Open in new window / Try shogun cloud
--- Log opened Tue Apr 30 00:00:29 2013
van51sonney2k: I may be doing something wrong, but valgrind still reports memory leaks with your commit alone00:01
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving]00:06
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun00:06
-!- mode/#shogun [+o iglesiasg] by ChanServ00:06
-!- cesarf [b2c73bf3@gateway/web/freenode/ip.178.199.59.243] has joined #shogun00:47
-!- [1]wencan [~wencan@c-71-61-182-121.hsd1.pa.comcast.net] has joined #shogun00:54
-!- wencan [~wencan@c-71-61-182-121.hsd1.pa.comcast.net] has quit [Ping timeout: 256 seconds]00:57
-!- [1]wencan is now known as wencan00:57
-!- cesarf [b2c73bf3@gateway/web/freenode/ip.178.199.59.243] has quit [Quit: Page closed]01:01
-!- van51 [~van51@athedsl-401908.home.otenet.gr] has quit [Quit: Leaving.]01:22
-!- van511 [~van51@athedsl-401908.home.otenet.gr] has joined #shogun01:22
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has quit [Ping timeout: 276 seconds]01:42
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout]01:59
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun02:04
-!- wencan [~wencan@c-71-61-182-121.hsd1.pa.comcast.net] has quit [Quit: HydraIRC -> http://www.hydrairc.com <- \o/]02:52
-!- alibezz [~alibezz@177.182.164.8] has quit [Quit: Leaving.]03:01
-!- alibezz1 [~alibezz@177.182.164.8] has joined #shogun03:01
-!- alibezz1 [~alibezz@177.182.164.8] has quit [Quit: Leaving.]03:08
-!- alibezz [~alibezz@177.182.164.8] has joined #shogun03:09
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving]03:29
-!- alibezz1 [~alibezz@177.182.164.8] has joined #shogun03:38
-!- alibezz [~alibezz@177.182.164.8] has quit [Ping timeout: 245 seconds]03:39
-!- alibezz1 [~alibezz@177.182.164.8] has left #shogun []04:36
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has joined #shogun06:12
-!- foulwall_ [~foulwall@li379-21.members.linode.com] has joined #shogun06:34
@sonney2kwiking, yeah I think the system might just have slightly different numerical precision...07:28
@sonney2kshogun-buildbot, force build 'cyg1 - libshogun'07:29
shogun-buildbotbuild forced [ETA 11m44s]07:29
shogun-buildbotI'll give a shout when the build finishes07:29
shogun-buildbotbuild #716 of cyg1 - libshogun is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/71607:30
@sonney2kshogun-buildbot, force build 'cyg1 - libshogun'07:36
shogun-buildbotbuild forced [ETA 11m44s]07:36
shogun-buildbotI'll give a shout when the build finishes07:36
shogun-buildbotbuild #717 of cyg1 - libshogun is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/71707:37
-!- mikhailBelous [~quassel@109.205.249.161] has quit [Remote host closed the connection]07:40
-!- foulwall_ [~foulwall@li379-21.members.linode.com] has quit [Ping timeout: 245 seconds]07:55
sonne|workwiking: the deviation is actually 1e-17 or 1e-1808:53
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun09:10
-!- mode/#shogun [+o iglesiasg] by ChanServ09:10
-!- admtuku [b4951f3b@gateway/web/freenode/ip.180.149.31.59] has joined #shogun09:14
-!- lisitsyn1 [~blackburn@37.61.181.55] has quit [Quit: Leaving.]09:25
@wikingsonne|work: lol09:34
-!- van511 [~van51@athedsl-401908.home.otenet.gr] has left #shogun ["PING 1367308428"]09:54
-!- gsomix_ [~Miranda@37.61.181.55] has quit [Ping timeout: 245 seconds]10:03
sonne|workwiking: this test is *very* strict it really checks for 100% reproducibility10:10
sonne|workif e.g. atlas would give a different result than just standard basic floatingpoint ops then we would get this10:11
-!- van51 [~van51@195.251.235.94] has joined #shogun10:29
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has quit [Quit: Leaving]10:30
-!- lisitsyn [~lisitsin@mxs.kg.ru] has joined #shogun10:35
-!- mode/#shogun [+o lisitsyn] by ChanServ10:35
-!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has joined #shogun10:36
-!- admtuku [b4951f3b@gateway/web/freenode/ip.180.149.31.59] has quit [Quit: Page closed]10:37
@lisitsynsonne|work: I'll arrive to schoenefeld :D10:38
sonne|workhaha10:38
sonne|workthat is not an airport10:38
@lisitsynsonne|work: nice :D10:39
sonne|workit is some temporary construction site10:39
@lisitsynhmm10:40
@lisitsynlet me check10:40
sonne|workyou will have to walk off the plane10:40
sonne|workfor 500m's10:40
@lisitsynsonne|work: that's usual here in samara :D10:40
sonne|workthen pass through some boxes and then walk again 200m's to get to a train station10:40
sonne|workbut hey10:40
@lisitsynhaha nice nice10:41
@lisitsynsonne|work: I checked - it is schoenefeld10:42
sonne|worklisitsyn: any ideas what we do about the failed test on travis? We now have the full detail...10:44
sonne|worklisitsyn: output is here https://travis-ci.org/shogun-toolbox/shogun/jobs/673925410:44
sonne|work(go to the very end)10:45
sonne|workyou will notice that the originally serialized object deviates by 1e-17 or even 1e-1810:45
@lisitsynsonne|work: interesting10:46
sonne|workI would totally understand that if other tests were failing too10:46
