Open in new window / Try shogun cloud
--- Log opened Wed Apr 17 00:00:28 2013
gsomixsometimes I think I'm not compatible with git at DNA level...00:32
-!- k_matsuda [~k_matsuda@h101-111-222-195.catv02.itscom.jp] has joined #shogun00:35
gsomixnew git workflow. got it.00:42
gsomixthen we need to go deeper...00:42
gsomixnite01:15
-!- votjak [~chatzilla@host-46-241-3-209.bbcustomer.zsttk.net] has joined #shogun02:42
-!- Yanglittle [deb20afc@gateway/web/freenode/ip.222.178.10.252] has joined #shogun03:03
Yanglittleis there any body?03:05
-!- Yanglittle [deb20afc@gateway/web/freenode/ip.222.178.10.252] has quit [Ping timeout: 245 seconds]03:53
-!- k_matsuda [~k_matsuda@h101-111-222-195.catv02.itscom.jp] has quit [Remote host closed the connection]04:04
shogun-buildbotbuild #364 of nightly_default is complete: Success [build successful]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/nightly_default/builds/36404:24
@sonney2klisitsyn, http://blog.mashape.com/post/48074869493/list-of-machine-learning-apis05:04
-!- dddddddddddd [ddd@tu073053.ip.tsinghua.edu.cn] has joined #shogun05:17
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun05:30
shogun-notifier-shogun: iglesias :develop * 5e9e870 / .gitignore: https://github.com/shogun-toolbox/shogun/commit/5e9e870001ca47d9814ac200cb5cc476d238680305:30
shogun-notifier-shogun: Update gitignore with automatically created file.05:30
shogun-notifier-shogun: Add interfaces/python_modular/abstract_type_extension.i to gitignore.05:30
shogun-notifier-shogun: Soeren Sonnenburg :develop * 57d9409 / .gitignore: https://github.com/shogun-toolbox/shogun/commit/57d9409aa6ffa35a78b120c842f9a986b5331ac305:30
shogun-notifier-shogun: Merge pull request #977 from iglesias/develop05:30
shogun-notifier-shogun:05:30
shogun-notifier-shogun: Update gitignore with automatically created file.05:30
shogun-notifier-shogun: Soeren Sonnenburg :develop * ee1eef7 / .travis.yml: https://github.com/shogun-toolbox/shogun/commit/ee1eef740c08b4548141ff6b25e8496c9ca28fe205:35
shogun-notifier-shogun: add python static to tests05:35
-!- travis-ci [~travis-ci@ec2-50-17-148-215.compute-1.amazonaws.com] has joined #shogun05:43
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/640327205:43
-!- travis-ci [~travis-ci@ec2-50-17-148-215.compute-1.amazonaws.com] has left #shogun []05:43
-!- travis-ci [~travis-ci@ec2-50-17-148-215.compute-1.amazonaws.com] has joined #shogun05:52
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/640341505:52
-!- travis-ci [~travis-ci@ec2-50-17-148-215.compute-1.amazonaws.com] has left #shogun []05:52
-!- asdf__ [6f5d05c6@gateway/web/freenode/ip.111.93.5.198] has joined #shogun06:56
-!- asdf__ [6f5d05c6@gateway/web/freenode/ip.111.93.5.198] has quit [Client Quit]06:59
-!- hoijui [~hoijui@dslb-088-074-124-102.pools.arcor-ip.net] has joined #shogun07:28
-!- hoijui [~hoijui@dslb-088-074-124-102.pools.arcor-ip.net] has quit [Ping timeout: 256 seconds]07:41
sonne|workdddddddddddd: ccccccccccccccccccc ?07:58
-!- erlenda [~erlenda@fw-oslo.intelcom.no] has joined #shogun08:33
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout]08:35
-!- hoijui [~hoijui@wlan-141-23-66-81.tubit.tu-berlin.de] has joined #shogun08:37
-!- sumit [73f91219@gateway/web/freenode/ip.115.249.18.25] has joined #shogun08:44
sumitVojtak: Sir , did you go through the mail I had sent you?08:45
-!- randomname [0e8b5206@gateway/web/freenode/ip.14.139.82.6] has joined #shogun08:46
randomnamehi08:47
randomnameI am trying to run documented examples from python_modular08:48
randomnameI am getting error in line "from shogun.Features import RealFeatures, MulticlassLabels"08:49
randomnameI did ./configure --interfaces=python_modular,python_static08:49
randomnameand then make08:49
randomnamewasn't it enough? Or am i missing something?08:49
randomnameSimilarly in python_static : The same line08:50
randomname" from sg import sg"08:50
randomnameAnyone there?08:51
randomnameshogun-toolbox: there?08:52
votjakrandomname: Do you run 'make install' after 'make'?08:57
randomnameoh..08:57
randomnamewill try it just now08:57
