Open in new window / Try shogun cloud
--- Log opened Sat May 11 00:00:36 2013
--- Day changed Sat May 11 2013
@sonney2kdvalcarce, or which branch is it?00:00
dvalcarcesonney2k, classification branch should work00:01
dvalcarcesonney2k, I can take a screenshot if you want00:01
shogun-notifier-shogun: Sergey Lisitsyn :develop * f511ae1 / examples/undocumented/python_modular/converter_factoranalysis_modular.py: https://github.com/shogun-toolbox/shogun/commit/f511ae16085d133b4db6e3e03e49d343312ca02700:02
shogun-notifier-shogun: Temporary fix for Factor Analysis example that is undeterministic,00:02
shogun-notifier-shogun: checks determinant of covariance matrix is positive00:02
shogun-notifier-shogun: cameron :develop * af7e5c1 / examples/undocumented/python_modular/graphical/interactive_gp_demo.py: https://github.com/shogun-toolbox/shogun/commit/af7e5c1639efa18b06b803e4df61c02dba1fb1b800:03
shogun-notifier-shogun: Additional features for the interactive gp regression demo (issue #948)00:03
shogun-notifier-shogun: Heiko Strathmann :develop * d906aa9 / examples/undocumented/python_modular/graphical/interactive_gp_demo.py: https://github.com/shogun-toolbox/shogun/commit/d906aa956e412fe64fbfaca5a9c56b74cf754af900:03
shogun-notifier-shogun: Merge pull request #1082 from goldbug/develop-gp-example00:03
shogun-notifier-shogun:00:03
shogun-notifier-shogun: Additional features for the interactive gp regression demo (issue #948)00:03
@sonney2kdvalcarce, hmmh I am on classification branch00:03
@sonney2kdvalcarce, which method then?00:03
@sonney2khttps://github.com/dvalcarce/shogun-web/branches00:03
@sonney2kit says last updated 13 days ago?00:04
dvalcarce13 days ago? hmm...00:04
dvalcarcesonney2k, try now00:06
@sonney2kdvalcarce, ok now I see the colorbar00:07
@sonney2kthough it still needs some numbers :D00:08
@sonney2kbut no contour00:08
dvalcarcesonney2k, I will add the numbers00:09
@sonney2kdvalcarce, hmmh  multiclass is broken now?00:09
dvalcarcesonney2k, multiclass works fine to me00:10
@sonney2kweird00:11
-!- goldbug [~quassel@global-2-1.nat.csx.cam.ac.uk] has quit [Remote host closed the connection]00:11
@sonney2kif you have time please do a real contour like in http://www.idlcoyote.com/tips/contour_user_selected.jpg00:11
dvalcarcesonney2k, do you receive some error message?00:12
@sonney2kI don't know how hard this is with d3 / or whether anyone has done it00:12
@sonney2kdvalcarce, no the result is just crap00:12
dvalcarcesonney2k, browser?00:13
-!- travis-ci [~travis-ci@ec2-23-20-77-42.compute-1.amazonaws.com] has joined #shogun00:13
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/706693900:13
-!- travis-ci [~travis-ci@ec2-23-20-77-42.compute-1.amazonaws.com] has left #shogun []00:13
dvalcarcesonney2k, binary classification has the same contour plotting as the perceptron00:13
@lisitsynsonney2k: where did you get that pic!!!!! :D00:13
@lisitsynthat's windows 95 for real!00:13
dvalcarcesonney2k, but perceptron countor plotting is ugly: just parallel lines00:14
@sonney2kdvalcarce, https://dl.dropboxusercontent.com/u/31480471/bad-mc.png00:14
@sonney2klisitsyn, google contour plot00:14
@sonney2kdvalcarce, yes colors & lines for 0 / 1 / or so00:15
@sonney2kwithout knowing the f(x)=0 contour line it is not clear where the decision boundary really is00:16
@sonney2kalright people I have to sleep00:16
@sonney2klisitsyn, please fixzebugz00:16
@lisitsynsonney2k: which?00:16
@sonney2kFA00:16
@lisitsynsonney2k: already?00:17
dvalcarcesonney2k, don't you see something like https://github.com/shogun-toolbox/shogun-web/pull/27#issuecomment-17746979 ?00:17
shogun-buildbotbuild #796 of cyg1 - libshogun is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/796  blamelist: cameron <cameronjml2@gmail.com>00:17
@sonney2kdvalcarce, yes but what am I seeing there?00:17
@sonney2kI meant black contour lines00:18
@sonney2kand below that the  *smooth* colors00:18
dvalcarcesonney2k, black?00:18
dvalcarcesonney2k, I have to find how to do that in D3js then00:19
-!- pickle27 [~kevin@rcv3-lab-pc.ee.queensu.ca] has joined #shogun00:20
-!- pickle27 [~kevin@rcv3-lab-pc.ee.queensu.ca] has quit [Client Quit]00:20
-!- pickle27 [~kevin@rcv3-lab-pc.ee.queensu.ca] has joined #shogun00:21
dvalcarcesonney2k, I use conrac.js library for contour plotting, I don't think I can add black contour lines without changing all plotting00:21
-!- pickle27 [~kevin@rcv3-lab-pc.ee.queensu.ca] has quit [Remote host closed the connection]00:22
-!- pickle27 [~kevin@rcv3-lab-pc.ee.queensu.ca] has joined #shogun00:23
@sonney2kpickle27, dvalcarce, van51 btw which timezone are you in?00:24
pickle27sonney2k: Canada - EDT00:24
* gsomix is tired. *bad, bad git* :(00:24
dvalcarcesonney2k, Madrid CET00:24
van51sonney2k: athens or UTC+200:25
gsomixnite00:25
@sonney2kok so we should not have cheng in too :D00:25
@sonney2klisitsyn, recall last year when we had our monthly meetings and it was already late at night for cheng while chris had to get up at 7 or so?00:26
@lisitsynsonney2k: haha yes00:26
@lisitsynwe were basically00:26
@lisitsyn-8 to +800:26
@sonney2kaustralia vs new york (city)00:26
@sonney2kwe could totally screw that with olvier (california)00:27
van51I think I would mind more the getting up at 7 :p00:27
@sonney2kand cheng00:27
@sonney2kvan51, it is 00:28 here00:28