sonne|workbut it seems to be just this one !?!10:47
sonne|workand only w/ one setting10:47
@lisitsynwe get lucky10:47
@lisitsynthat's the only thing I can say haha10:47
sonne|workwell but what do we do about it10:51
sonne|workwe could10:51
sonne|worka) just drop the first setting10:52
sonne|workor change the test to return some numpy values10:52
sonne|workor ...?10:52
sonne|worklisitsyn, wiking any ideas?10:53
@lisitsynsonne|work: it sounds unreasonable to test values with such precision10:54
sonne|worklisitsyn: yes and no10:55
sonne|workwhat this test does is serialize a file10:55
sonne|workerr object10:55
sonne|workand then it does the same thing again10:55
sonne|workand compares10:55
@lisitsynsonne|work: it may be just gmnpsvm tihng10:55
sonne|workso stuff should match 100%10:55
sonne|workI don't see how10:56
sonne|workgmnpsvm tests don't have this issue10:56
@lisitsyndon't we train it in this example?10:56
@lisitsynwe should not train it there10:56
thoralfsonne|work: Sorry if I'm bugging - but did you think about my email?  i.e. how to use stream_features->get_vector() properly?10:56
* sonne|work overload10:57
sonne|worklisitsyn: yes we train10:57
sonne|workit is meant to be a complex example10:57
sonne|worklisitsyn: we have a) multiclass machine - so it has multiple SVMs10:57
sonne|workeach having its kernel set & features10:57
sonne|workand labels10:57
sonne|workand preprocessors attached10:58
sonne|workand if you currently train GMNPsvm on that example10:58
sonne|workthen serialize to .json etc etc10:58
sonne|workand then load from json etc10:58
sonne|workand train again10:58
sonne|workyou will get exactly the same result (except for json - slightly worse accuracy)10:59
@lisitsynsonne|work: well I mean we could just initialize it rather hard-coded11:00
@lisitsynand then we will not have that issues11:00
@lisitsynissue*11:00
@lisitsynit makes sense to test only serialization here11:01
@lisitsynbut not gmnpsvm11:01
sonne|worklisitsyn: makes sense. Then I would just say we remove the first test and simplify it later11:03
sonne|worklisitsyn: by may 6th we have to request the number of slots11:10
sonne|workso we should have students <-> mentors loosely assigned by then11:10
@lisitsynsonne|work: alright11:18
-!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has quit [Ping timeout: 245 seconds]11:22
@lisitsynsonne|work: do you think I'll get to tiergarten from schoenefeld somewhat at 22:00?11:27
@lisitsynby public transport? or rather taxi?11:27
sonne|workwhy go to tiergarten?11:29
sonne|workthat is the other airport11:29
@lisitsynsonne|work: hotel is there11:29
sonne|workmore or less11:29
sonne|workI still don't get why you dont' take the ostel11:29
sonne|workbut yes sure11:29
@lisitsynyes I know tegel is closer but my tickets are to schoenefeld11:29
@lisitsynostel is not cheapier than what I chose11:29
@lisitsynmay be just a little11:29
@lisitsynso why not to live in something more comfortable? :)11:30
@lisitsynsonne|work: not really tiergarten - cross of schillstrasse and kurfuerstenstrasse11:32
sonne|workyour holidays :)11:32
@lisitsynsonne|work: I just don't know how good public transport is - that's why I am asking11:33
@lisitsynin samara it could be a trouble to get to the city that late :D11:33
@lisitsynbut we are wild you know11:33
@lisitsynI was surprised you have night transport11:33
sonne|workwell it is berlin11:34
@lisitsynwe have nothing like buses on roads after midnight11:34
@iglesiasgno roads haha11:34
sonne|work20% of the people wake up around midnight here11:34
@iglesiasglisitsyn: according to the pictures you have shown, you roads also have some troubles during daylight11:34
@iglesiasg:P11:34
@lisitsyniglesiasg: oh that keeps you awake11:37
@lisitsynyou can't be no cautious here11:37
@lisitsynsonne|work: iglesiasg: btw happy bday to gsomix11:38
@iglesiasgcool11:39
van51sonne|work: hey, did you see the messages I left you? can it be that it still leaks?11:39
van51also if not, should i start another entry issue?11:39
-!- abinash [75ef5e6e@gateway/web/freenode/ip.117.239.94.110] has joined #shogun11:40
abinashiglesiasg: please could you review my GSoC application and comment me if any changes are to be done?11:41
sonne|workvan51: looks like my messages got dropped11:42
@iglesiasgabinash: I can glance over it but I am just a simply mortal student :P11:42
sonne|workvan51: yes - it might still leak memory11:42
sonne|workI was only ensuring that it gives the correct result / doesn't crash11:42
@iglesiasgabinash: I mean for you to know that I am not a mentor or so11:42
abinashiglesiasg: ok11:42
@iglesiasgabinash: link me to it please11:42
abinashiglesiasg: just a moment I am sending you the link11:44
-!- gsomix [~Miranda@188.168.2.34] has joined #shogun11:44
@iglesiasggsomix: feliz cumplea?os!11:45
gsomixiglesiasg: thanks. :)11:45
van51sonne|work: ok.. those wrong results/crashes did you see them from travis or from "make tests"?11:48