randomnametried .. but didn't work08:59
randomnameThere were no errors on make install08:59
randomnamedo you want me to put the output of make install?09:00
-!- sumit [73f91219@gateway/web/freenode/ip.115.249.18.25] has quit [Ping timeout: 245 seconds]09:04
-!- dddddddddddd [ddd@tu073053.ip.tsinghua.edu.cn] has left #shogun []09:07
-!- dddddddddddd [ddd@tu073053.ip.tsinghua.edu.cn] has joined #shogun09:08
-!- dddddddddddd [ddd@tu073053.ip.tsinghua.edu.cn] has left #shogun []09:08
randomnameIs there any sort of linking that needs to be done?09:10
randomnameAm running "python <filename>.py"09:10
-!- madcc [madcc@tu073053.ip.tsinghua.edu.cn] has joined #shogun09:11
sonne|workrandomname: no. might be that you need to run ldconfig as root?09:12
randomnameldconfig in?09:13
randomnameI hope my os Fedora 17 is not the issue, as in README it says "Debian GNU/Linux" under platforms?09:22
-!- hoijui [~hoijui@wlan-141-23-66-81.tubit.tu-berlin.de] has quit [Ping timeout: 255 seconds]09:28
randomnameanyone?09:30
-!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has joined #shogun09:30
randomnamesonne|work: I did run make install as "sudo make install"09:30
randomnameso ldconfig must have also ran with root permissions09:30
-!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has quit [Ping timeout: 245 seconds]09:38
-!- lisitsyn [~blackburn@83.234.54.56] has quit [Quit: Leaving.]09:39
randomnamevotjak: ?09:41
-!- iglesiasg [~nando@n132-p169.kthopen.kth.se] has joined #shogun09:43
-!- mode/#shogun [+o iglesiasg] by ChanServ09:43
-!- madcc [madcc@tu073053.ip.tsinghua.edu.cn] has quit []09:44
-!- randomname [0e8b5206@gateway/web/freenode/ip.14.139.82.6] has quit [Ping timeout: 245 seconds]09:52
-!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has joined #shogun10:02
-!- hoijui [~hoijui@wlan-141-23-66-81.tubit.tu-berlin.de] has joined #shogun10:03
deerishiin the kernel pca algorithm in python modular, i am trying to plot the data after implementing kernel pca.just one doubt ,the preprocessor.apply_to_feature_matrix(features) returns the new linearly separable data?rite/10:07
deerishi?10:07
deerishi?10:08
deerishi?10:10
-!- foulwall [~foulwall@li379-21.members.linode.com] has joined #shogun10:11
foulwallHello guys10:11
deerishi@sonney2k any ideas?10:11
deerishi?10:11
-!- madcc [madcc@tu073053.ip.tsinghua.edu.cn] has joined #shogun10:12
-!- foulwall [~foulwall@li379-21.members.linode.com] has quit [Ping timeout: 240 seconds]10:18
-!- lisitsyn [~lisitsin@mxs.kg.ru] has joined #shogun10:20
-!- mode/#shogun [+o lisitsyn] by ChanServ10:20
@iglesiasgdeerishi: is kernel pca a preprocessor? I thought it was a converter10:22
@lisitsyniglesiasg: preprocessor still10:22
@iglesiasgaham I see10:23
@iglesiasglisitsyn: isn't included in tapkee then?10:23
-!- sonne|work [~sonnenbu@sams-office-nat.tomtomgroup.com] has quit [Quit: Leaving.]10:23
deerishii did preprocessor=KernelPCA(kernel) ,now m trying to plot the data10:23
@lisitsyniglesiasg: it is -but currently it uses old code10:23
-!- sonne|work [~sonnenbu@sams-office-nat.tomtomgroup.com] has joined #shogun10:23
@iglesiasglisitsyn: understood10:24
deerishie.g.  the input data is 2x302 , but m getting a data of  1x302 after kernelpca10:24
@iglesiasgdeerishi: maybe you should transpose the data? no real idea, just wild guess10:25
deerishiwilll just try that10:25
@iglesiasgdeerishi: or wait, it sort of makes sense10:25
@iglesiasgdeerishi: you have 302 points in 2D right? two points in R^302 sounds weird - too few points10:26
deerishii have 2 not separable datasets ,i.e total of 302 points in 2d10:27
deerishithe data is 2 semicircular arcs10:29
deerishiin  the same quadrants10:29
@iglesiasgdeerishi: why are you talking about separable datasets?10:30
-!- sonne|work [~sonnenbu@sams-office-nat.tomtomgroup.com] has quit [Remote host closed the connection]10:30
@lisitsynkpca has nothing to do with separability10:30
-!- sonne|work [~sonnenbu@sams-office-nat.tomtomgroup.com] has joined #shogun10:32
deerishibut one should be able to distinguish 2 sets os data,after kernel pca,rite?10:32