@sonney2kso 18:28 at your place00:28
@sonney2k6 hrs00:28
@lisitsynlucky you 02:28 here00:28
@sonney2kif we had a meeting at 13:0000:29
@sonney2kit would indeed be 7 am for you00:29
van51no, 00:28 sounds fine00:29
van51:p00:29
@wikingsonney2k: how should a load and save method look like for a preprocessor00:29
@sonney2knot when you have to get up in 5 hours :D00:30
@sonney2kwiking, a what?00:30
@sonney2kI guess that is obsolete00:30
gsomixsonney2k, sorry for another PR. I have problems with "amending".00:30
@wikingsonney2k: is it?00:30
@wikingsonney2k: somehow i need to serialize some parameters of the preprocessor00:30
@sonney2kgsomix, jsut do git reset --soft <version>00:30
-!- travis-ci [~travis-ci@ec2-54-242-186-200.compute-1.amazonaws.com] has joined #shogun00:31
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/706718600:31
-!- travis-ci [~travis-ci@ec2-54-242-186-200.compute-1.amazonaws.com] has left #shogun []00:31
@sonney2kand then git commit everything at once00:31
@sonney2kand push --force00:31
@sonney2kwiking, but then SG_ADD00:31
@wikingsonney2k: and then save/load functions will be handled automagically? :)00:31
@sonney2kdvalcarce, maybe you can add contour lines on top? I have no clue what conrac.js is00:32
@sonney2kwiking, what?00:32
shogun-buildbotbuild #1091 of deb3 - modular_interfaces is complete: Failure [failed test python_modular]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/1091  blamelist: Viktor Gal <viktor.gal@maeth.com>00:32
@sonney2kload and save should die00:32
@wikingsonney2k: ok i'll remove00:33
gsomixsonney2k, hm, ok. I'll try it next time. :(00:33
@sonney2klisitsyn, I like gsomix' label factory patch00:35
@lisitsynsonney2k: no I like it too, just naming00:36
@lisitsynit is not factory00:36
@sonney2kwhat else then?00:36
@lisitsynsonney2k: factory creates objects00:36
shogun-buildbotbuild #797 of cyg1 - libshogun is complete: Success [build successful]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/79700:37
@sonney2klisitsyn, LabelUtils?00:39
* sonney2k Zzzzz....00:39
@lisitsynsonney2k: utils is better indeed00:39
gsomixgood night, finally00:39
@sonney2kgsomix, ok then00:39
@sonney2kLabelUtils00:39
gsomixsonney2k, ok.00:39
@sonney2kzzzzz....00:39
@lisitsynLabelsUtils::cast_to_structured00:40
@lisitsynI'd call it so00:40
@lisitsyngsomix: ^00:40
@lisitsynI like short names but Utils doesn't tell the user what it does00:40
@wikingwhy do i get segfault for sg_adding a SGVector? :)00:46
@wikingok just another make clean && make story00:48
-!- travis-ci [~travis-ci@ec2-23-20-77-42.compute-1.amazonaws.com] has joined #shogun00:48
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/706734300:48
-!- travis-ci [~travis-ci@ec2-23-20-77-42.compute-1.amazonaws.com] has left #shogun []00:48
van51sonney2k: are you gone?00:49
@wikinglisitsyn: apply your magic to unit test compilation as well! :)00:57
@lisitsynwiking: what magic?00:57
shogun-notifier-shogun: Viktor Gal :develop * 54ea3ad / / (4 files): https://github.com/shogun-toolbox/shogun/commit/54ea3ad40ecd4ac5ce1704d783d1378eeea5cb3300:57
shogun-notifier-shogun: Refactor RescaleFeatures and fix PruneVarSubMean::cleanup RescaleFeatures should be used with init() in order to aquire the min and range values for each vector. PruneVarSubMean::cleanup() did not set initialized to false00:57
@wikinglisitsyn: to hide the actual compilation command00:58
@lisitsynahh00:58
@lisitsynyeah tomorrow00:58
@wikingok so now the rescalefeature should actually work :P00:58
@wikingcool00:58
@lisitsynwiking: are you still using clang_complete?00:58
@lisitsynI just updated and I noticed it is faster now00:59
@wikingit's cool that we have now already 86 tests in unit tests00:59
@wikinglisitsyn: in a way yes....00:59
shogun-buildbotbuild #1092 of deb3 - modular_interfaces is complete: Failure [failed test python_modular]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/1092  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>, Sergey Lisitsyn <lisitsyn.s.o@gmail.com>, cameron <cameronjml2@gmail.com>01:00
@lisitsynwiking: whooop01:04
@lisitsynno more ctags01:04
@lisitsynwiking: try jumping01:05
@lisitsynCtrl + ] here01:05
@lisitsynfinally01:05
pickle27hey is there any docs on how to build the test suit for unit tests?01:05
@lisitsynpickle27: build unit tests or what?01:06
pickle27lisitsyn: yeah I am trying to build and run the unit tests to see of the new one I'm working on works01:07
@lisitsynpickle27: make in shogun/tests/unit01:07
@lisitsyndoes it for you01:07
@wikinglisitsyn: didn't hide swig compilation01:07
@lisitsynwiking: yeah I know01:07
@lisitsynwiking: but hey I said jumping works01:08
@lisitsyn:D01:08
pickle27lisitsyn: yes but what version of gtest and gmock because the build is failing01:08
@lisitsynpickle27: error?01:08
pickle27lisitsyn: gtest/test/gtest_unittest.cc:69:36: fatal error: src/gtest-internal-inl.h: No such file or directory01:08
pickle27lisitsyn: I just downloaded gtest and gmock and copied them to /usr/src01:09
@lisitsynpickle27: did you reconfigure after that?01:09
pickle27yup01:09
pickle27lisitsyn: they both showed up as recognized after ./config01:10
@lisitsynpickle27: so /usr/src/gtest/src/gtest-internal-inl.h exists01:10
@lisitsyn?01:10
pickle27lisitsyn: yes01:11
