Open in new window / Try shogun cloud
--- Log opened Tue Mar 11 00:00:33 2014
-!- thoralf [] has joined #shogun00:02
-!- bumblebee [~quassel@] has joined #shogun00:04
-!- bumblebee is now known as Guest9610900:04
-!- Guest96109 [~quassel@] has quit [Remote host closed the connection]00:05
@HeikoSthoralf: heya00:05
thoralfStill stammtisch, right?00:08
thoralfIt's so quiet in here...00:08
-!- HeikoS [] has quit [Ping timeout: 240 seconds]00:22
-!- HeikoS [] has joined #shogun00:29
-!- mode/#shogun [+o HeikoS] by ChanServ00:29
-!- HeikoS [] has quit [Quit: Leaving.]00:42
-!- shogun-notifier- [] has quit [Quit: transmission timeout]01:31
-!- thoralf [] has quit [Quit: Konversation terminated!]01:54
-!- khalednasr [2945ac5f@gateway/web/freenode/ip.] has joined #shogun02:14
-!- kislay [~kislay@] has joined #shogun02:38
-!- khalednasr [2945ac5f@gateway/web/freenode/ip.] has quit [Ping timeout: 245 seconds]03:06
-!- sunil1337 [0e8bd057@gateway/web/freenode/ip.] has joined #shogun03:18
-!- sunil1337 [0e8bd057@gateway/web/freenode/ip.] has quit [Ping timeout: 245 seconds]03:23
-!- abinash_panda [75ef5e6f@gateway/web/freenode/ip.] has joined #shogun03:30
-!- lambday [67157f54@gateway/web/freenode/ip.] has joined #shogun04:58
-!- dhruv13 [~dhruv13@] has quit [Ping timeout: 240 seconds]05:08
-!- dhruv13 [~dhruv13@] has joined #shogun05:41
-!- kislay [~kislay@] has quit [Ping timeout: 244 seconds]06:00
-!- lisitsyn [] has quit [Quit: Leaving.]06:03
-!- dhruv13 [~dhruv13@] has quit [Ping timeout: 265 seconds]06:11
-!- vivekrai [uid10305@gateway/web/] has joined #shogun06:52
-!- abinash_panda [75ef5e6f@gateway/web/freenode/ip.] has quit [Ping timeout: 245 seconds]07:03
-!- vasanth [~vasanth@] has joined #shogun07:39
-!- kislay [~kislay@] has joined #shogun07:48
-!- shogun-notifier- [] has joined #shogun08:09
shogun-notifier-shogun: dhruv13j :develop * e12675b / src/shogun/base/init.cpp:
shogun-notifier-shogun: Issue #1955; change loglevel through environment variable08:09
shogun-notifier-shogun: dhruv13j :develop * c5dc355 / src/shogun/base/init.cpp,src/shogun/base/init.h:
shogun-notifier-shogun: fixes #1955; added a function init_from_env08:09
shogun-notifier-shogun: dhruv13j :develop * e246070 / src/shogun/base/init.cpp:
shogun-notifier-shogun: fixes #1955; minor change08:09
shogun-notifier-shogun: Viktor Gal :develop * 3c6d101 / src/shogun/base/init.cpp,src/shogun/base/init.h:
shogun-notifier-shogun: Merge pull request #1965 from dhruv13J/develop08:09
shogun-notifier-shogun: Issue #1955; change loglevel through environment variable08:09
-!- kislay [~kislay@] has quit [Ping timeout: 240 seconds]08:15
-!- HeikoS1 [] has joined #shogun08:26
-!- HeikoS1 [] has quit [Client Quit]08:27
-!- kislay [~kislay@] has joined #shogun08:28
-!- kislay [~kislay@] has quit [Ping timeout: 240 seconds]08:45
-!- PirosB3 [~pirosb3@] has joined #shogun09:07
-!- kprah [67157f4c@gateway/web/freenode/ip.] has joined #shogun09:13
kprahwiking: hey! you there?09:14
@wikingkprah: sure09:19
@wikingkprah: if u wanna discuss things about that task i really suggest u to either use mailing list or that given issue09:19
-!- kprah [67157f4c@gateway/web/freenode/ip.] has quit [Ping timeout: 245 seconds]09:22
-!- vasanth [~vasanth@] has quit [Quit: Konversation terminated!]09:45
-!- lisitsyn1 [] has joined #shogun10:03
sonne|workbesser82: around?10:10
jiaolongHi, I got an error when trying graph factor model in ipython notebook: AttributeError: 'DualLibQPBMSOSVM' object has no attribute 'get_helper'10:47
jiaolongCan anybody help me?10:47
-!- HeikoS1 [] has joined #shogun11:05
-!- shogun-notifier- [] has quit [Quit: transmission timeout]11:09
-!- HeikoS1 [] has quit [Ping timeout: 244 seconds]11:29
-!- HeikoS1 [] has joined #shogun11:44
@besser82sonne|work: I'm here now  ;)11:48
sonne|workbesser82: hey!11:49
sonne|workbesser82: I did the shogun3.2.0 package upload11:49
sonne|workand did build everything with clang on ppc etc11:49