abinashiglesiasg: here is the link http://www.google-melange.com/gsoc/proposal/review/google/gsoc2013/abinashpanda/111:48
van51so I can try to spot them on my own from now on11:48
@iglesiasgabinash: I think it looks good ;)11:54
abinashiglesiasg: ok, are there any changes to be made particularly in the project details and in the timeline11:54
gsomixiglesiasg: friends gave me a bottle of Spanish non-alcoholic wine today. not vodka, but ok.11:56
@iglesiasgabinash: from top level it looks fine I think. The thing is that I am not completely aware of the contents of this project, maybe Vojtech can give you better feedback11:57
@iglesiasggsomix: is it called mosto?11:57
sonne|workvan51: I just ran the serialize complex example from python11:57
@iglesiasgabinash: I think there is a type in the timeline btw11:58
@iglesiasgabinash: july 15 - august 29 should be july 29 I think11:59
@iglesiasgtype -> typo*11:59
abinashiglesiasg: yeah just noticed it, thanks11:59
@iglesiasgabinash: in any case, from my experience I think that it is normally more important to make contributions rather than writing a fancy application12:00
abinashiglesiasg: ok, I have contributed to one issue previously and now working on another one12:01
@iglesiasgabinash: at first sight, I think that the part of interfacing libqp from modular interfaces should be easy to do12:02
@iglesiasgmaybe it could be an idea to get that part done and include it in the application12:03
@iglesiasgabinash: only my opinion :)12:03
gsomixiglesias: yes, I think so. but I don't know technology of production exactly. http://bit.ly/189Go0r12:04
abinashiglesiasg: ok, then I should write atleast the python interfaces and the rest of modular interfaces can be left as a part of GSoC project12:04
@iglesiasggsomix: ahhh ok, I think it is not mosto though12:06
-!- abinash [75ef5e6e@gateway/web/freenode/ip.117.239.94.110] has quit [Ping timeout: 245 seconds]12:13
-!- wencan [~wencan@c-71-61-182-121.hsd1.pa.comcast.net] has joined #shogun12:28
-!- wencan [~wencan@c-71-61-182-121.hsd1.pa.comcast.net] has left #shogun []12:31
-!- wencan [~wencan@c-71-61-182-121.hsd1.pa.comcast.net] has joined #shogun12:32
-!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has joined #shogun12:33
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving]12:34
-!- sids_aquarius [~sidi_@li400-124.members.linode.com] has quit [Remote host closed the connection]12:53
-!- sids_aquarius [~sidi_@li400-124.members.linode.com] has joined #shogun12:57
-!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has quit [Ping timeout: 245 seconds]13:03
@wikingsonne|work: ping13:07
sonne|workwiking: poiiing13:08
@wikingsonne|work: ok so question: i couldn't find any specific macros/variables for freebsd in ./configure13:10
sonne|workwiking: so?13:10
@wikingwhich one(s) were u referring to yesterday?13:10
sonne|workjust add them13:10
@wikingsonne|work: heheh ok cool that's even more work...13:10
sonne|workthere is a function called freebsd()13:10
sonne|worknot really13:10
van51sonne|work: hey again :P so with your changes and mine, I don't see neither that test failing nor any memory leaks13:13
van51would you like me to make another PR so you can have another look?13:13
sonne|workvan51: is it still the same one you do?13:13
-!- van51 [~van51@195.251.235.94] has quit [Read error: Connection reset by peer]13:13
sonne|workhmmhh13:13
-!- van51 [~van51@195.251.235.94] has joined #shogun13:15
@wikingsonne|work: what if i add this into test_modular ?13:16
@wikingas it's swig related13:16
sonne|workwiking: check_wordsize maybe13:16
sonne|workit is a workaround for wrong wordsize so I think it fits best there13:17
@wikingsonne|work: yeah although that script runs fine as it uses simple c13:17
@wikingnot c++13:17
sonne|workjust add the if freebsd ; then ...13:17
sonne|workwhat?13:17
sonne|workDEFINES_SWIG="${DEFINES_SWIG} -DSWIGWORDSIZE64"13:18
sonne|workafter this line13:18
sonne|workjust add this13:18
sonne|workif13:18
@wikingsonne|work: yeah as i said yesterday wordsize is 64 if u compile with c and not c++13:18
sonne|worksorry what?13:18
sonne|workjust do the fix there and be done13:19
van51sonne|work: yeap the same one13:19
@wikingsonne|work: shall it be part of DEFINES_SWIG ?13:19
sonne|workyes it is only affecting swig13:22
@wikingok lets see13:22
@wikingwe still need --swig as well13:22
sonne|workvan51: I think the reference count increase there is justified - what is missing is to decrement the counter once the structure is no longer used13:23
@wikingsince the swig binary is swig2.0 and not swig13:23
sonne|workvan51: so you found the issue but I would solve it differently13:23
sonne|workwiking: not really difficult isn't it?13:23
sonne|workjust do it13:23
@wikingyeah and as well as -ldl removal13:24
-!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has joined #shogun13:24
@wikingi dont know how that's going to effect though python_modular on linux13:24
@wikingloool13:25
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has joined #shogun13:25