@lisitsynpickle27: no idea then :D01:12
@lisitsynpickle27: just check they are included01:12
@lisitsynpickle27: shogun/src/.config should have01:12
@lisitsynGTEST_DIR01:12
@lisitsynGMOCK_DIR01:12
@lisitsynset to /usr/src/gtest01:13
@lisitsynand /usr/src/gmock01:13
pickle27lisitsyn: yeah thats what they are set to01:13
@lisitsynpickle27: sorry I am powering off just like sonney2k now01:15
pickle27lisitsyn: no worries I'll try and figure this out!01:15
@lisitsynpickle27: this error means one thing - include paths are not here01:16
pickle27lisitsyn: yeah I'm trying a rebuild now and I'll take another look at exactly what it was building]01:16
pickle27lisitsyn: looks like it might be okay now01:17
@lisitsynoh these lovely side effects01:17
shogun-notifier-shogun: Sergey Lisitsyn :develop * e885cc8 / src/shogun/lib/tapkee/ (4 files): https://github.com/shogun-toolbox/shogun/commit/e885cc84c9efe6b269a006aacf0ec1318656e2f401:21
shogun-notifier-shogun: Fixes for tapkee01:21
shogun-notifier-shogun:01:21
shogun-notifier-shogun: - Proper ratio limits for landmark methods01:21
shogun-notifier-shogun: - Fixed crasher in exact t-SNE01:21
@wikinglisitsyn: went already to sleep? :)01:24
@lisitsynwiking: on the way01:24
@lisitsynwiking: what's up?01:25
-!- travis-ci [~travis-ci@ec2-23-20-77-42.compute-1.amazonaws.com] has joined #shogun01:25
travis-ci[travis-ci] it's Sergey Lisitsyn'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/706798401:25
-!- travis-ci [~travis-ci@ec2-23-20-77-42.compute-1.amazonaws.com] has left #shogun []01:25
shogun-buildbotbuild #799 of cyg1 - libshogun is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/799  blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com>01:25
@wikinglisitsyn: decision tree implementation01:26
@wikinganybody started that?01:26
@lisitsynwiking: monica may be01:27
@lisitsynshe wanted to start with id301:28
@lisitsynwiking: we rather need RF actually01:28
@lisitsynone tree is useless01:28
@wikinglisitsyn: i know....01:30
@wikinglisitsyn: that's what i wanted but stupid librf supports only binary classification which is kind of stupid01:31
-!- travis-ci [~travis-ci@ec2-54-242-220-92.compute-1.amazonaws.com] has joined #shogun01:41
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/706801701:41
-!- travis-ci [~travis-ci@ec2-54-242-220-92.compute-1.amazonaws.com] has left #shogun []01:41
-!- dvalcarce [5b757556@gateway/web/freenode/ip.91.117.117.86] has quit [Quit: Page closed]01:41
shogun-buildbotbuild #1093 of deb3 - modular_interfaces is complete: Failure [failed test python_modular]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/1093  blamelist: Viktor Gal <viktor.gal@maeth.com>01:44
-!- travis-ci [~travis-ci@ec2-54-242-186-200.compute-1.amazonaws.com] has joined #shogun01:53
travis-ci[travis-ci] it's van51's turn to pay the next round of drinks for the massacre he caused in van51/shogun: http://travis-ci.org/van51/shogun/builds/706977201:53
-!- travis-ci [~travis-ci@ec2-54-242-186-200.compute-1.amazonaws.com] has left #shogun []01:53
van51hmm01:58
van51how do I turn that off?01:58
van51:)01:59
pickle27lisitsyn: aaah all I needed was a make clean...02:04
pickle27lisitsyn: oh nvm problem persists02:04
pickle27lisitsyn: ahhh now I got it running silly mistake on my part02:06
shogun-buildbotbuild #1094 of deb3 - modular_interfaces is complete: Failure [failed test python_modular]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/1094  blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com>02:11
-!- van51 [~van51@athedsl-320452.home.otenet.gr] has left #shogun ["QUIT :Leaving."]02:17
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has joined #shogun02:35
-!- pickle27 [~kevin@rcv3-lab-pc.ee.queensu.ca] has quit [Ping timeout: 245 seconds]03:08
-!- pickle27 [~kevin@rcv3-lab-pc.ee.queensu.ca] has joined #shogun03:09
-!- nube [~rho@49.244.141.97] has joined #shogun03:19
-!- zxtx_ [~zv@cpe-75-83-151-252.socal.res.rr.com] has joined #shogun03:27
-!- zxtx [~zv@cpe-75-83-151-252.socal.res.rr.com] has quit [Ping timeout: 245 seconds]03:27
-!- flxb_ [~flxb@master.ml.tu-berlin.de] has joined #shogun03:34
-!- flxb [~flxb@130.149.108.2] has quit [Ping timeout: 245 seconds]03:37
-!- travis-ci [~travis-ci@54.242.186.200] has joined #shogun03:37
travis-ci[travis-ci] it's Sergey Lisitsyn'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/706958003:37
-!- travis-ci [~travis-ci@54.242.186.200] has left #shogun []03:37
-!- kakashi_1 [~kakashi_@bumbum.lsi.upc.edu] has joined #shogun03:58
-!- shogun-buildbot_ [~shogun-bu@7nn.de] has joined #shogun04:01
-!- sijin_ [~smuxi@144.214.222.109] has joined #shogun04:01
-!- Netsplit *.net <-> *.split quits: kakashi_, sijin04:03
-!- shogun-buildbot [~shogun-bu@7nn.de] has quit [Ping timeout: 252 seconds]04:07
-!- shogun-buildbot [~shogun-bu@7nn.de] has joined #shogun04:08
-!- shogun-buildbot_ [~shogun-bu@7nn.de] has quit [Ping timeout: 245 seconds]04:09
-!- sijin_ [~smuxi@144.214.222.109] has quit [Ping timeout: 245 seconds]04:09
-!- sijin_ [~smuxi@144.214.222.109] has joined #shogun04:09
-!- Netsplit *.net <-> *.split quits: kakashi_104:11
-!- Netsplit over, joins: kakashi_104:15
shogun-buildbotbuild #391 of nightly_default is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/nightly_default/builds/39104:17