@besser82wiking: lemme see what i can do about replacing it with some cmake-magic;  having it not present makes build of doxy-docs b0rk3n...11:49
sonne|workand guess what11:49
sonne|worksame thing!11:49
@besser82sonne|work: numpy, i suppose :D11:50
@besser82sonne|work: for ppc11:50
@besser82sonne|work: 3.2.0 is FC testing-repos  ;)11:50
sonne|workwell there is only that additional thing python & numpy that we are linking too11:50
@besser82sonne|work: as I said numpy && ppc(64) == BOOM-BANG!  :D11:51
sonne|workno idea if this is ppc 64 though11:51
@besser82sonne|work: I assured that with redhat-releng folks11:51
@besser82sonne|work: it bangs on either11:51
@besser82sonne|work: and tht issue is know by numpy-devs11:51
sonne|workyeah ppc6411:52
@besser82sonne|work: Has been reported by rh-folks during development of RHEL6  ;)11:52
@besser82sonne|work: so no fix within last ~4 years  :(11:52
sonne|workhmmhh it should just be some asm code in there then11:53
@besser82sonne|work: yes, but numpy-devs said: "That's not our target audiance..."11:54
sonne|workok then so no ppc...11:54
@besser82sonne|work: to be serious:  Who uses numpy on ppc?!?11:54
sonne|workI did but loong time ago11:55
sonne|workon some ppc mac11:55
@besser82sonne|work: iMac G5 or sth?!?11:55
sonne|workmacbook pro titanium :)11:55
sonne|workthe name was powerbook11:55
@besser82sonne|work: yeah, i know these11:55
@besser82sonne|work: they have been the same gen as G5 iMac, afaik11:56
@besser82sonne|work: but who has on today?11:56
@besser82sonne|work: one11:56
@besser82sonne|work: it's like keeping compat with ie5  :-P11:56
@besser82sonne|work: not worth the trouble11:56
@besser82sonne|work: but other bindings should work on ppc11:57
lisitsyn1hey guys11:57
lisitsyn1as I am thiking of plugin stuff11:57
@besser82lisitsyn1, yes?11:57
lisitsyn1what platforms do we need to support11:57
sonne|workbesser82: swig 3.0 will be released this week11:57
lisitsyn1linux .so11:57
lisitsyn1windows .dll11:57
@besser82sonne|work: hot stuff!11:57
lisitsyn1mac .dylib11:57
lisitsyn1what else?11:57
@besser82lisitsyn1: osx, win, *nix11:58
lisitsyn1so .so, .dll and .dylib11:58
sonne|workso I can just get the packaging going11:58
lisitsyn1is this enough?11:58
@besser82lisitsyn1, I don't recall anything else in the wild11:58
lisitsyn1besser82: some solaris shit may be11:59
@besser82lisitsyn1, but some really exotic stuff not commonly used, but that isn't target audiance...11:59
lisitsyn1nobody cares about these I guess11:59
@besser82lisitsyn1, solaris is *.so too11:59
lisitsyn1then no idea11:59
sonne|worklisitsyn1: android native lib?11:59
@besser82lisitsyn1, covering those schould be enough11:59
lisitsyn1sonne|work: do they have some format on their own?12:00
@besser82sonne|work: don't be silly  ---> android  :-P12:00
sonne|worklisitsyn1: never built it for android12:00
@besser82sonne|work: smartphone isn't that powerful enough...12:00
sonne|workbesser82: why? some nice apps are possible12:00
sonne|workbesser82: oh sure it is nowadays12:00
@besser82sonne|work: but they can be build on native so-libs12:01
@besser82sonne|work: we don't need any speciuhl adroid stuff12:01
lisitsyn1sonne|work: can we handle some plugins repository? :D12:02
@besser82sonne|work: if one wants to use that quite productive, I bet he would do it on a jailbreak-android ;)12:02
@besser82lisitsyn1: steal that from ruby-gems  :-P12:02
lisitsyn1besser82: well I mean we need some hardware to run this on12:02
@besser82lisitsyn1: do we?12:02
lisitsyn1besser82: any way around that?12:03
@besser82lisitsyn1: we need some httpd, but not really more  ;)12:03
@besser82lisitsyn1, around what?12:03
lisitsyn1besser82: yeah but still12:03
@besser82lisitsyn1, btw. on which laptop did backlight work with FC20???12:03
lisitsyn1besser82: did not work you mean?12:03
lisitsyn1samsung np53012:03
sonne|worklisitsyn1: that would be cool!12:03
lisitsyn1sonne|work: I am now developing some system to do that12:04
lisitsyn1next we just integrate it and make first plugin12:05
