
@iglesiasgyes yes22:24
@iglesiasga row of A22:24
@iglesiasgwiking, do you have the doc page?22:25
wikingah and it replaces n number of elements22:25
@iglesiasgI can try to help to find it22:25
wikingfor the new api?22:25
@iglesiasgmm n number of elements22:25
wikinghttp://docs.mosek.com/7.0/capi/Changes_and_new_features_in_MOSEK.html22:26
@iglesiasgso that method was done to dine a row of A22:26
@iglesiasgbut you do it in a sparse way22:26
@iglesiasgso you give22:26
@iglesiasg1) the number of values in a row you want to change22:26
@iglesiasg2) the indices within the row, i.e. col idxs22:26
@iglesiasg3) the actual values22:26
@iglesiasgwiking,  http://docs.mosek.com/7.0/capi/MSK_putarow_.html?22:27
@iglesiasgwiking, to me it looks like it must be that one22:28
@iglesiasgwiking, now you have m_rescode = MSK_putavec(m_task, MSK_ACC_CON, con_idx, nnz+1,22:28
@iglesiasg    asub.vector, aval.vector);22:28
@iglesiasgI think it must be pretty much like22:29
@iglesiasgm_rescode = MSK_putarow(m_task, con_idx, nnz+1,  asub.vector, aval.vector);22:29
@iglesiasgnot sure right now why the +1 nnz+1 but with any luck I wrote a comment for that22:30
-!- van51 [~van51@athedsl-409794.home.otenet.gr] has joined #shogun22:39
-!- lisitsyn [~lisitsyn@fb2-lo1.global63.net] has joined #shogun22:49
-!- iglesiasg_ [~iglesias@s83-179-44-135.cust.tele2.se] has joined #shogun22:51
-!- iglesiasg [~iglesias@s83-179-44-135.cust.tele2.se] has quit [Quit: Ex-Chat]22:51
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout]22:55
wiking?23:00
iglesiasg_wiking, is that for me?23:00
-!- iglesiasg_ is now known as iglesiasg23:00
-!- mode/#shogun [+o iglesiasg] by ChanServ23:01
wikingi was looking for replacing MSK_putaveclist23:02
wikingnot MSK_putavec23:02
@iglesiasgwell then MSK_putarowlist23:02
wikingthat'll be good ! thnx23:07
wikingok only one more is missing MSK_getsolutionstatus23:09
wikingok this should be it23:11
wikingMSK_getprosta23:11
@iglesiasgwelcome23:12
wikingoh damn23:12
@sonney2kwiking, btw we cannot use any of coin-or due to licence conflicts23:12
wikinghttp://docs.mosek.com/7.0/capi/Problem_status_keys.html23:12
wikingthey are toootaly different23:13
wikingwhy they use what?23:13
@sonney2kwiking, they used to use CPL23:15
@sonney2kI forgot what they switched to23:16
@sonney2kbut it was again in conflict to GPL23:16
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun23:21
shogun-notifier-shogun: Soeren Sonnenburg :develop * 6197bb4 / examples/undocumented/python_modular/ (6 files): https://github.com/shogun-toolbox/shogun/commit/6197bb4ab279aec6cb3d84c6c5d6fbe87ccf22a823:21
shogun-notifier-shogun: convert more examples to CSVFile23:21
shogun-notifier-shogun-demo: van51 :master * 2bd9761 / / (3 files): https://github.com/shogun-toolbox/shogun-demo/commit/2bd9761c84f52dfae94fdce22bd1c504b62b0bd123:23
shogun-notifier-shogun-demo: Changed prompt and about in language detection demo23:23
shogun-notifier-shogun-demo: Soeren Sonnenburg :master * 6569227 / / (3 files): https://github.com/shogun-toolbox/shogun-demo/commit/6569227662c69ce5645d298cd82eeb6676f2701d23:23
shogun-notifier-shogun-demo: Merge pull request #24 from van51/master23:23
shogun-notifier-shogun-demo:23:23
shogun-notifier-shogun-demo: Changed prompt and about in language detection demo23:23
shogun-buildbotbuild #105 of precise - libshogun is complete: Success [build successful]  Build details are at http://buildbot.shogun-toolbox.org/builders/precise%20-%20libshogun/builds/10523:49
shogun-buildbotbuild #1849 of deb3 - modular_interfaces is complete: Failure [failed test python modular]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb3%20-%20modular_interfaces/builds/1849  blamelist: Soeren Sonnenburg <sonne@debian.org>23:56
--- Log closed Fri Sep 20 00:00:06 2013