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has quit [Quit: Leaving]04:18
-!- zxtx__ [~zv@cpe-75-83-151-252.socal.res.rr.com] has joined #shogun04:23
-!- Netsplit *.net <-> *.split quits: zxtx_, pickle27, shogun-notifier-04:26
-!- flxb [~flxb@master.ml.tu-berlin.de] has joined #shogun04:53
-!- Netsplit *.net <-> *.split quits: shogun-buildbot, flxb_04:53
-!- Netsplit over, joins: shogun-buildbot04:54
-!- sijin_ [~smuxi@144.214.222.109] has quit []05:50
-!- lambday [3ba2174d@gateway/web/freenode/ip.59.162.23.77] has joined #shogun06:55
@sonney2klisitsyn, then we should not call it LabelUtils btw sth else07:06
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has joined #shogun07:42
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has quit [Client Quit]07:44
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has joined #shogun07:48
-!- nube [~rho@49.244.141.97] has quit [Remote host closed the connection]08:01
-!- nube [~rho@49.244.141.97] has joined #shogun08:02
-!- nube [~rho@49.244.141.97] has quit [Quit: Leaving.]08:27
gsomixsonney2k, mb sth like LabelsCast::to_binary?08:33
-!- nube [~rho@49.244.141.97] has joined #shogun08:34
gsomixok, good morning finally09:33
-!- van51 [~van51@athedsl-320452.home.otenet.gr] has joined #shogun10:37
-!- van51 [~van51@athedsl-320452.home.otenet.gr] has left #shogun ["QUIT :Leaving."]10:43
-!- van51 [~van51@athedsl-320452.home.otenet.gr] has joined #shogun10:45
-!- goldbug [~quassel@global-2-1.nat.csx.cam.ac.uk] has joined #shogun10:51
-!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has joined #shogun11:19
van51hello, since last night whenever I'm compiling I'm getting " a /usr/lib/gcc/i686-linux-gnu/4.6/include/emmintrin.h:32:3: error: #error "SSE2 instruction set not enabled" " message11:24
van51anyone knows what's that about?11:24
-!- nube [~rho@49.244.141.97] has quit [Ping timeout: 276 seconds]11:50
goldbugthis might help11:53
goldbughttp://stackoverflow.com/questions/16410149/error-sse2-instruction-set-not-enabled-when-including-emmintrin-h11:53
van51goldbug: thanks, I had seen it.. and I think I have set the flags it says, but I am still getting this error12:11
goldbugmaybe reinstall gcc and cmake12:13
van51actually I was building with make12:14
goldbugvan51: can't help then, didn't have that problem at all, sorry12:19
van51goldbug: no worries, thanks for your time anyway12:19
@lisitsynvan51: which file it is?12:23
@lisitsynI mean is it shogun?12:24
van51lisitsyn: yea12:24
@lisitsynvan51: where emmintrin.h is included?12:28
van51lisitsyn: yea, the error comes from that file12:29
@lisitsynvan51: and it must be included somewhere else12:30
van51what do you mean?12:32
goldbuglisitsyn: Hi, just wondering whether shogun has a periodic kernel?12:40
-!- goldbug [~quassel@global-2-1.nat.csx.cam.ac.uk] has quit [Remote host closed the connection]12:52
@sonney2klisitsyn, what is you name preference then? I am totally fine with factory  btw...13:04
@wikingvan51: is that produced by ./configure or make?13:05
van51wiking: by make13:06
@wikingvan51: woah13:06
van51it spams it in the beginning13:06
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun13:06
shogun-notifier-shogun: Evgeniy Andreev :develop * 2ab348a / / (60 files): https://github.com/shogun-toolbox/shogun/commit/2ab348a759ec59698f804665deaba8029887315013:06
shogun-notifier-shogun: Issue 1058: added factory for labels13:06
shogun-notifier-shogun: Soeren Sonnenburg :develop * 90a5fc4 / / (60 files): https://github.com/shogun-toolbox/shogun/commit/90a5fc4c32a171a38c4ad21bc1f98458800aca8513:06
shogun-notifier-shogun: Merge pull request #1083 from gsomix/feature/issue105813:06
shogun-notifier-shogun:13:06
shogun-notifier-shogun: added factory for labels13:06
@wikingcan you copy-paste the context plz13:06
van51wiking: what do you mean by context? :)13:07
@wikingvan51: well like 2 lines before this error13:08
van51it has13:08
van51make -C shogun13:08
van51make[1]: Entering directory `/home/van51/shogun/src/shogun'13:08
van51Generating dependencies13:08
@wikingmmmmhm13:09
van51and then it spams it for a while13:09
@wikingvan51: can you send me your configure.log ?13:09
@wikingvan51: wiking@maeth.com13:09
@wikingthnx13:09
gsomixsonney2k, oh... it was WIP. :(13:09
shogun-buildbotbuild #1152 of deb1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/1152  blamelist: Soeren Sonnenburg <sonne@debian.org>13:10
van51wiking: sent13:11
@wikingvan51: and after this basically you do a make and it complains that error: #error "SSE2 instruction set not enabled" " messag13:14
van51wiking: exactly13:14
@wikingshit13:14
@wikingi wonder why13:15
van51wiking: dunno, but it is working fine on another machine I have13:15
van51so it has something to do with the configuration on my laptop13:15
van51the weirder thing is that it suddenly started last night :S13:15
shogun-buildbotbuild #1151 of deb1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/1151  blamelist: Evgeniy Andreev <gsomix@gmail.com>13:16
@wikingvan51: i guess it's because you upgraded to HEAD of the repo13:17
@sonney2kgsomix, you label everything WIP so how could I figure this out?13:18
@wikingvan51: can u paste on file's compilation line.... ?13:19
@wikinglisitsyn: how do i turn off this 'hide compilation mode'13:19