sonne|workbesser82, lisitsyn1well I think some android based port would be nice for sure12:05
lisitsyn1and see how it goes12:05
sonne|worklisitsyn1: 'just' do it12:05
@besser82lisitsyn1, yes, did not work?!?12:05
lisitsyn1besser82: yeah with keyboard12:05
lisitsyn1but with slider12:05
lisitsyn1it worked12:05
lisitsyn1besser82: actually they just patched kernel to support this samsung shit12:06
lisitsyn1after 2 years of being reported12:06
lisitsyn1but actually it works in my ubuntu ;)12:07
lisitsyn1so that's quite interesting thing12:07
sonne|workwow cool - did you see that mark reid wants to co-mentor the SaaS project?12:07
@besser82lisitsyn1, might be canonical has some additional patches12:07
@besser82lisitsyn1, that is always the same sh*t with samsung....12:08
lisitsyn1sonne|work: yeah12:08
lisitsyn1that's cool12:08
sonne|workyes very!12:08
@besser82lisitsyn1, on my lenovo x220 && Clevo P170HMx everything works oob12:09
lisitsyn1I have no idea about mark but he seems to be some *the* guy12:09
@wikingsonne|work: win port before android imo12:09
lisitsyn1sonne|work: I have no idea about his previous work but he seems to be pretty cool12:09
lisitsyn1besser82: yeah I know12:10
lisitsyn1besser82: but samsung is always like that :D12:10
@besser82lisitsyn1: I know, same shit was with Q45 in 2007  :-P12:10
lisitsyn1besser82: battery is killed already12:10
@besser82lisitsyn1: That problem I never had on mine...12:10
lisitsyn1besser82: they are doing their best to make the worst hardware12:10
@besser82lisitsyn1: hehe  :D12:11
@besser82lisitsyn1: buy lenovo next time  :D12:11
@besser82lisitsyn1: they do best to build good hw in cheap case  :-P12:11
sonne|workwiking: do we have one who does it?12:11
@wikingsonne|work: none.... only a guy who really would like to have a win port :D12:12
@wikingsonne|work: i mean it's totally nonesense that somebody already came to do an android port12:12
lisitsyn1besser82: yeah they are much better12:12
@wikingbut nobody would like to do VS port :)12:12
lisitsyn1we have ms evangelist here12:13
@wikinglisitsyn1: yeah yeah :D12:13
lisitsyn1sonne|work: ^12:13
lisitsyn1shoot him down!12:13
-!- HeikoS1 [] has quit [Read error: Connection reset by peer]12:17
@besser82lisitsyn1, Who likes M$ on the chan?!?12:19
@besser82lisitsyn1, Btw. if you want to stick to Fedora:  file a bug against kernel with the wish of making bl-dim working on your hw  :D12:20
@besser82lisitsyn1, josh boyer or theodore tso will take care of it then12:20
lisitsyn1besser82: I think this will be fixed with current version of kernel12:21
lisitsyn1they added some routines to reset some queue12:22
@besser82lisitsyn1: which kernel ver?12:22
@besser82lisitsyn1: FC20 has kernel-0:3.13.6-200.fc20.x86_64 recently...12:22
@besser82lisitsyn1: which should be latest stable12:23
lisitsyn1besser82: cutting edge ;)12:25
@besser82lisitsyn1: allright, that will be in 3.14 then  ;)12:25
@besser82lisitsyn1: which is currently in rc6, so should populate to FC20 pretty soon ;)12:26
@besser82lisitsyn1: if you want to try backport, I can supply you with 3.14.0-rc6 kernel pkgs for FC2012:28
-!- HeikoS1 [] has joined #shogun12:28
lisitsyn1besser82: I am not in hurry ;)12:28
@besser82lisitsyn1: I build that for me, too  ;)12:29
@besser82lisitsyn1: Just need to upload the resulting rpms  :-P12:29
@besser82lisitsyn1: %changelog says:  Fix stale EC events on Samsung systems (rhbz 1003602)12:32
@besser82lisitsyn1:  There you go:
lisitsyn1besser82: I was thinking about some things like security12:42
@besser82lisitsyn1: security in which means / manner?12:43
lisitsyn1it would be pretty funny if we distribute some 'wrong' plugins12:43
lisitsyn1besser82: can you please ask your guys who know anything about that12:44
lisitsyn1how can we sign them up12:44
lisitsyn1or so12:44
@besser82lisitsyn1: you mean like pki?12:44
lisitsyn1yeah may be12:44
@besser82lisitsyn1: or adding some gpg.asc?12:44
lisitsyn1I have no idea12:44