@lisitsynah yes10:32
@iglesiasgdeerishi: one should be, but the method is not going to tell you about that10:32
deerishik.I am trying to solve the bug about kernelpca test/check .10:33
deerishii should be able to see after plotting that kernel pca works,10:34
deerishithe dimension of the data i am getting is just a single row,how can i plot that and see if the algorithim has worked or not?10:35
@lisitsynwell plot points on a line10:36
@iglesiasgif your intuition is correct, then you should see two parts of the line populated with points10:37
-!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has quit [Ping timeout: 245 seconds]10:54
-!- madcc [madcc@tu073053.ip.tsinghua.edu.cn] has quit []10:54
-!- hoijui [~hoijui@wlan-141-23-66-81.tubit.tu-berlin.de] has quit [Ping timeout: 245 seconds]10:55
-!- HeikoS [~heiko@nat-178-28.internal.eduroam.ucl.ac.uk] has joined #shogun10:57
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun11:03
shogun-notifier-shogun: lambday :develop * 72a036a / src/shogun/mathematics/ (4 files): https://github.com/shogun-toolbox/shogun/commit/72a036aadff7e50028a2245aac79c05667029eb811:03
shogun-notifier-shogun: added EigenSparseUtil class11:03
shogun-notifier-shogun: lambday :develop * f66ba35 / src/shogun/mathematics/Statistics.h: https://github.com/shogun-toolbox/shogun/commit/f66ba35331d597cc044cd6f079dfee3f7bf2d97711:03
shogun-notifier-shogun: added EigenSparseUtil class11:03
shogun-notifier-shogun: lambday :develop * 6312941 / src/shogun/mathematics/ (4 files): https://github.com/shogun-toolbox/shogun/commit/6312941d6c3b94e02f04a470f2c87e05e4af9f8511:03
shogun-notifier-shogun: Merge branch 'feature/eigen_sparse_util' into develop11:03
shogun-notifier-shogun: Heiko Strathmann :develop * d385b95 / src/shogun/mathematics/ (4 files): https://github.com/shogun-toolbox/shogun/commit/d385b95b8b83c37180454704fe0b02328c97de8211:03
shogun-notifier-shogun: Merge pull request #982 from lambday/develop11:03
shogun-notifier-shogun:11:03
shogun-notifier-shogun: Added EigenSparseUtil class for ease of coverting SGSparseMatrix to Eigen3 SparseMatrix11:03
-!- sumit [73f91219@gateway/web/freenode/ip.115.249.18.25] has joined #shogun11:07
sonne|workHeikoS: any comments about the workshop announcement mail?11:10
sonne|workIf not I will send it out this evening.11:10
HeikoSsonne|work: ah sorry, haven't looked yet, will do now11:10
@lisitsynHeikoS: I am afraid you speak about PCA11:11
@lisitsynnot kPCA11:11
@lisitsynHeikoS: there is no such thing like projection matrix in kpca11:12
HeikoSlisitsyn: what does the projection matrix give you then?11:12
HeikoSor transformation matrix11:13
HeikoSthis is a matrix that the data is projected on11:13
HeikoSat least this is what the apply method does11:13
HeikoSnot correct?11:13
@lisitsynHeikoS: yes, already projected data11:13
HeikoSlisitsyn: so the get_transformation matrix method returns the projected data?11:14
HeikoSwhat does apply_to_features do?11:14
HeikoSif I look at the code this looks like a projection11:14
@lisitsynlet me check11:14
@lisitsynHeikoS: ehmm11:15
-!- sumit [73f91219@gateway/web/freenode/ip.115.249.18.25] has quit [Ping timeout: 245 seconds]11:16
@lisitsynHeikoS: it is something strange11:16
HeikoSlisitsyn: haha :)11:16
HeikoSgood that we started investigating :)11:16
@lisitsynHeikoS: well may be I misunderstand something11:16
@lisitsynbut kpca is not a linear method11:17
@lisitsynin my head11:17
HeikoSit is in the feature space right?11:17
@lisitsynHeikoS: top eigenvectors of centered kernel matrix11:17
HeikoSyes11:17
@lisitsynthat is the projection11:17
@lisitsynnot a projection matrix to be multiplied11:17
HeikoSso the projection is11:20
@lisitsynHeikoS: ah it seems everything is ok here11:21
HeikoSlisitsyn: so whats the difference between apply and get_transform?11:22
@lisitsynbut my understanding of kpca was not complete11:22
-!- travis-ci [~travis-ci@ec2-54-242-66-169.compute-1.amazonaws.com] has joined #shogun11:22
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/640897011:22