gsomixsonney2k, hm. I usually say when it's ready to merge...13:23
gsomixsonney2k, sorry. I'll try to be more careful with WIP tag.13:23
van51wiking: if you mean the line of the error, then there is nothing more to the output13:24
@wikinglisitsyn: "We just can't be fucked to implement C11 support, and variadic templates were just too hard to implement in a year. (But ohmygosh we turned "^" into a reference-counted pointer operator. Oh, and what's a reference cycle?) " :D13:26
-!- travis-ci [~travis-ci@ec2-54-242-186-200.compute-1.amazonaws.com] has joined #shogun13:27
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/707703813:27
-!- travis-ci [~travis-ci@ec2-54-242-186-200.compute-1.amazonaws.com] has left #shogun []13:27
van51wiking: after a while this error message stops and it does a  "( find . -name "*.i" | cut -c 3- ; find ../shogun/ -name "*.h" ... "13:27
van51it re-enters the directory13:27
van51and says that it is compiling13:27
van51but if I run any test after that, it fails with a segfault13:27
@sonney2kgsomix, well now it is too late fix the rest ASAP13:27
gsomixsonney2k, aha, I'm fixing now.13:29
gsomixsonney2k, btw, what do you think about "LabelsCast" naming?13:31
-!- iglesiasg [c1934d0c@gateway/web/freenode/ip.193.147.77.12] has joined #shogun13:54
-!- mode/#shogun [+o iglesiasg] by ChanServ13:55
@iglesiasggreetings people13:55
@iglesiasgI am finally back13:55
gsomixiglesiasg, \o/13:56
-!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has quit [Ping timeout: 245 seconds]13:57
@iglesiasggsomix: let met ask you14:00
@iglesiasgwhy is it necessary to add gsomix the declaractions class CLabels, CBinaryLabels, etc in LabelsFactory.h14:00
@iglesiasgI never really understand it14:00
@iglesiasgalthough I think it is required14:01
gsomixiglesiasg, it's just predeclaration. I have included LabelsFactory.h in Labels.h - at this moment BinaryLabels and etc. are not really defined.14:03
gsomixiglesiasg, btw, it seems, that CLabels declaration is not necessary.14:03
gsomixiglesiasg, ah, nope. It's needed too.14:07
@iglesiasggsomix: aham, I see it. I didn't notice BinaryLabels.h, etc are not included.14:08
@iglesiasggsomix: I think there is something related to SWIG that makes you have to do it even if you include them14:09
gsomixsonney2k, ok, now it's really WIP.14:32
gsomixsonney2k, once it is ready, I will write to you and will remove tag "WIP".14:33
-!- van51 [~van51@athedsl-320452.home.otenet.gr] has left #shogun ["QUIT :Leaving."]14:37
-!- nube [~rho@49.244.48.223] has joined #shogun14:41
-!- nube [~rho@49.244.48.223] has quit [Quit: Leaving.]15:23
-!- nube [~rho@49.244.48.223] has joined #shogun15:25
gsomixsonney2k, bit.ly/W6drdM15:44
gsomixsonney2k, I mean PR is good to merge now. :)15:44
-!- nube [~rho@49.244.48.223] has quit [Quit: Leaving.]15:47
-!- nube [~rho@49.244.48.223] has joined #shogun15:47
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout]16:06
-!- abinash [75c8eae6@gateway/web/freenode/ip.117.200.234.230] has joined #shogun16:10
abinashHello, please someone help me in getting to know about maximum likelihood II framework for Gaussian process regression?16:26
@iglesiasgabinash: hi!16:32
@iglesiasgare you stuck in any part of it?16:32
-!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has joined #shogun16:32
abinashiglesiasg: actually I am trying to work on the issue #108016:33
abinashI am unable to understand what is actually maximum likelihood II framework for Gaussian Process Regression16:34
abinashiglesiasg: please help me on this16:38
@iglesiasgabinash: have you read in any book, paper or other reference about this?16:41
abinashiglesiasg: no, but I would prefer to read16:41
@iglesiasgabinash: I guess it is a good idea to check a reference in order to understand it ;)16:42
abinashiglesiasg: ok, I would do that.16:42
@iglesiasgvotjakovr: according to the issue votjakovr is taking a look at that, maybe he can suggest a reference16:43
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has quit [Ping timeout: 264 seconds]16:49
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has joined #shogun16:49
abinashvotjakovr: can you suggest me any reference for maximum likelihood II framework for Gaussian process regression16:55
votjakovrabinash: Hi, ML2 framework make it possible to perform gradient descent to select the best hyperparameters. I think this http://www.youtube.com/watch?v=1CFof6hU3cI video lecture will be useful for introduction in ML2.17:03
-!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has quit [Ping timeout: 245 seconds]17:22
-!- iglesiasg [c1934d0c@gateway/web/freenode/ip.193.147.77.12] has quit [Quit: Page closed]17:52
-!- nube [~rho@49.244.48.223] has quit [Ping timeout: 245 seconds]18:10
@lisitsynre18:12
@lisitsynwiking: why?18:12
-!- nube [~rho@49.244.40.224] has joined #shogun18:23
-!- Netsplit *.net <-> *.split quits: kakashi_118:28
-!- Netsplit over, joins: kakashi_118:29
-!- zxtx_ [~zv@cpe-75-83-151-252.socal.res.rr.com] has joined #shogun19:02
-!- Netsplit *.net <-> *.split quits: votjakovr, zxtx__19:07
-!- Netsplit over, joins: votjakovr19:15
-!- hushell [43bd6474@gateway/web/freenode/ip.67.189.100.116] has joined #shogun19:18
-!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has joined #shogun19:25
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun19:27