@wikingsonne|work: DEFINES_SWIG is actually not good13:25
sonne|workwhy not?13:25
@wikingswig2.0 -w473 -w454 -w312 -w325 -fvirtual -builtin -modern -modernargs -c++ -python -I../../shogun -I../.. -I.. -I. -DSWIGWORDSIZE64 -DUINTPTR_MAX=0xffffffffffffffff -DUINT64_MAX=0xffffffffffffffff -DSWIG_TYPE_TABLE=shogun -DSHOGUN -DFREEBSD -DUSE_BOOL -DUSE_CHAR -DUSE_UINT8 -DUSE_UINT16 -DUSE_INT32 -DUSE_INT64 -DUSE_UINT64 -DUSE_FLOAT32 -DUSE_FLOAT64 -DHAVE_SQRTL -DHAVE_LOG2 -DHAVE_PTHREAD -DHAVE_JSON -DHAVE_XML -DHAVE_ATLAS -DHAVE_LAPACK -DHAVE_CATLAS13:25
@wiking# error "SWIG wrapped code invalid in 32 bit architecture, regenarete code using -DSWIGWORDSIZE32"13:25
@wikingmmm13:26
@wikingi think thi one has been cropped13:26
@wikingsonne|work: http://pastebin.com/RZ8s1HQ413:26
@wikingit's better here13:26
@wikingso basically that variable is passed for swig13:26
@wikingwhich we don't really need13:26
@wikingbut we need it for clang++13:26
@wikinghence it should be appeneded to COMP_...13:27
sonne|workok13:27
@wikingCOMPFLAGS_CPP13:27
@wikingtesting13:27
@wikingbtw13:28
@wikingthis is not the best:13:28
@wikingfind . -name "*.c" -o -name "*.cpp" \13:28
@wiking-exec gcc -MM -DSWIG_TYPE_TABLE=shogun -DSHOGUN -DFREEBSD -DUSE_BOOL -DUSE_CHAR -DUSE_UINT8 -DUSE_UINT16 -DUSE_INT32 -DUSE_INT64 -DUSE_UINT64 -DUSE_FLOAT32 -DUSE_FLOAT64 -DHAVE_SQRTL -DHAVE_LOG2 -DHAVE_PTHREAD -DHAVE_JSON -DHAVE_XML -DHAVE_ATLAS -DHAVE_LAPACK -DHAVE_CATLAS -DHAVE_EIGEN3 -DUSE_GLPK -DUSE_SNAPPY -DUSE_GZIP -DUSE_BZIP2 -DUSE_LZMA -DHAVE_LARGEFILE -DUSE_SHORTREAL_KERNELCACHE -DUSE_BIGSTATES -DUSE_HMMCACHE -DUSE_REFERENCE_COUNTING -DUSE_SVML13:28
@wikingit's using gcc.... but it should use clang13:28
@wikingas that the compiler based on ./configure13:28
@wikingsssshsss13:30
@wikingwoah13:30
@wikingwhy that clang++ does not use COMPFLAGS_CPP13:30
@wikingCOMPFLAGS_CPP += $(DEFINES)13:31
@wikingCOMPFLAGS_C += $(DEFINES)13:31
@wikingmmmm13:31
@wikingwhere DEFINES += $(DEFINES_PYTHON_MODULAR)13:31
sonne|workwiking: that is for dependency generation - ignore that13:32
@wikingstill13:32
@wikingi have to somehow pass those 2 defines13:33
@wikingso i'll try now DEFINES=....13:33
@wikingah you mean that find....13:33
@wikingsonne|work: ok got the fix working13:36
-!- gsomix [~Miranda@188.168.2.34] has quit [Ping timeout: 260 seconds]13:39
@wikingok before commiting13:39
@wikingyou can change the configuration for freebsd bot13:40
@wikingchange the ./configure that it has --interfaces=python_modular and --swig=swig2.013:40
@wikinghere13:45
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun13:45
shogun-notifier-shogun: Viktor Gal :develop * 71ac672 / src/configure: https://github.com/shogun-toolbox/shogun/commit/71ac672e34734d79edd490ea371fb88351137eb213:45
shogun-notifier-shogun: Fix for issue #1023 Fix for compiling python_modular on FreeBSD Add new --swig command line argument for ./configure script to be able to define the actual swig executable to use13:45
@wikinglet's check what happens with other buildbots that build python_modular :P13:46
-!- iglesiasg [~nando@n177-p222.kthopen.kth.se] has joined #shogun13:50
-!- mode/#shogun [+o iglesiasg] by ChanServ13:50
@wikingdoh why only deb1 builds :(13:50
shogun-buildbotbuild #718 of cyg1 - libshogun is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/718  blamelist: Viktor Gal <viktor.gal@maeth.com>13:55
@wikingah shit i can test my hypotesis now about RNG :)14:01
@wikingsonne|work: seems that this patch hasn't broke modular interface so you can change the configure command of freebsd bot14:02
sonne|workwiking: what?14:03
-!- travis-ci [~travis-ci@ec2-50-16-143-238.compute-1.amazonaws.com] has joined #shogun14:03
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/675530014:03
-!- travis-ci [~travis-ci@ec2-50-16-143-238.compute-1.amazonaws.com] has left #shogun []14:03
@wikingsonne|work: for freebsd bot14:03
@wikingchange the ./configure command14:03
@wikingadd --swig=swig2.014:04
@wikingand --interfaces=python_modular14:04
@wikingand run python modular tests as well on that bot14:04
van51sonne|work: I can fix the memory leak by making an additional decrease to the reference counter.. but I cannot really explain why it needs it :/14:04
sonne|workvan51: the get increases the refcount and since things are put in a list that is mangled later it is better to have everything in there refcount increased14:05
sonne|worknot that sth dies14:05
sonne|workwiking: can only do this tonight14:06
@wikingokok14:06
@wikingoh lol14:06
@wikingrunning features_sparse_modular.py .. ERROR14:06
van51so should I leave it at this?14:06
@wikingrunning serialization_complex_example.py .. ERROR14:07
-!- gsomix [~Miranda@188.168.128.160] has joined #shogun14:18
-!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has quit [Ping timeout: 245 seconds]14:18
sonne|workvan51: no decrease it when everything is parsed/destroyed14:37
sonne|workat the end14:37
sonne|workwiking: ?14:37
sonne|workdoes sth work?14:37