-!- travis-ci [~travis-ci@ec2-54-242-66-169.compute-1.amazonaws.com] has left #shogun []11:22
@lisitsynHeikoS: get_transformation_matrix?11:24
@lisitsynHeikoS: it is eigenvectors before normalization11:24
gsomixhi11:26
@lisitsynHeikoS: err not normalization11:26
@lisitsynmultiplication with kernel values11:26
HeikoSlisitsyn: so in order to plot the projected input data, the apply method is the right one right?11:27
-!- abinashpanda [75ef5e6e@gateway/web/freenode/ip.117.239.94.110] has joined #shogun11:27
HeikoSso, the question is, do I have to correct the email? :)11:27
HeikoSgsomix: hi!11:28
HeikoSlisitsyn: very cool that travis has python now, once with works, I will be very happy :)11:28
gsomixHeikoS, about my pull request. it's seems I'm careless, but at night I had problem with compilation. need check again.11:30
@lisitsynHeikoS: yes it seems so11:30
HeikoSgsomix: ok11:30
-!- hoijui [~hoijui@wlan-141-23-66-81.tubit.tu-berlin.de] has joined #shogun11:30
@lisitsynthe careless whisper of a good friend11:30
@lisitsynto the heard and mind11:30
@lisitsynla la la11:30
@lisitsyn:D11:30
HeikoSlisitsyn: ehm?11:30
HeikoS:)11:30
@lisitsynHeikoS: there is a song11:31
HeikoSI see  :)11:31
@lisitsynHeikoS: george michael! you must be not listening to gay music11:31
@lisitsyn:D11:31
-!- abinashpanda [75ef5e6e@gateway/web/freenode/ip.117.239.94.110] has quit [Ping timeout: 245 seconds]11:31
gsomixlisitsyn, too gay11:31
@lisitsynHeikoS: gsomix told he is careless that's why I recalled it11:31
HeikoShaha ;)11:32
wikingyo11:33
gsomixwiking, \(^_^)/11:33
wikinggsomix: heeheey11:34
wikingltnc11:34
sonne|workHeikoS: travis - once with works?11:38
sonne|workHeikoS: what do you mean?11:38
wiking./check.sh: line 21: gdb: command not found11:38
wikingdo we need gdb for pythom modular examples11:38
wiking?11:38
sonne|workwiking: so just add gdb11:38
sonne|workyes11:38
wikingok11:38
sonne|workwhen something crashes11:38
wikingas well as for python_static i guess?11:38
sonne|workit will runt he thing through gdb and show the backtrace11:38
sonne|workyes even standard test11:39
sonne|works11:39
sonne|works/tests/examples/11:39
wikingah ok11:39
wikingso even for libshogun examples?11:39
sonne|workyes11:39
wikingokey11:39
wikingone11:40
wiking*done11:40
shogun-notifier-shogun: Viktor Gal :develop * f6a16c2 / .travis.yml: https://github.com/shogun-toolbox/shogun/commit/f6a16c2f486e16efc733ec2fb31b806a25c371e911:40
shogun-notifier-shogun: Add gdb package for debugging errors in examples11:40
sonne|workwiking: btw could we in principle build for all architectures on travis?11:43
sonne|workor would that kill travis?11:43
wikingsonne|work: you mean x-compile?11:43
wikinglike for arm?11:43
sonne|workerr all interfaces I meant11:43
wikingmmm yes11:43
wikingi was meaning to ask11:43
wikingif anybody knows the exact package dependency for the other interfaces11:43
wikingplease give them to me11:43
sonne|workit might make sense now11:43
wikingand i'll add those to travis11:44
sonne|workwiking: just look at shogun's debian package11:44
wikingso for r_*, lua_*, csharp_* etc11:44
wikingok11:44
sonne|workahh true11:44
sonne|workit got removed11:44
sonne|workI should put creating debian package high on my prio list again11:44
wikingmmm i dont have it on ubuntu11:45
wikinghttp://packages.debian.org/squeeze/shogun-r11:46
wikingthis one is r_modular?11:46
-!- randomname [0e8b5206@gateway/web/freenode/ip.14.139.82.6] has joined #shogun11:47
sonne|workwiking: http://bollin.googlecode.com/svn/shogun/trunk/debian/control11:47
sonne|worklook here11:47
sonne|workthat's the stuff I used for shogun 2.011:47
wikingmmm11:52
wikingsonne|work: octave dependency?11:52
gsomixHeikoS, mhm, my pull request doesn't make sense11:52
sonne|workwiking: I guess  liboctave-dev, octave11:54
wikingscipy!11:55
wikingshit that we miss :)11:55
-!- travis-ci [~travis-ci@ec2-50-17-148-215.compute-1.amazonaws.com] has joined #shogun11:59
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/640974911:59
-!- travis-ci [~travis-ci@ec2-50-17-148-215.compute-1.amazonaws.com] has left #shogun []11:59