lisitsyn1I haven't ever cared about security12:44
lisitsyn1but this would be important12:45
@besser82lisitsyn1, will ask those rpm-guys how they do that  ;)12:45
@besser82lisitsyn1, iirc  they generate sha512 from file and sign that with gpg ...12:46
-!- PirosB3 [~pirosb3@] has quit [Quit: PirosB3]12:46
lisitsyn1besser82: we can actually download sources12:46
lisitsyn1and compile locally12:46
lisitsyn1this makes sense performance wise12:46
@besser82yes, but src needs to have some signing too  ;)12:46
lisitsyn1yeah sure12:46
lisitsyn1this doesn't solve security issues12:47
lisitsyn1I mean just distributing sources12:47
@besser82so we can do like create tar.$compressed, gen sha512 of uncompresse && compressed tarball, sign those sums with gpg-key ...12:49
@besser82What do oyu think?12:49
@besser82lisitsyn1, ^^^12:50
lisitsyn1besser82: yeah sure12:54
lisitsyn1should be safe enough12:54
@besser82lisitsyn1: i think so too12:57
@besser82lisitsyn1: at least that should be safe enough to make sure the src is fine12:58
@besser82lisitsyn1: but someone might still do some malicious dns-spoof and hijack  :-P12:58
lisitsyn1besser82: that's headache for someone else13:02
lisitsyn1we don't pretend to have super safe package system <tm>13:03
-!- HeikoS1 [] has quit [Ping timeout: 252 seconds]13:03
sonne|workwiking: android is easier than win*13:04
shogun-buildbot_build #214 of debian wheezy - memcheck is complete: Failure [failed memory check]  Build details are at  blamelist: Viktor Gal <>, dhruv13j <>13:05
@wikingsonne|work: true... :)13:12
@wikingah code reviewing... such a fuuuuuuun.... *not*!13:15
-!- PirosB3 [~pirosb3@] has joined #shogun13:16
PirosB3Hi all13:18
@wikingI see you have a very good point! :) I'm satisfied.13:20
@wikingwoah i'm allowed to b right :D13:20
-!- lambday [67157f54@gateway/web/freenode/ip.] has quit [Ping timeout: 245 seconds]13:25
lisitsyn1wiking: haha13:27
-!- kprah [67157d52@gateway/web/freenode/ip.] has joined #shogun13:35
kprahwiking: hey! are you free?13:35
kprahFound some redundancies, wanted to discuss.13:35
sonne|workkprah: always ask the question please don't ask to ask :)13:51
kprahok.... I just wanted to ask if we can use templates with SG_CALLOC13:52
sonne|workkprah: what do you mean? make an example13:56
kprahthere is a redundancy where we are continually checking if a pointer which has been SG_CALLOC'ed with a given datatype say float64_t, to be null or not.14:03
kprahi.e. matrix = (uint32_t*) LIBBMRM_CALLOC(aDim , uint32_t);14:04
sonne|workkprah: you mean the cast here?14:05
kprahi don't intend to use cast14:05
kprahcast will be too heavy.14:06
kprahI want to use template14:06
sonne|worksorry I dont' know what you mean...14:06
kprahin this way:14:06
kprahtemplate<Type T>14:06
kprahmatrix = (T*) LIBBMRM_CALLOC(aDim, sizeof(T));14:07
kprahcan I do this?14:07
kprahI mean the above line is obviously a part of a function14:07
sonne|workkprah: no the problem is LBMRM is not written by us and it would be bad to deviate too much in case we want to merge in newer versions14:10
kprahWell , I am working on LIBBMRM. Guess I have to read it thoroughly then. Thanks.14:11
sonne|workkprah: ahh sorry then you cannot know. vojtech wrote most of it and libbmrm is also an external library14:21
lisitsyn1kprah: what do you want to get with templating it?14:21
kprahactually there were many error checking calls of similar pattern in libbmrm.cpp14:22
kprahi wanted to get around it by writing a template function14:22
lisitsyn1yeah but that's just float or double right?14:23
kprahno bool14:24
kprahuint32 , float64_t14:24
kprahinfact a couple of types14:24
lisitsyn1what would be templated then?14:25
lisitsyn1the whole function?14:25
@wikingsonne|work: no actually it's not written by vojtech14:29
@wikingsonne|work: and it's not being maintained anywhere else14:29
@wikingjust in our codebase14:29
lisitsyn1yeah it is not developed as open source :D14:30
sonne|workwiking: errm why do we have a libbmrm anyways? libocas has a bmrm mode...14:31