shogun-notifier-shogun: Evgeniy Andreev :develop * 5d991aa / / (28 files): https://github.com/shogun-toolbox/shogun/commit/5d991aa9beada440888e45b5651efa810eccc3aa19:27
shogun-notifier-shogun: fixed examples and tests have been broken by PR#108319:27
shogun-notifier-shogun: Sergey Lisitsyn :develop * 67690b8 / / (28 files): https://github.com/shogun-toolbox/shogun/commit/67690b8ab2da441c754c15f8a64c9f118a59ace219:27
shogun-notifier-shogun: Merge pull request #1084 from gsomix/feature/issue105819:27
shogun-notifier-shogun:19:27
shogun-notifier-shogun: Fixed examples and tests have been broken by PR#108319:28
shogun-buildbotbuild #800 of cyg1 - libshogun is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/800  blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com>19:34
shogun-buildbotbuild #1153 of deb1 - libshogun is complete: Success [build successful]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/115319:41
gsomixlisitsyn, thanks!19:42
shogun-buildbotbuild #801 of cyg1 - libshogun is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/801  blamelist: Evgeniy Andreev <gsomix@gmail.com>19:43
-!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has quit [Ping timeout: 245 seconds]19:44
-!- travis-ci [~travis-ci@ec2-23-20-77-42.compute-1.amazonaws.com] has joined #shogun19:45
travis-ci[travis-ci] it's Sergey Lisitsyn'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/708322519:45
-!- travis-ci [~travis-ci@ec2-23-20-77-42.compute-1.amazonaws.com] has left #shogun []19:45
shogun-buildbotbuild #1095 of deb3 - modular_interfaces is complete: Failure [failed compile java_modular]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/1095  blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com>20:06
shogun-buildbotbuild #1096 of deb3 - modular_interfaces is complete: Failure [failed compile java_modular]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/1096  blamelist: Evgeniy Andreev <gsomix@gmail.com>20:09
gsomixshogun-buildbot, hmm...20:13
gsomixmkay20:13
-!- Netsplit *.net <-> *.split quits: naywhayare, hushell, nube, abinash, lambday, @wiking, @sonney2k, shogun-notifier-, @lisitsyn20:14
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has quit [Quit: Leaving]20:20
-!- gsomix [~gsomix@109.169.146.48] has quit [Quit: Leaving]20:23
-!- gsomix [~gsomix@109.169.146.48] has joined #shogun20:24
-!- Netsplit over, joins: nube20:31
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun20:32
-!- naywhayare [~ryan@spoon.lugatgt.org] has joined #shogun20:32
-!- lisitsyn [~blackburn@188-122-243-31.clients.tlt.100megabit.ru] has joined #shogun20:32
-!- wiking [~wiking@huwico/staff/wiking] has joined #shogun20:32
-!- ServerMode/#shogun [+oo lisitsyn wiking] by pratchett.freenode.net20:32
-!- sonney2k [~shogun@7nn.de] has joined #shogun20:36
-!- ServerMode/#shogun [+o sonney2k] by pratchett.freenode.net20:36
-!- lisitsyn [~blackburn@188-122-243-31.clients.tlt.100megabit.ru] has quit [Quit: Leaving.]20:41
-!- Netsplit *.net <-> *.split quits: nube20:44
-!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has joined #shogun20:49
-!- Netsplit over, joins: nube20:49
-!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has quit [Client Quit]20:50
-!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has joined #shogun20:51
-!- lisitsyn [~blackburn@188-122-243-31.clients.tlt.100megabit.ru] has joined #shogun20:55
-!- mode/#shogun [+o lisitsyn] by ChanServ20:55
-!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has quit [Ping timeout: 245 seconds]20:57
-!- gsomix_ [~gsomix@109.169.150.253] has joined #shogun20:58
-!- gsomix [~gsomix@109.169.146.48] has quit [Ping timeout: 245 seconds]21:02
@sonney2kwiking, why is SFMT included in swig stuff?21:12
@sonney2kno wait it is just about doxygen21:13
-!- nube [~rho@49.244.40.224] has quit [Ping timeout: 276 seconds]21:20
-!- nube [~rho@49.244.107.88] has joined #shogun21:34
shogun-notifier-shogun: Soeren Sonnenburg :develop * 4ad0e6f / src/interfaces/modular/Features.i,src/shogun/labels/LabelsFactory.h: https://github.com/shogun-toolbox/shogun/commit/4ad0e6fe994719635b206bf08fb92cf91657d9d421:38
shogun-notifier-shogun: derive labelfactory from sgobject21:38
shogun-buildbotbuild #802 of cyg1 - libshogun is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/802  blamelist: Soeren Sonnenburg <sonne@debian.org>21:46
-!- nube [~rho@49.244.107.88] has quit [Ping timeout: 264 seconds]21:47
gsomix_sonney2k, thanks!21:56
-!- gsomix_ is now known as gsomix21:56
-!- travis-ci [~travis-ci@ec2-174-129-127-8.compute-1.amazonaws.com] has joined #shogun21:57
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/708574321:57
-!- travis-ci [~travis-ci@ec2-174-129-127-8.compute-1.amazonaws.com] has left #shogun []21:57
@lisitsynsonney2k: are you under eclipse now?21:59
@lisitsynI mean in general do you use eclipse now?22:00
-!- nube [~rho@49.244.111.155] has joined #shogun22:00
@lisitsynI just wanted to say clang_complete has made a few steps to cover most things22:00
@lisitsynIDE covers22:00
@sonney2klisitsyn, where do I get it?22:04
@sonney2kI miss vim keys...22:04
@lisitsynsonney2k: ok do you know what pathogen is?22:04
@sonney2kno22:08
@lisitsynsonney2k: https://github.com/tpope/vim-pathogen22:10
@lisitsynsonney2k: check installation22:11