shogun-notifier-shogun: Viktor Gal :develop * 408f5ae / .travis.yml: https://github.com/shogun-toolbox/shogun/commit/408f5aeff6a95b5354872772da4c606caf12a8e411:59
shogun-notifier-shogun: Add r,java,rubycsharp,lua modular interfaces to travis11:59
shogun-notifier-shogun: Viktor Gal :develop * ef72d49 / .travis.yml: https://github.com/shogun-toolbox/shogun/commit/ef72d49486caeb26723e37148f03512e782495c811:59
shogun-notifier-shogun: Add python-scipy package to python_modular on travis11:59
wikingthis should fix python_modular...12:00
wikinglol12:01
wikingpoor travis we are killing it :D12:01
wikinghttps://travis-ci.org/shogun-toolbox/shogun12:01
sonne|workwiking: do we really need to build for clang & gcc for everything?12:05
sonne|workI mean sure it is nice but hmhhh12:05
sonne|workpoor travis12:05
sonne|worknow consider we would add arm / ppc / ...12:05
sonne|workand octave modular needs 4GB memory to compile12:06
@iglesiasgpeople from travis will blacklist us soon :D12:06
wikingmmm i dont know what's the VPS limit for a travis build12:06
wikingbut i suppose it won't have 4gb :(12:06
wikingsonne|work: we can change the build matrix12:06
-!- hoijui [~hoijui@wlan-141-23-66-81.tubit.tu-berlin.de] has quit [Ping timeout: 245 seconds]12:06
wikingthat the modular interfaces are only built with one compiler12:07
-!- hoijui [~hoijui@wlan-141-23-66-81.tubit.tu-berlin.de] has joined #shogun12:08
-!- votjak [~chatzilla@host-46-241-3-209.bbcustomer.zsttk.net] has quit [Quit: ChatZilla 0.9.90 [Firefox 20.0.1/20130409194949]]12:08
wikingbtw: is it normal that serialization_complex_example.py setting 1/2 fails?12:20
wikingin python_modular?12:20
sonne|workno12:20
wikingok it dies on travis12:21
wikingsonne|work: can u plz take a look : https://travis-ci.org/shogun-toolbox/shogun/jobs/641010212:22
wiking?12:22
sonne|workwiking: no idea how to debug that12:24
sonne|workstuff that fails on travis but works elsewhere is very hard to debug12:25
wikingindeed12:25
wikingmmm for some reason it cannot find (configure script) ruby-narray.... mmm maybe i need a -dev package of it?12:26
wikingmmm there's no such thing12:26
shogun-notifier-shogun: Viktor Gal :develop * 88a18c7 / .travis.yml: https://github.com/shogun-toolbox/shogun/commit/88a18c7b25330a375bab34ab4a7becc76c566f7012:48
shogun-notifier-shogun: Travis: add missing package for java and csharp12:48
-!- hoijui [~hoijui@wlan-141-23-66-81.tubit.tu-berlin.de] has quit [Ping timeout: 252 seconds]12:50
wikingsonne|work:13:07
wikingError in library(shogun) : 'shogun' is not a valid installed package13:07
wikingit's with ruby_modular13:07
wikingno sorry13:07
wikingr_modular13:07
wikingshould i give some extra path or something?13:07
wikingwhat are those .m files in tests/integration/python_modular/test_all.sh ?13:16
-!- k_matsuda [~k_matsuda@h101-111-222-195.catv02.itscom.jp] has joined #shogun13:21
-!- travis-ci [~travis-ci@ec2-54-242-66-169.compute-1.amazonaws.com] has joined #shogun13:28
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/641095013:28
-!- travis-ci [~travis-ci@ec2-54-242-66-169.compute-1.amazonaws.com] has left #shogun []13:28
wiking:(13:30
-!- iglesiasg [~nando@n132-p169.kthopen.kth.se] has quit [Read error: Operation timed out]13:34
-!- iglesiasg [~nando@n132-p169.kthopen.kth.se] has joined #shogun13:51
-!- mode/#shogun [+o iglesiasg] by ChanServ13:51
-!- iglesiasg [~nando@n132-p169.kthopen.kth.se] has quit [Quit: Reconnecting]14:29
-!- iglesiasg [~nando@n132-p169.kthopen.kth.se] has joined #shogun14:29
-!- mode/#shogun [+o iglesiasg] by ChanServ14:29
-!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has joined #shogun14:41
randomnamehi14:41
@iglesiasghey randomname14:42
randomnameIs there any class similar to SGVector for priority_queue or heap ?14:42
randomnameI mean something similar to SGHeap14:42
@iglesiasgrandomname: IIRC there is some fibonacci head somewhere14:42
@iglesiasgfibonacci heap, not head hehe14:43
@lisitsynI removed that as it was not needed14:43
@lisitsynrandomname: what for?14:43
@iglesiasgrandomname: anyway, if you just need to use it in your cpp (not in header) feel free to use STL14:43