kprahby vojtech are you referring to Michal?14:31
-!- striker [ca8362aa@gateway/web/freenode/ip.] has joined #shogun14:31
strikerhi, i want to take part in gsoc 2014, how can i proceed???14:33
lisitsyn1striker: please see
@wikingsonne|work: don't ask me... vojtech mentored this project...14:35
@wikingkprah: templating out calloc functions is not the biggest problem here...14:36
lisitsyn1hah yeah interesting codestyle is the issue14:37
-!- striker [ca8362aa@gateway/web/freenode/ip.] has quit [Ping timeout: 245 seconds]14:37
@wikingindeed i dont see why do you want to introduce template for malloc14:37
@wikingkprah: that's what you get when u port code from matlab...14:37
kprahwell I am editing libbmrm.cpp :)14:37
kprahin clean_icp function14:38
-!- striker [ca8362aa@gateway/web/freenode/ip.] has joined #shogun14:38
kprahwe are repeatedly checking for the nullity of many pointers.14:39
strikeri want to do project in machine learning algorithms14:39
lisitsyn1striker: you are for sure welcome to apply to any project14:40
-!- striker [ca8362aa@gateway/web/freenode/ip.] has quit [Quit: Page closed]14:44
-!- [paladin] [73f91219@gateway/web/freenode/ip.] has joined #shogun14:50
kprahwhich branch of shogun should I push to?14:56
@besser82kprah: make a pull-request to develop  ;)15:14
@besser82kprah:  :D15:15
kprahbtw in what modules of shogun, the quasi-newton methods are implemented?15:16
kprahwiking : are you familiar?15:17
-!- PirosB3 [~pirosb3@] has quit [Quit: PirosB3]15:21
kprahwell If I know I can unify all of them , i.e. bundle methods and bmrm15:24
kprahsorry bfgs15:24
-!- pl8787 [9fe22b60@gateway/web/freenode/ip.] has joined #shogun15:41
pl8787Hi everyone.15:44
-!- lisitsyn1 [] has quit [Quit: Leaving.]15:47
pfm_hello pl878715:47
pl8787It's my first time to be here15:48
pl8787My pull requst is travis passed. Does anyone review?15:52
-!- HeikoS1 [] has joined #shogun15:57
@wikingHeikoS1: ping16:03
pl8787Hi HeikoS1 my request passed travis16:04
@wikingpl8787: trust me we see it w/o a warning.... we'll try to answer/act ASAP but please understand that all of us does this as a hobby and not full time work16:06
HeikoS1wiking: pong16:06
HeikoS1pl8787: nice, I will check things soon16:06
HeikoS1pl8787: just too many emails currently, I havent checked this morning so now its double ;)16:06
HeikoS1wiking:  yeah indeed ^16:06
@wikingHeikoS1: about entropy16:07
@wikingHeikoS1: why do u use logspace probabilities? :)16:07
HeikoS1wiking: probabilities should always be stored in lgo-domain16:07
pl8787I just test this webchat :)16:07
HeikoS1wiking: to avoid underflows16:07
@wikingHeikoS1: oh ok16:07
HeikoS1wiking: when multiplying then16:07
@wikingHeikoS1: good to know16:07
@wikingHeikoS1: but makes sense :)16:08
HeikoS1wiking: and there is also log-sum-exp if one wants to average in non-log domain16:08
HeikoS1wiking: so if you have numbers log(a) log(b) log(c) but you want to compute log(exp(log(a))+exp(log(b))+exp(log(c))) ,thats in shogun16:10
-!- kprah [67157d52@gateway/web/freenode/ip.] has quit [Ping timeout: 245 seconds]16:14
-!- kprah [67157d4c@gateway/web/freenode/ip.] has joined #shogun16:26
-!- kprah [67157d4c@gateway/web/freenode/ip.] has quit [Quit: Page closed]16:32
-!- shogun-notifier- [] has joined #shogun16:33
shogun-notifier-shogun: Wu Lin :develop * 44250ac / .gitignore:
shogun-notifier-shogun: add cscope indexing files at .gitignore16:33
shogun-notifier-shogun: Wu Lin :develop * 00241d5 / src/shogun/machine/gp/ (2 files):
shogun-notifier-shogun: add/modify some comments at the CGaussianLikelihood class16:33
shogun-notifier-shogun: Heiko Strathmann :develop * d54b1ba / / (3 files):
shogun-notifier-shogun: Merge pull request #1977 from yorkerlin/develop16:33
shogun-notifier-shogun: add some comments at the CGaussianLikelihood class16:33
-!- pl8787 [9fe22b60@gateway/web/freenode/ip.] has quit [Quit: Page closed]16:36
shogun-notifier-shogun: pl8787 :develop * ff75832 / / (2 files):