@wikingsonne|work: mmm yes some of them... but seems there were missing deps for the tests14:38
van51sonne|work: yeah that's what i meant14:38
@wikingfirst it was scipy14:38
@wikingnow it was hdf514:38
van51should I make another PR?14:38
@wikingsonne|work: so still trying to fill the gaps14:38
sonne|workyeah will take time14:38
@wikingok serialization now passes14:40
@wikingso it was 'just' a dependency problem14:40
@wikingfeatures_string_char_compressed_modular.py fails14:40
@wikingi think there's still some package missing14:40
@wikingafaik it was lzo...14:42
@wikinglet's see the new build14:42
@wikingat least now when you'll change the fbsd bot's configure it'll run14:43
@wikingsonne|work: ok examples of python_modular runs fine...14:48
@wikinglet's see the integration tests14:48
@wikingyep14:48
@wikingthe same integration tests fails on freebsd as on osx14:48
@wiking=> they have the same RNG14:49
@wikingso at least now we'll have a buildbot that'll show us when we managed to fix all the integration tests14:50
@wiking:P14:50
sonne|workso we won't enable tests on *bsd/sx just yet14:50
@wikingwell yes and no :P14:51
@wikingsonne|work: so how do you feel about depending on openssl?14:51
@wikingthat has a 'quite' stable RNG14:52
sonne|workwiking: I would rather borrow from opencv14:53
-!- FSCV [~FSCV@187.210.54.166] has joined #shogun14:54
van51sonne|work: I made a PR, please have a look when you can14:58
van51got to go for a while now14:58
van51cya all14:58
-!- van51 [~van51@195.251.235.94] has left #shogun ["QUIT :Leaving."]14:58
-!- ianlini [8c701e90@gateway/web/freenode/ip.140.112.30.144] has joined #shogun15:40
-!- mdragan [8d55007a@gateway/web/freenode/ip.141.85.0.122] has joined #shogun15:53
mdraganHello everyone!15:56
sonne|workhey there.15:56
mdragan@iglesiasg or @lisitsyn  Is anyone of you here? :)15:56
@iglesiasgmdragan: yep15:57
mdraganI just made a ne pull request for issue 863 - dev branch.15:57
mdraganIs it now OK?15:57
@iglesiasgmdragan: yeah, the target branch is fine now :)15:58
mdraganok, fine :)15:59
-!- ianlini [8c701e90@gateway/web/freenode/ip.140.112.30.144] has quit [Ping timeout: 245 seconds]15:59
mdraganThis is what I wanted to check. See you!15:59
@iglesiasgbye16:00
-!- mdragan [8d55007a@gateway/web/freenode/ip.141.85.0.122] has quit [Quit: Page closed]16:00
-!- FSCV [~FSCV@187.210.54.166] has quit [Quit: Leaving]16:08
@wikingsonne|work: ok so take some RNG implementation from an opensouce project and put it into shogun...16:10
sonne|workwiking: yeah with some meaning == working & maintained and well tested16:10
sonne|workohh and fast :)16:10
-!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has joined #shogun16:28
-!- iglesiasg [~nando@n177-p222.kthopen.kth.se] has quit [Quit: leaving]16:44
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout]16:45
@lisitsynre!16:47
sonne|worker!16:49
@lisitsynsonne|work: er!16:49
* gsomix is on the road to home...16:59
gsomixsee you, guys16:59
-!- gsomix [~Miranda@188.168.128.160] has quit [Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org]17:00
-!- sonne|work [~sonnenbu@sams-office-nat.tomtomgroup.com] has quit [Quit: Leaving.]17:06
-!- HeikoS [~heiko@nat-171-79.internal.eduroam.ucl.ac.uk] has joined #shogun17:22
-!- mode/#shogun [+o HeikoS] by ChanServ17:22
thoralfHey Heiko.17:30
@wikingsonney2k: https://github.com/Itseez/opencv/blob/master/modules/core/src/rand.cpp17:40
-!- lisitsyn [~lisitsin@mxs.kg.ru] has left #shogun []17:52
-!- van51 [~van51@athedsl-401908.home.otenet.gr] has joined #shogun17:54
@HeikoSthoralf: hi!18:07
@HeikoSthoralf: I haven't yet looked into the code, but will later today are at latest tomorrow18:07
thoralfHeiko: Okay, thanks.18:13
thoralfHeiko: The example is sooo easy, that there will be an easy explanation, too. ;)18:14
-!- alibezz [~alibezz@177.182.164.8] has joined #shogun18:19
votjakovrHeikoS: I added test case, in which GP regression returns negative variances. So when you'll have a time, please, look at it :)18:29
@HeikoSvotjakovr: does it still return negative variance?18:30
votjakovrHeikoS: Nope18:30
@HeikoSah nice18:30
@HeikoSIll check18:31
votjakovrHeikoS: Thanks18:31
@HeikoShere we ho18:32
@HeikoSgo18:32
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun18:32
shogun-notifier-shogun: Roman Votyakov :develop * bf4df78 / src/shogun/regression/GaussianProcessRegression.cpp: https://github.com/shogun-toolbox/shogun/commit/bf4df784376fbfc011150e304213bc0f99939cf718:32
shogun-notifier-shogun: Fixed bug with negative values for the variance18:32
shogun-notifier-shogun: Roman Votyakov :develop * b30b8ea / / (4 files): https://github.com/shogun-toolbox/shogun/commit/b30b8ea7cc368dfe94ff098dd1986dd64816360418:32
shogun-notifier-shogun: Merge branch 'develop' of https://github.com/shogun-toolbox/shogun into develop18:32
shogun-notifier-shogun: Roman Votyakov :develop * 964362e / tests/unit/regression/GaussianProcessRegression_unittest.cc: https://github.com/shogun-toolbox/shogun/commit/964362ef24b3b154f3de00f5ba1a5a1fd5ec597d18:32