randomnamewe can use STL ?14:44
randomnamethats great :)14:44
deerishihttps://github.com/deerishi/KernelPCA-test-check/commit/80b334747ebd176d3e8602bacbff996dbcce54cf#commitcomment-302946514:44
@iglesiasgrandomname: but only in implementation files please, not headers14:44
deerishihave posted the kernelpca test.check14:44
randomnamelisitsyn: I need for a patch on Ball Tree for KNN14:44
deerishican anyone see the result and get back to me14:44
deerishii have also posted the plots of the results14:45
@iglesiasgdeerishi: the second plot is the density of the points after kpca I guess14:45
@iglesiasgdeerishi: or?14:45
deerishiyes14:45
@iglesiasgso no actual coordinates right?14:45
@iglesiasgI think it looks fine14:46
@iglesiasgI wonder why one of them is in 50 and the other in 200 though, since the original are of radii 50 and 10014:46
@iglesiasgbut I guess it is fine14:46
deerishii just took the vector returned after apply_to _feature_matrices and plotted it14:46
@lisitsynrandomname: yes, use stl priority heap in .cpp14:46
randomnameI have another question: I should be making another class BallTreeNodeData.h or just a simple struct def in the code will do ?14:47
randomnamelisitsyn: thanks :)14:47
deerishi@iglesiasg ,so did i do as per required in the task?14:47
deerishi?14:47
@lisitsynrandomname: that's ok to have struct here I believe14:48
randomnamelisitsyn: thanks :) that helped ... I was on my way to make separate class files for Ball Tree Node :)14:49
@iglesiasgdeerishi: do you think the second plot with the results makes sense?14:49
randomnamebut now I can do it with just a struct :)14:49
-!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has quit [Ping timeout: 245 seconds]14:52
-!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has joined #shogun14:52
deerishi@iglesiasg : no but i am having doubt in what data to plot to check if kernel pca has worked or not14:52
deerishiany suggestions on how to improve14:53
@iglesiasgdeerishi: you can draw inspiration from http://scikit-learn.org/dev/auto_examples/decomposition/plot_kernel_pca.html14:53
@lisitsynit actually makes sense to make an homomorphous mapping here 2d->2d14:54
@lisitsynas it is implicitly \inf d -> 2d14:54
@lisitsynin case of gaussian kernel14:55
@iglesiasgwhat is homomorphous in this context?14:55
@iglesiasgsimply that it doesn't reduce dim?14:55
@lisitsyniglesiasg: X -> X14:55
@iglesiasgok14:55
@lisitsyndoesn't change the space14:55
@lisitsyniglesiasg: as we are actually mapping rkhs of X -> X14:56
@lisitsynI think that's ok14:56
@iglesiasgrkhs?14:56
@lisitsynreproducing kernel hilbert space14:56
@lisitsyniglesiasg: see we have some youth!14:59
@lisitsyn:)15:00
-!- deerishi_ [73f88294@gateway/web/freenode/ip.115.248.130.148] has joined #shogun15:00
@iglesiasglisitsyn: hehe yeah, I like the Lincoln's speeches stuff15:00
@lisitsyniglesiasg: born 199415:01
@lisitsyniglesiasg: younger than me for 3 years15:01
-!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has quit [Ping timeout: 245 seconds]15:02
@iglesiasgiglesiasg: oh I thought you were 90!15:02
@lisitsyniglesiasg: no I was born 1991 :)15:02
-!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has joined #shogun15:02
@iglesiasghaha look I have come back to talk to myself :D15:02
@lisitsynlisitsyn: that's ok15:03
@lisitsyniglesiasg: ^ :D15:03
deerishi@iglesiasg : i have a doubthttp://www.shogun-toolbox.org/doc/en/1.0.0/KernelPCA_8cpp_source.html#l00133 .according to this the matrix returned by the function should be the matrix after kpca.(that is what i had plotted).15:03
@iglesiasg:)15:03
deerishi?15:03
deerishior am i interpreting it wrong15:03
@iglesiasgdeerishi: what's the question? :)15:04
-!- deerishi_ [73f88294@gateway/web/freenode/ip.115.248.130.148] has quit [Ping timeout: 245 seconds]15:04
deerishii want to see the data after kpca15:04
deerishiand the above function returns that matrix15:04
deerishirite?15:04
@iglesiasgCSimpleFeatures, that code is outdated I am afraid15:06
@iglesiasgdeerishi: apply to feature matrix returns the features after kPCA has been applied15:08