shogun-notifier-shogun:  #1953 memory leak and uninitialised value fixed.16:43
shogun-notifier-shogun: pl8787 :develop * c92a781 / / (8 files):
shogun-notifier-shogun: Merge remote-tracking branch 'shogun/develop' into develop16:43
shogun-notifier-shogun: pl8787 :develop * 2a3cff1 / / (43 files):
shogun-notifier-shogun: Merge remote-tracking branch 'shogun/develop' into develop16:44
shogun-notifier-shogun: pl8787 :develop * 04b626d / tests/unit/classifier/svm/
shogun-notifier-shogun: Add LibLinear unit test with real data and compare with liblinear. Pass valgrind check.16:44
shogun-notifier-shogun: pl8787 :develop * 8e99789 / src/shogun/classifier/svm/LibLinear.cpp:
shogun-notifier-shogun: Use new T[] instead of SG_MALLOC.16:44
shogun-notifier-shogun: pl8787 :develop * 574f2c2 / tests/unit/classifier/svm/
shogun-notifier-shogun: fixed unit test error on some boundary condition.16:44
shogun-notifier-shogun: pl8787 :develop * beae221 / tests/unit/statistics/
shogun-notifier-shogun: #1931 HSIC unit test added.16:44
shogun-notifier-shogun: pl8787 :develop * fee9354 / tests/unit/statistics/
shogun-notifier-shogun: Remove one EXPECT_NEAR on p value16:44
shogun-notifier-shogun: pl8787 :develop * c759b8e / tests/unit/classifier/svm/
shogun-notifier-shogun: Simplify the header16:44
-!- witness___ [uid10044@gateway/web/] has quit [Quit: Connection closed for inactivity]16:44
jiaolongHi, when I add or remove a new class under /src/shogun/io/, I couldn't pass the compile.  The error comes from ../src/shogun/base/class_list.cpp. How to solve this problem? Thank you!16:54
-!- sunil1337 [0e8bd057@gateway/web/freenode/ip.] has joined #shogun16:58
-!- sunil1337 [0e8bd057@gateway/web/freenode/ip.] has quit [Ping timeout: 245 seconds]17:13
-!- HeikoS1 [] has quit [Ping timeout: 240 seconds]17:29
-!- lisitsyn [~lisitsyn@] has joined #shogun17:31
-!- vivekrai [uid10305@gateway/web/] has quit [Quit: Connection closed for inactivity]17:42
-!- abinash_panda [01261510@gateway/web/freenode/ip.] has joined #shogun17:49
-!- abinash_panda [01261510@gateway/web/freenode/ip.] has quit [Ping timeout: 245 seconds]17:55
-!- zxtx [] has quit [Ping timeout: 241 seconds]18:58
-!- [paladin] [73f91219@gateway/web/freenode/ip.] has quit [Ping timeout: 245 seconds]19:11
@wikinglisitsyn: ping19:21
-!- bumblebee [~quassel@] has joined #shogun19:23
-!- kislay [~kislay@] has joined #shogun19:24
-!- cameo54321 [2729794f@gateway/web/freenode/ip.] has joined #shogun19:25
-!- sonne|osx [] has joined #shogun19:42
-!- shogun-notifier- [] has quit [Quit: transmission timeout]19:44
-!- gif [~gif@] has joined #shogun19:50
-!- PirosB3 [~pirosb3@] has joined #shogun20:16
-!- kprah [67157d4c@gateway/web/freenode/ip.] has joined #shogun20:17
-!- brothersam [~brothersa@] has joined #shogun20:17
kprahRegarding issue 1969: there is already an unified interface available in CDualLibQPBMSOSVM. Am I wrong?20:18
-!- dhruv13 [~dhruv13@] has joined #shogun20:18
dhruv13wiking: hey! I got a mail for buildbot failure:
dhruv13wiking: any clue why?20:19
dhruv13wiking: it's from the patch i submitted for the environment variable thing...20:21
shogun-buildbot_build #215 of debian wheezy - memcheck is complete: Failure [failed memory check]  Build details are at  blamelist: Heiko Strathmann <>, Wu Lin <>20:22
dhruv13okay, it seems to be recursing...20:23
dhruv13same problem as above...20:23
-!- brothersam [~brothersa@] has quit [Remote host closed the connection]20:24
kprahhey, dhruv1320:26
-!- abinash_panda [01261510@gateway/web/freenode/ip.] has joined #shogun20:28
@wikingkprah: i would suggest to close this pr, as it's really a bit messy20:29
dhruv13kprah: hello!20:29
kprahwell I cleaned it up.20:29
@wikingkprah: yeah but currently you have 9 commits that on the end changed 8 lines20:30
@wikingkprah: that's really too much20:30
kprahok. I'll close it.20:30
@wikingkprah: and honestly i expected something more in depth than this20:30