shogun-notifier-shogun: Added test case for GP regression, which checks negative value for the variance18:32
shogun-notifier-shogun: Heiko Strathmann :develop * 0f133a8 / / (2 files): https://github.com/shogun-toolbox/shogun/commit/0f133a89573cfab4eb0492351df0a5cc38d059f518:32
shogun-notifier-shogun: Merge pull request #1030 from votjakovr/develop18:33
shogun-notifier-shogun:18:33
shogun-notifier-shogun: Fixed bug with negative values for the variance in GP regression18:33
shogun-buildbotbuild #719 of cyg1 - libshogun is complete: Success [build successful]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/71918:51
-!- travis-ci [~travis-ci@ec2-50-16-50-250.compute-1.amazonaws.com] has joined #shogun18:56
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/676283118:56
-!- travis-ci [~travis-ci@ec2-50-16-50-250.compute-1.amazonaws.com] has left #shogun []18:56
-!- thoralf [~thoralf@enki.zib.de] has quit [Quit: Konversation terminated!]19:08
-!- sids_aquarius [~sidi_@li400-124.members.linode.com] has quit [Remote host closed the connection]19:13
-!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has quit [Ping timeout: 245 seconds]19:13
-!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has joined #shogun19:30
@sonney2kHeikoS, I think it is too hard to ask people to generate test data - it took ages until we managed to do that ourselves19:33
@HeikoSsonney2k: I think I agree!19:33
@HeikoSjust realised during this epic series of PR19:34
shogun-notifier-shogun: Soeren Sonnenburg :develop * 3823b83 / src/shogun/multiclass/MulticlassSVM.h: https://github.com/shogun-toolbox/shogun/commit/3823b83752592faa508894d014f37685742978ad19:35
shogun-notifier-shogun: fix typos19:35
shogun-notifier-shogun: Soeren Sonnenburg :develop * 94e0a02 / / (3 files): https://github.com/shogun-toolbox/shogun/commit/94e0a023709d86c471b0ccbe39e718e5138b9af019:35
shogun-notifier-shogun: return svm bias in serialization example for all types19:35
@sonney2kHeikoS, after all these are students that are new to everything19:38
@sonney2kso we have to teach them19:38
@sonney2kit is not like they can do *everything* already no matter how good19:38
-!- travis-ci [~travis-ci@ec2-50-16-50-250.compute-1.amazonaws.com] has joined #shogun19:38
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/676463219:38
-!- travis-ci [~travis-ci@ec2-50-16-50-250.compute-1.amazonaws.com] has left #shogun []19:38
shogun-buildbotbuild #721 of cyg1 - libshogun is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/721  blamelist: Soeren Sonnenburg <sonne@debian.org>19:42
shogun-buildbotbuild #1020 of deb3 - modular_interfaces is complete: Failure [failed test ruby_modular]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/1020  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>, Roman Votyakov <votjakovr@gmail.com>19:55
shogun-buildbotbuild #722 of cyg1 - libshogun is complete: Failure [failed compile]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/722  blamelist: Soeren Sonnenburg <sonne@debian.org>19:55
-!- romi__ [~mizobe@187.74.6.200] has quit [Ping timeout: 245 seconds]19:57
@sonney2kshogun-buildbot, force build 'cyg1 - libshogun'19:59
shogun-buildbotbuild forced [ETA 12m16s]19:59
shogun-buildbotI'll give a shout when the build finishes19:59
@wikingsonney2k: how about fbsd bot ?19:59
@sonney2kwiking, what is the configure line again?20:00
@wikingsonney2k: ./configure --cc=clang --cxx=clang++ --interfaces=python_modular --swig=swig2.0 --includes=/usr/local/include --libs=/usr/local/lib --gtest=/home/wiking/gmock-1.6.0/gtest --gmock=/home/wiking/gmock-1.6.0 --prefix=`pwd`/build20:02
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has quit [Quit: Leaving]20:02
@sonney2kwiking, how do we want to do it - change libshogun_freebsd into just freebsd?20:07
-!- romi__ [~mizobe@187.74.6.200] has joined #shogun20:09
@sonney2kwiking, or add another modular_freebsd?20:10
shogun-buildbotbuild #723 of cyg1 - libshogun is complete: Success [build successful]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/72320:14
@sonney2kwiking, tada! https://travis-ci.org/shogun-toolbox/shogun/jobs/676463620:15
-!- shogun-buildbot [~shogun-bu@7nn.de] has quit [Quit: buildmaster reconfigured: bot disconnecting]20:17
-!- shogun-buildbot [~shogun-bu@7nn.de] has joined #shogun20:17
shogun-buildbotbuild #1021 of deb3 - modular_interfaces is complete: Success [build successful]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/102120:17
-!- travis-ci [~travis-ci@ec2-50-16-50-250.compute-1.amazonaws.com] has joined #shogun20:18
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/676463220:18
-!- travis-ci [~travis-ci@ec2-50-16-50-250.compute-1.amazonaws.com] has left #shogun []20:18
@sonney2kshogun-buildbot, force build 'bsd1 - libshogun'20:18
shogun-buildbotbuild forced [ETA 3m33s]20:18
shogun-buildbotI'll give a shout when the build finishes20:18
shogun-buildbotbuild #885 of bsd1 - libshogun is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/88520:18