@iglesiasgcheck doc in the current code15:08
@iglesiasghttps://github.com/shogun-toolbox/shogun/blob/master/src/shogun/preprocessor/KernelPCA.h15:08
-!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has quit [Ping timeout: 245 seconds]15:11
-!- hoijui [~hoijui@dslb-088-074-124-102.pools.arcor-ip.net] has joined #shogun15:22
-!- VladyslavGorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has joined #shogun15:26
-!- randomname [0e8b5206@gateway/web/freenode/ip.14.139.82.6] has quit [Ping timeout: 245 seconds]15:31
-!- hoijui [~hoijui@dslb-088-074-124-102.pools.arcor-ip.net] has quit [Read error: Operation timed out]15:37
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout]15:48
wikingblah15:49
wikinglisitsyn: here?15:49
wikinglisitsyn: any ideas what could be the problem in these 2 cases: https://travis-ci.org/shogun-toolbox/shogun/jobs/6410958 https://travis-ci.org/shogun-toolbox/shogun/jobs/641096215:50
@lisitsynwiking: no idea bout R15:56
-!- gsomix_ [~gsomix@37.61.179.101] has joined #shogun15:57
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun16:01
shogun-notifier-shogun: Viktor Gal :develop * 408f4b6 / .travis.yml: https://github.com/shogun-toolbox/shogun/commit/408f4b6aadd91c3b0c8da05e553a8774d01659db16:01
shogun-notifier-shogun: Travis: compile the interfaces only with clang16:01
-!- gsomix [~gsomix@83.234.54.56] has quit [Ping timeout: 245 seconds]16:01
gsomix_wiking, uh, oh, R interface. at my local machine I can't even compile it.16:01
-!- gsomix_ is now known as gsomix16:01
wikingwhat about ruby? :)16:01
wikingok this commit didn't work as i expected it16:02
-!- iglesiasg [~nando@n132-p169.kthopen.kth.se] has quit [Read error: Operation timed out]16:02
-!- gsomix [~gsomix@37.61.179.101] has left #shogun []16:02
-!- gsomix [~gsomix@37.61.179.101] has joined #shogun16:02
gsomixwiking, have not tried16:03
shogun-notifier-shogun: Viktor Gal :develop * 9690a0c / .travis.yml: https://github.com/shogun-toolbox/shogun/commit/9690a0cbf6c439ef69171a46003adc408808301c16:05
shogun-notifier-shogun: Travis: matrix exclusion only works with exact match16:05
-!- lisitsyn [~lisitsin@mxs.kg.ru] has quit [Quit: Leaving.]16:06
wikingok this works... ;)16:06
gsomixwiking, hm, it's all ok with ruby an my machine. just intstalled ruby-narray16:09
-!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has joined #shogun16:11
-!- Yanglittle [deb20afe@gateway/web/freenode/ip.222.178.10.254] has joined #shogun16:12
-!- VladyslavGorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has quit [Ping timeout: 245 seconds]16:15
-!- erlenda [~erlenda@fw-oslo.intelcom.no] has quit [Remote host closed the connection]16:19
wikinggsomix: yeah that's weird16:20
wikingi do the same16:20
wikingstill the configure script dies with not finding ruby-narray16:21
-!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has quit [Ping timeout: 245 seconds]16:32
YanglittleHey, how can i do if i have the three kernel matrix of three modular between traindat and traindat  in mkl, I read the examples, they're all the same traindat.16:40
-!- wiking [~wiking@info2k1.hu] has quit [Changing host]16:41
-!- wiking [~wiking@huwico/staff/wiking] has joined #shogun16:41
-!- travis-ci [~travis-ci@ec2-54-242-66-169.compute-1.amazonaws.com] has joined #shogun16:46
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/641532816:46
-!- travis-ci [~travis-ci@ec2-54-242-66-169.compute-1.amazonaws.com] has left #shogun []16:46
wikingHeikoS: ping16:46
HeikoSwiking: pong, hi!16:50
HeikoSsonne|work, I meant the python for travis16:50
wikingHeikoS: what's with those integration scripts? .m ?16:55
wikingHeikoS: one is failing for some reason on travis16:55
HeikoSehm16:55
HeikoSdoes it fail on the buildbot?16:55
HeikoSno idea16:55
wikingmmm lemme check16:55
wikingbtw have u ever used ruby_modular?16:58
wikingit seems it does not fail on buildbot :D16:58
wikingclassified: 1.717182e-01, bias: 0.000000e+00 <--- accuracy: 1.000000e-0416:59
HeikoSweird16:59
wikingin /SubGradientSVM_30_1en05_16_05_False.m16:59