kprahRegarding issue 1969: there is already an unified interface available in CDualLibQPBMSOSVM. Am I wrong?20:30
@wikingkprah: that's another story20:31
kprahWell I'm not used to git and all.20:31
-!- fullface [~fullface@] has joined #shogun20:32
dhruv13wiking: okay, sorry for the confusion, i thought i added an unnecessary SG_UNREF(), which caused the build problem...20:32
@wikingkprah: yeah i can tell.... but this is why we have entrance tasks so that you get familiarised with the whole development story20:32
-!- jiaolong [9e6d09e4@gateway/web/freenode/ip.] has left #shogun []20:33
kprahSorry. Please wait for the next PR. I'll make up for the mistakes i've made.20:34
-!- bumblebee [~quassel@] has quit [Read error: Operation timed out]20:36
dhruv13besser82: hello!20:40
@besser82dhruv13: Hey! 'sup?20:41
dhruv13besser82: nothing much ;-) looking for datasets...20:41
dhruv13besser82: found something interesting...20:41
@besser82dhruv13: tried  ?20:41
@besser82dhruv13: which interessenting?20:42
-!- thoralf [] has joined #shogun20:42
dhruv13besser82: evaluation of ann -
dhruv13besser82: ann = approx nearest neighbours20:42
-!- fullface [~fullface@] has quit [Remote host closed the connection]20:42
@besser82dhruv13: just seen  it  ;)20:42
@besser82dhruv13: looks awesome  :D20:44
@besser82thoralf: Hey, man!20:44
@besser82thoralf: How is it going?20:44
dhruv13besser82: yeah! really specific, just what i need20:44
@besser82dhruv13: but files are a bit big  ;)20:44
thoralfbesser82: Busy-busy-all-the-time.20:44
thoralfbesser82: You're fine?20:45
@besser82thoralf: With me all is fine  :D20:45
-!- webprotech [~webprotec@] has joined #shogun20:45
@besser82thoralf: Just busy too atm20:45
@besser82thoralf: playing with webmining and related stuff20:45
@besser82thoralf: like with pthon-pattern  ;)20:46
thoralfbesser82: What exactly are you trying to mine?20:46
@besser82thoralf: text from a database20:46
dhruv13besser82: i'll use a subset, i think... need to test my nanoflann programs...20:46
@besser82thoralf: mostly customer responses or problem reports20:46
-!- webprotech [~webprotec@] has quit [Remote host closed the connection]20:47
@besser82thoralf: to get them categorized by use of speech or buzz-words20:47
thoralfbesser82: Scary. ;)20:47
@besser82thoralf: indeed  :P20:47
dhruv13besser82: have you looked into opinosis?
thoralf"Ey geht nicht.  Is kaputt."20:47
thoralfTry to mine that!20:47
-!- sonne|osx [] has quit [Quit: sonne|osx]20:47
dhruv13besser82: non conventional ;-)20:47
dhruv13thoralf: lol20:47
-!- sejDino [] has joined #shogun20:48
@besser82dhruv13: you made my day!   :DD20:48
dhruv13besser82: i'm working on something similar for my final year project ;-)20:49
@besser82thoralf: talking 'bout stuff like this:
dhruv13besser82: :-)20:49
@besser82thoralf, or that
dhruv13don't understand german :-( translate only takes you so far...20:52
abinash_pandathoralf: Hi !20:56
thoralfabinash_panda: Hey!20:56
abinash_pandathoralf: I have been working on the issue
abinash_pandathoralf: I have started creating methods for loading/saving Factor Graphs in UAI format.20:57
abinash_pandaRight now, I have implemented a method for saving the model in UAI format.20:57
thoralfabinash_panda: Didn't look into it right now -- just reading my mail the first time since this morning. ;)20:58
thoralfabinash_panda: Are you speeding up with shogun?20:58
thoralfDo you like it?20:58
abinash_pandaYeah. I liked it a lot :)20:58
abinash_pandathoralf: So, should I send a PR right now, or should I wait till I complete both the methods?20:59
thoralfabinash_panda: Send a PR and link it to the issue, so we can have a look.21:00
abinash_pandaOk. Sending it right now :)21:00
thoralfabinash_panda: It's always a good idea to PR on half-way, just to see if there are misunderstandings.21:00
abinash_pandathoralf: Ok.21:00
-!- kprah [67157d4c@gateway/web/freenode/ip.] has quit [Quit: Page closed]21:02
abinash_pandathoralf: Sent a PR. If you could have a look at it.21:02
thoralf1984?  George Orwell :)21:04