@sonney2kwiking, travis is green for the first time: https://travis-ci.org/shogun-toolbox/shogun/builds/676463220:18
@wikingsonney2k: amaaaazing!20:20
@sonney2kwiking, any idea why the bsd bot doesn't even configure?20:21
@wikingsonney2k: lolface20:21
@wikinglemme check20:21
@wikingmmmm this is weird20:21
@wikingare we sure it's trying on develop branch?20:21
@wikingmmm20:22
@wikingsonney2k: [wiking@sandbox ~/shogun-buildbot/bsd1_-_libshogun/source]$ git branch20:22
@wiking* master20:22
@wiking:(20:22
@wikingwrong branch20:22
-!- vgorbati_ [d4029f22@gateway/web/freenode/ip.212.2.159.34] has joined #shogun20:22
@wikinghttp://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/88520:23
-!- sids_aquarius [~sidi_@li400-124.members.linode.com] has joined #shogun20:23
@wikingcheck the source stamp20:23
@wikingGot Revisiond10e022a23a26b51e73a217c5014f3cf8d879ee620:23
@wikinghttps://github.com/shogun-toolbox/shogun/commit/d10e022a23a26b51e73a217c5014f3cf8d879ee620:24
@wikingand that's from master branch20:24
@sonney2kwiking, please erase the build dir then20:24
@wikingjust a sec20:24
@wikingbut the source dir is wrong as well20:24
-!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has quit [Ping timeout: 245 seconds]20:24
@sonney2kthat too then20:24
@wikingerased20:24
@wikingoh wait20:25
@wikingdon't start yet20:25
@sonney2kwiking, it might be though that it will try to build master & develop20:25
@wikingok20:25
@wikingshogun-buildbot, force build 'bsd1 - libshogun'20:25
shogun-buildbotbuild forced [ETA 3m33s]20:25
shogun-buildbotI'll give a shout when the build finishes20:25
@wikinglets see20:26
@wikingsonney2k: it again checked out master branch :(20:27
@wikingbut ok it's still fetching20:27
@wikingso maaaybe it'll change after20:27
@sonney2kwiking, it will checkout master & develop20:28
shogun-buildbotbuild #886 of bsd1 - libshogun is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/88620:29
@wikingok now what? :D20:29
@wikingsince this is only master20:30
@wiking:(20:30
@sonney2kwiking, maybe submit to develop?20:30
@wikinghehehe either we force now a dummy commit20:31
@wikingor wait for a new commit20:31
@wikingand see what happens20:31
shogun-notifier-shogun: Soeren Sonnenburg :develop * abf0f1d / src/NEWS: https://github.com/shogun-toolbox/shogun/commit/abf0f1df47ec92031c407a9ef13d3f6dd59fecff20:31
shogun-notifier-shogun: update news20:31
@wikingheheheh20:31
shogun-buildbotbuild #724 of cyg1 - libshogun is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/724  blamelist: Soeren Sonnenburg <sonne@debian.org>20:34
-!- sids_aquarius [~sidi_@li400-124.members.linode.com] has left #shogun ["Off to code | sleep"]20:36
@sonney2kwiking, seems like it is doing develop now20:36
@wikingyeps20:38
@wikingok configure seems went well with detecting everything for python_modular20:39
shogun-buildbotbuild #887 of bsd1 - libshogun is complete: Success [build successful]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/88720:39
@wikingand yeah cool20:39
@wikingeverything went fine20:39
@wikingalthough it did not run the integration tests20:40
@wikingwhich is cool20:40
@wikingas it would have failed20:40
@sonney2kwiking, any thoughts how we can mirror the github repository to shogun-toolbox.org & code.google.com?20:46
@wikingsonney2k: mmm why do u want to mirror?20:54
@sonney2kwiking, everything20:54
@wikingsonney2k: i mean for shogun-toolbox.org should be easy as i guess u have a shell for it :P20:54
@wikingsonney2k: so just a simple cron script for mirroring20:54
@sonney2kI dont' know if it is good - but this year we will have likely 20-30 student proposals only20:55
@wikingonly? :)20:55
@sonney2kI think we are raising the bar to enter year by year getting fewer submissions20:55
@wikinglast year we had more?20:55
@sonney2kwiking, yes - 2 years ago we had >7020:55
@wikingah and last year?20:55
@sonney2klast year we hard ~4520:56
@sonney2khad20:56
@wikingah ok20:56
@wikingdoes the quality keeps at least the same?20:56
@sonney2kbut again quality of proposals improved year by year20:56
@wikingah ok20:56
@wikingwell20:56
@wikinghow many slots did u apply for last year?20:56
@wikingi mean this year20:56
@wiking?20:56
@sonney2kit is good for us (not so many reviews) but not clear if this will reduce #slots20:57
@sonney2knot sure yet20:57
@wikingah slots are not decided yet?20:57
@sonney2kI think about 1020:57
@sonney2kbut that woudl be really a lot20:57
@sonney2kif we get them20:57
@sonney2kwiking, btw here is the graph from 2012 http://shogun-toolbox.org/page/Events/gsoc201220:58
@sonney2k(scroll down a bit)20:58
@wikingyeah i remember :P20:58
@wikingthe last minute proposals20:58
@wikingwell we have 3 more days ;)20:59
@sonney2kwe got 12 applications by now21:00
@sonney2kmaybe 7 HQ21:00
@sonney2kdefinite want21:00
@wikingbut hey last year we had like 9 no?21:08
@wikingi mean i think that 7 will get the slots :)21:09
@wikingapache gets like crazy amount of slots21:09