@lisitsynonce you are done22:11
@lisitsynjust go to ~/.vim/bundle22:11
@lisitsynand git clone git://github.com/Rip-Rip/clang_complete.git22:12
@sonney2kand then?22:15
@lisitsynsonney2k: should work now22:15
@lisitsynsonney2k: just try to jump to any definition22:15
@lisitsynCtrl + ] here22:15
@sonney2ksays no tag found22:15
@lisitsynsonney2k: hmm okay22:16
@lisitsynsonney2k: what happens when you do22:16
@lisitsynsomeobject.22:16
@lisitsynI mean some real object22:16
@lisitsynoh22:16
@lisitsynsonney2k: did you add anything related to pathogen to your .vimrc?22:17
@lisitsynsonney2k: add 'execute pathogen#infect()' to the very beginning if you didn't22:18
@sonney2khmmh nothing22:21
@sonney2kthere must be mroe settings22:21
@sonney2kmore22:21
@lisitsynsonney2k: yeah a few more but I thought this should work already22:22
@lisitsynsonney2k: ahh one more22:22
@lisitsyndo you have libclang installed?22:23
@sonney2kof course not22:24
@lisitsynsonney2k:22:24
@lisitsyn 45 let g:clang_close_preview = 122:24
@lisitsyn 46 let g:clang_complete_auto = 122:24
@lisitsyn 47 let g:clang_auto_select = 022:24
@lisitsyn 48 let g:clang_snippets = 022:24
@lisitsyn 49 let g:clang_use_library = 122:24
@lisitsyn 50 let g:clang_debug = 022:24
@lisitsynfrom my vimrc22:24
@lisitsynbtw it is damn slow in tapkee due to a lot of headers22:25
@lisitsynand fast with shogun22:25
@lisitsynsonney2k: ohhh I think I know22:26
shogun-buildbotbuild #1097 of deb3 - modular_interfaces is complete: Failure [failed test python_modular]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/1097  blamelist: Soeren Sonnenburg <sonne@debian.org>22:26
@sonney2klisitsyn, do I need to set paths somehow?22:26
@lisitsynsonney2k: yes yes22:26
@sonney2kstill nothing22:26
@lisitsynsonney2k: I just got it jumps to my local install of shogun22:26
@lisitsynsonney2k: add .clang_complete file22:26
@lisitsynin the dir you run it22:26
@lisitsynshogun/src I guess22:26
@lisitsynthen add -I./shogun22:26
@sonney2klisitsyn, shouldn't we add this to shogun/src ?22:27
@lisitsynahh wait22:27
@lisitsyni am wrong22:27
@lisitsyn-I.22:27
@lisitsynthat's what should be in shogun/src22:27
@lisitsynsonney2k: will commit it in a minute22:28
@sonney2khmmhh doesn't work :/22:28
@sonney2kcan I somehow debug?22:28
@lisitsynsonney2k: any error or just nothing?22:28
shogun-notifier-shogun: Sergey Lisitsyn :develop * cd6de7c / src/.clang_complete: https://github.com/shogun-toolbox/shogun/commit/cd6de7c8f144b6e6b621887c96a88910fa02a48e22:29
shogun-notifier-shogun: Added clang complete file22:29
@sonney2kjust nothing22:29
@lisitsynsonney2k: Ctrl + ]?22:29
@sonney2kno idea what to do22:29
@sonney2kcstag: tag not found22:30
@lisitsynargh22:30
@lisitsynsonney2k: :scriptnames22:30
@lisitsynsonney2k: does it contain clang_complete?22:30
@sonney2k 29: ~/.vim/bundle/clang_complete/plugin/clang_complete.vim22:31
@sonney2kyes22:31
@lisitsynsonney2k: so you run your vim in shogun/src22:31
@lisitsynright?22:32
@lisitsynand you have that .clang_complete file I just pushed22:32
@lisitsynthat contains -I.22:32
@sonney2kyes22:35
@lisitsynsonney2k: and you pushed these settings from my vimrc22:36
@lisitsyntoo?22:36
@sonney2klisitsyn, yes22:36
@lisitsyndamn22:37
@lisitsynsonney2k: trying to find any reason22:40
shogun-buildbotbuild #803 of cyg1 - libshogun is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/803  blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com>22:42
@lisitsynsonney2k: so neither . completion nor jump work?22:43
@sonney2klisitsyn, exactly22:44
@sonney2kfinally an error22:47
@sonney2kLoading libclang failed, completion won't be available22:47
@sonney2kConsider setting g:clang_library_path22:47
@sonney2k-rw-r--r-- 1 root root 8.3M Mar 31 11:04 /usr/lib/libclang.so.122:48
@lisitsynsonney2k: ohh that's better22:48
@lisitsynat least we know an error22:48
@sonney2kI am installing libclang-dev22:48
@sonney2kyes22:49
@sonney2ksomething works22:49
-!- travis-ci [~travis-ci@ec2-23-20-59-228.compute-1.amazonaws.com] has joined #shogun22:49
travis-ci[travis-ci] it's Sergey Lisitsyn'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/708681822:49
-!- travis-ci [~travis-ci@ec2-23-20-59-228.compute-1.amazonaws.com] has left #shogun []22:49
@lisitsynsonney2k: nice22:50
@sonney2kI need ctrl+space though22:50
@sonney2kfor autocomplete22:50
@lisitsynsonney2k: ohh again22:50
@lisitsyn:D22:50
@sonney2kany idea?22:50
@lisitsynsonney2k: I did that once for you22:50
@lisitsynsonney2k: http://www.shogun-toolbox.org/irclogs/%23shogun.2012-05-21.log.html22:51
@lisitsynsonney2k: search by ctrl+space22:51
@lisitsynsonney2k: so patch clang_complete.vim22:52
@lisitsynthen add that imap <Nul> stuff22:52
@lisitsynto .vimrc22:52
@lisitsynsonney2k: line 147 now22:53
@lisitsynsonney2k: works here22:54
@sonney2kyeah works22:57
@lisitsynsonney2k: anything else?22:57
@lisitsyn:D22:57
@sonney2klisitsyn, yeah fix the build22:57
@lisitsynsonney2k: vim related?22:57
@lisitsyn:D22:57
@sonney2kconverter_factoranalysis_modular.py setting 1/1              ERROR22:57
@lisitsynstill?22:58
@lisitsynbah..22:58
@lisitsyndidn't know22:58
@lisitsynsonney2k: what is c?22:58
@lisitsynin output22:58
@lisitsyn-T +c22:58
@sonney2kevaluation_multiclassovrevaluation_modular.py setting 1/1    ERROR22:58
@sonney2k too hmmhh22:58