-!- cameo54321 [2729794f@gateway/web/freenode/ip.] has quit [Ping timeout: 245 seconds]21:05
dhruv13thoralf: lol21:06
-!- gif [~gif@] has quit [Remote host closed the connection]21:07
-!- PirosB3 [~pirosb3@] has left #shogun []21:14
-!- PirosB3 [~pirosb3@] has joined #shogun21:14
-!- gif [~gif@] has joined #shogun21:17
@wikingabinash_panda: i've just closed that pr21:22
@wikingabinash_panda: today i've sent an email about this that we really dont want to have that function part of FactorGraph21:23
@wikingabinash_panda: so please refactor your PR that's implementing the same functionality but in src/shogun/io...21:24
abinash_pandaOk. Let me have a look at the mail.21:24
abinash_pandawiking: I have a few doubts.21:25
abinash_pandawiking: So, I have to implement the same functionality but in src/shogun/io..., this means that I have to create a new class called UAIFile. Am I right ?21:27
naywhayaresonney2k: wiking: lisitsyn: I've contacted the scikit-learn guys about the benchmarking project:
naywhayareno response yet, but I will let you know what I find out21:30
@wikingabinash_panda: yeah that'd be the idea21:31
@wikingnaywhayare: great!21:31
-!- sejDino [] has left #shogun []21:32
@wikingnaywhayare: although i cannot open that link 'We're sorry but we weren't able to process this request.'21:32
naywhayareodd... does this one work?
naywhayarefor now, at least, my message is still the most recent, so it is at the top21:32
abinash_pandawiking: Ok. I got it.21:35
@wikingnaywhayare: yeah got this way :)21:37
naywhayareok, good21:37
@wikingnaywhayare: let's hope they get on board as well21:39
naywhayareI am interested to hear their ideas.  I think we need to be careful that each change we make to the system is deliberate and well-planned, so that the system goes one (or a few) directions well, instead of a million directions simultaneously but poorly21:40
-!- bumblebee [~quassel@] has joined #shogun21:53
-!- bumblebee is now known as Guest9803221:53
-!- Guest98032 [~quassel@] has quit [Remote host closed the connection]21:53
-!- gif [~gif@] has quit [Remote host closed the connection]21:57
-!- iglesiasg [] has joined #shogun21:59
-!- mode/#shogun [+o iglesiasg] by ChanServ21:59
-!- wiking_ [] has joined #shogun22:22
-!- _flatline_ [~flatline@2607:f470:22:8:beae:c5ff:fe5a:903f] has joined #shogun22:26
-!- gif [~gif@] has joined #shogun22:27
-!- besser82 [quassel@fedora/besser82] has quit [Ping timeout: 245 seconds]22:28
-!- wiking [~wiking@huwico/staff/wiking] has quit [Ping timeout: 245 seconds]22:28
-!- Netsplit *.net <-> *.split quits: thoralf22:33
-!- besser82 [quassel@fedora/besser82] has joined #shogun22:34
-!- mode/#shogun [+o besser82] by ChanServ22:34
-!- Netsplit over, joins: thoralf22:39
-!- Netsplit *.net <-> *.split quits: naywhayare, gif, @iglesiasg, PirosB3, lisitsyn22:39
-!- Netsplit over, joins: gif, @iglesiasg, PirosB3, lisitsyn, naywhayare22:41
shogun-buildbot_build #216 of debian wheezy - memcheck is complete: Failure [failed memory check]  Build details are at  blamelist: Heiko Strathmann <>, pl8787 <>22:48
-!- zxtx [] has joined #shogun23:06
-!- dhruv13 [~dhruv13@] has quit [Remote host closed the connection]23:25
-!- iglesiasg [] has quit [Quit: Leaving]23:33
-!- shogun-notifier- [] has joined #shogun23:36
shogun-notifier-shogun: frank0523 :develop * 6ba98c6 / doc/md/
shogun-notifier-shogun: Add namespace in sample code and add some hint on downloading data23:36
shogun-notifier-shogun: frank0523 :develop * eff9b57 / doc/md/
shogun-notifier-shogun: Fix code format23:36
shogun-notifier-shogun: frank0523 :develop * 978037a / doc/md/
shogun-notifier-shogun: Fix typo23:36
shogun-notifier-shogun: Fernando Iglesias :develop * 41a96fd / doc/md/
shogun-notifier-shogun: Merge pull request #1980 from frank0523/develop23:36
shogun-notifier-shogun: #1978 Update README_developer.md23:36
thoralfwiking_: Finding unused imports using clang:
thoralfwiking_: Did you try it?23:39
--- Log closed Wed Mar 12 00:00:34 2014