@wikingafaik they apply as apache foundation in gsoc21:10
@wikingbut there's like mahout, lucene, hadoop etc.21:10
@wikingand that's like at least 5 slots apache subproject21:11
-!- gsomix [~Miranda@109.169.243.24] has joined #shogun21:11
gsomixgood evening21:11
gsomixI'm at home finally21:11
@wiking\o/ gsomix21:12
@wikinghow many kms have u travelled?21:12
gsomixwiking: just 100+ kms, but 3 hours in bus: hard traffic in the Samara21:17
@wikingufff21:18
@wikingcrazy21:18
* wiking just got graphsearch on his fb... it's pretty nice what they put together :P21:18
gsomixwiking: graphsearch?21:24
vgorbati_ gsomix: do you live in the same town with lisitsyn?21:24
gsomixvgorbati_: yep.21:27
vgorbati_gsomix: do you know each other?)21:27
@wikingvgorbati_: :)21:27
@wikingi guess samara has only 2 resident who uses shogun ;P21:28
@wikingthat's gsomix and lisitsyn21:28
vgorbati_:D21:28
gsomixvgorbati_: yes, we live in one room in the hostel21:28
vgorbati_gsomix: ah, I did not know :)21:29
@wikingvgorbati_: they are a couple ;P21:29
@wikinggsomix: can u come to the workshop?21:29
vgorbati_wiking: aha21:29
@wikingor will u?21:29
gsomixwiking: nope. I will keep the law and order in Samara, while he is in Berlin :321:33
@wikingtoo bad :(21:33
gsomix*he == lisitsyn21:36
-!- lisitsyn [~blackburn@109-226-120-160.clients.tlt.100megabit.ru] has joined #shogun21:36
-!- mode/#shogun [+o lisitsyn] by ChanServ21:36
@lisitsynsonney2k: so you are into reviewing now? ;)21:36
gsomixlisitsyn: hello, honey :]21:37
@lisitsynno money no honey!21:37
gsomix:(21:38
shogun-notifier-shogun: vladislav.horbatiuk@gmail.com :develop * 5a2df25 / / (4 files): https://github.com/shogun-toolbox/shogun/commit/5a2df2554ad6b4c08c462ec2cfcac4c9d37dfc3021:53
shogun-notifier-shogun: Added a wrapper for the t-SNE DR method; added an integration test for it (black-listed for now).21:53
shogun-notifier-shogun: Sergey Lisitsyn :develop * 7df41b2 / / (4 files): https://github.com/shogun-toolbox/shogun/commit/7df41b21aa46e905d4f40da27f8bad05352ea24a21:53
shogun-notifier-shogun: Merge pull request #1027 from vladislav-horbatiuk/vgorbati-develop21:53
shogun-notifier-shogun:21:53
shogun-notifier-shogun: Added a wrapper for the t-SNE DR method; added an integration test for it (black-listed for now).21:53
gsomixnew photo from hipsters^WCurioisty http://cs405418.vk.me/v405418490/c09d/GZ0Oe0jKu58.jpg21:58
gsomixCurioisty is in touch again21:58
shogun-buildbotbuild #888 of bsd1 - libshogun is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/888  blamelist: vladislav.horbatiuk@gmail.com <vladislav.horbatiuk@gmail.com>21:58
shogun-buildbotbuild #725 of cyg1 - libshogun is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/725  blamelist: vladislav.horbatiuk@gmail.com <vladislav.horbatiuk@gmail.com>21:59
gsomix*Curiosity22:03
shogun-buildbotbuild #726 of cyg1 - libshogun is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/726  blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com>22:07
shogun-buildbotbuild #889 of bsd1 - libshogun is complete: Success [build successful]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/88922:12
-!- wencan [~wencan@c-71-61-182-121.hsd1.pa.comcast.net] has quit [Quit: wencan]22:37
shogun-notifier-shogun: Sergey Lisitsyn :develop * b370f6b / src/shogun/lib/tapkee/ (2 files): https://github.com/shogun-toolbox/shogun/commit/b370f6b807d58953f8e29b6c7cdb6fb433f1cf6e22:42
shogun-notifier-shogun: Ported a few improvements from tapkee22:42
@lisitsynsonney2k: wiking: we are not going to have 20 proposals, that's not true22:54
@lisitsyn16 already + last day/hour/minute..22:54
gsomixnite23:01
shogun-buildbotbuild #727 of cyg1 - libshogun is complete: Success [build successful]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/72723:03
-!- vgorbati_ [d4029f22@gateway/web/freenode/ip.212.2.159.34] has quit [Quit: Page closed]23:10
shogun-buildbotbuild #1025 of deb3 - modular_interfaces is complete: Failure [failed test ruby_modular]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/1025  blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com>23:17
shogun-buildbotbuild #1026 of deb3 - modular_interfaces is complete: Success [build successful]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/102623:27
alibezzlisitsyn: sorry for the dull question, but how do i clone the develop branch?23:29
alibezzi'm finally down to work on my issue ;)23:29
alibezzlisitsyn: maybe that i guess: https://github.com/shogun-toolbox/shogun.git i'll check it out23:33
alibezzlisitsyn: nope it always drives me to the master branch. is the develop branch remote??23:34
alibezzlisitsyn: chill i found it. it was a remote one. ;)23:36
@lisitsynalibezz: git checkout develop23:36
@lisitsynin your local repo23:36
alibezzyup found it thx!23:36
@lisitsynalibezz: sorry have to sleep now - if you have any questions please mail me or ask here I'll answer and then you can read in logs23:38
alibezzlisitsyn: no worries ;) thx23:39
--- Log closed Wed May 01 00:00:29 2013