@sonney2klisitsyn, must have been some exception22:59
@lisitsynsonney2k: okay lets see if that happens again23:02
@lisitsynahh happened two times already23:02
@lisitsynsonney2k: could you check what does it return on your machine?23:03
* sonney2k compiling23:08
@lisitsynsonney2k: is that ok to return boolean at all?23:18
@lisitsynwell I do that in multidimensional scaling already23:18
@lisitsynI observe nothing random23:19
@lisitsynin both examples23:19
@sonney2ksomething is weird23:20
@sonney2kone returns a boolean23:20
@sonney2kone returns DenseFeatures23:20
@lisitsynsonney2k: what ex returns features?23:22
@sonney2klisitsyn, you changed the test23:23
@sonney2kreturn covdet > 023:23
@lisitsynbut didn't change data23:23
@lisitsynhahaha23:23
shogun-buildbotbuild #1098 of deb3 - modular_interfaces is complete: Failure [failed test python_modular]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/1098  blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com>23:25
shogun-notifier-shogun: Sergey Lisitsyn :develop * b491e83 / data: https://github.com/shogun-toolbox/shogun/commit/b491e83ab095a782dc18091a55e649fb6d280a9a23:29
shogun-notifier-shogun: Updated data submodule reference23:29
-!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has joined #shogun23:32
vgorbatilisitsyn: hi, here?23:33
@lisitsynvgorbati: hey23:33
@lisitsynyes23:33
vgorbatilisitsyn: have you seen my e-mail?23:33
@lisitsynyeah sure, didn't manage to write an answer yet though23:34
@lisitsynvgorbati: what do I think is that it is nice you pointed out a possibility to speed it up23:34
@lisitsynbut I am unsure it is worth the deal23:34
shogun-buildbotbuild #804 of cyg1 - libshogun is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/804  blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com>23:34
vgorbatilisitsyn: so, implement without it for now?23:35
@lisitsynvgorbati: yeah lets just do simplest23:35
vgorbatilisitsyn: ok, less work for me:) But the resulting method will be quite slow I think)23:35
@lisitsynvgorbati: oh well all DR methods are quite slow23:36
vgorbatilisitsyn: just curious: does the idea I supposed at least make some sense?23:36
@lisitsynvgorbati: I didn't find any error so probably yes23:38
@lisitsynbut again lets go simplest23:38
vgorbatilisitsyn: ok, maybe I will convince you later:)23:39
@lisitsynvgorbati: there is a bunch of more useful methods :)23:39
vgorbatilisitsyn: well, at least wikipedia says that manifold sculpting gives better results for some cases23:40
@lisitsynvgorbati: ha! each method has a dataset it works the best among others23:41
@lisitsynvgorbati: LTSA is the ideal algorithm for swissroll23:42
vgorbatilisitsyn: :) PCA is good for a straight lines I guess. Unfortunately, real datasets are not like this)23:42
@lisitsynheh but PCA is relevant still23:43
-!- pickle27 [~kevin@rcv3-lab-pc.ee.queensu.ca] has joined #shogun23:43
vgorbatilisitsyn: I guess it is like a first-try-method - if it succeeds, then there's no need to invent bycicles)23:44
@lisitsynvgorbati: you mean PCA?23:45
vgorbatilisitsyn: yes23:45
shogun-notifier-shogun: Soeren Sonnenburg :develop * 42ca7de / data,tests/integration/python_modular/tester.py: https://github.com/shogun-toolbox/shogun/commit/42ca7de15c98c26cf75ad10e5b56109292058d6923:45
shogun-notifier-shogun: fix tester and require new mcroceval data23:45
@lisitsynyeah some kind of - you just drop some features and project others23:45
@sonney2kvgorbati, usually the simplest thing is just enough in practise23:46
@lisitsynvgorbati: and it is damn fast23:46
@lisitsynand stable23:46
@sonney2konly scientists want to tell you that it is not23:46
@sonney2kbut they have to publish papers23:46
@lisitsynhaha23:46
@lisitsynsonney2k: I just wonder23:46
@sonney2kand a paper on PCA yet again is well *yawn*23:46
@lisitsynhad you been thinking so finishing your phd?23:47
@lisitsynand more important when you started working on it?23:47
gsomixgood night, guys23:48
@lisitsynnite23:48
@sonney2klisitsyn, of course not23:53
@sonney2kbut when you grow old you learn how papers are made23:53
@sonney2kwhat gets you in a conference / a paper accepted and what not23:53
@sonney2kthe simple good working ideas never made23:53
@sonney2kit23:53
@lisitsynsonney2k: do you realize you get us old :D23:53
@sonney2klook at neural nets23:54
@sonney2kno one could publish a NN paper after SVMs / kernel machines hyped23:55
@sonney2k'they have local minima'23:55
@sonney2kand perceptron cannot even solve x-or23:55
@sonney2kbla bla23:55
@lisitsynsonney2k: haha23:55
@sonney2kI guess it is the same with SVMs nowadays23:56
-!- travis-ci [~travis-ci@ec2-23-20-64-221.compute-1.amazonaws.com] has joined #shogun23:57
travis-ci[travis-ci] it's Sergey Lisitsyn'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/708780123:57
-!- travis-ci [~travis-ci@ec2-23-20-64-221.compute-1.amazonaws.com] has left #shogun []23:57
@sonney2kmy lesson learned is that any method is as good as the person using it23:57
* sonney2k argh. managed to attract a more than reasonable number of mosquitoes23:58
@lisitsynhehe23:58
@lisitsynsonney2k: what's with your garden?23:58
@lisitsynsonney2k: conquered it?23:58
@sonney2khttps://github.com/shogun-toolbox/shogun/pull/1085/files23:59
--- Log closed Sun May 12 00:00:41 2013