HeikoSmaybe some library not available?16:59
wikingeverything else works fine16:59
wikingonly this one fails16:59
wikinge.g. SubGradientSVM_30_1en05_1_05_False.m17:00
wikingaaaah17:00
wikingwhy might SubGradientSVM_30_1en05_1_05_False.m be skipped?17:00
wikingbecause of some library missing?17:00
HeikoSno I mean sometimes we have fallback code17:00
HeikoSwhen a library is not available17:01
HeikoSwhich might be less accurate17:01
wikingaccording to diff -uN ./data/testsuite/classifier/SubGradientSVM_30_1en05_16_05_False.m ./data/testsuite/classifier/SubGradientSVM_30_1en05_1_05_False.m17:02
wikingthe only difference is the number of threads17:02
HeikoSmmh17:02
HeikoSthat doesnt sound good ;)17:02
wikinganyhow it skips SubGradientSVM_30_1en05_1_05_False.m and SubGradientSVM_30_1en05_16_05_False.m fails with that error17:05
wikingif anybody has any idea what's going on please let me know, i.e. knows what package might be missing etc :17:06
wiking:)17:06
HeikoSwiking: sonney2k might know17:08
gsomixHeikoS, hey, have a minute? just little question. as you know lisitsyn came up new syntax for modelselection.17:15
gsomixlike ModelSelectionParameter("kernel").as(ValuesSet).from([gaussian_kernel, linear_kernel, chi2_kernel]);17:15
HeikoSgsomix:  sure17:15
HeikoSI think its documented in the issue17:15
gsomixhow do you think is it good to store ModelselectionParameter in SGObjects?17:16
gsomixfor syntax like yet_another_machine.paramete('kernel').as(...)...17:16
HeikoSgsomix: I dont understand17:17
HeikoSthe model selection parameters are already stored in the SGObject17:17
HeikoSbut we need to construct the parameter tree from this syntax17:18
HeikoScurently, those trees are created by hand17:18
HeikoSsee python modelselection examples17:18
HeikoSthis should be easier17:18
gsomixHeikoS, err. I mean to store parameters tree in SGObject.17:19
gsomixfor automagically constructing without new local variables in code17:19
HeikoSI see17:20
HeikoSso I would just make this syntax return a parameter tree17:20
HeikoSand those can then be appended to each other17:20
HeikoSand the modelselection class takes parameter trees as inputs17:20
-!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has joined #shogun17:23
gsomixHeikoS, got it, thanks. I have to think about...17:26
HeikoSgsomix: this is a messy topic17:26
HeikoSmaybe try to construct some trees by hands17:26
HeikoSI think most cases are covered in the examples17:26
HeikoSthis helped me a lot back then17:26
-!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has quit [Ping timeout: 245 seconds]17:29
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun17:33
-!- mode/#shogun [+o iglesiasg] by ChanServ17:33
-!- Yanglittle [deb20afe@gateway/web/freenode/ip.222.178.10.254] has quit [Quit: Page closed]17:48
shogun-notifier-shogun: lambday :develop * e9ed683 / src/shogun/mathematics/ (3 files): https://github.com/shogun-toolbox/shogun/commit/e9ed683c4153758ab3eb5e8bd2082060dc8a2bf718:13
shogun-notifier-shogun: minor edits in eigen318:13
shogun-notifier-shogun: Heiko Strathmann :develop * 0bf664c / src/shogun/mathematics/ (3 files): https://github.com/shogun-toolbox/shogun/commit/0bf664c80d073e996f3fbf42f936525afae6b1c918:13
shogun-notifier-shogun: Merge pull request #983 from lambday/develop18:13
shogun-notifier-shogun:18:13
shogun-notifier-shogun: Few edits in regarding EigenSparseUtil class in eigen3.h and eigen3.cpp18:13
-!- HeikoS [~heiko@nat-178-28.internal.eduroam.ucl.ac.uk] has left #shogun []18:34
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has quit [Read error: Operation timed out]18:44
-!- travis-ci [~travis-ci@ec2-54-242-66-169.compute-1.amazonaws.com] has joined #shogun18:45
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/641935718:45
-!- travis-ci [~travis-ci@ec2-54-242-66-169.compute-1.amazonaws.com] has left #shogun []18:45
-!- _rohan [~rohan@122.179.42.32] has joined #shogun19:09
-!- sumit [73f91219@gateway/web/freenode/ip.115.249.18.25] has joined #shogun19:30
-!- lisitsyn [~blackburn@109-226-120-160.clients.tlt.100megabit.ru] has joined #shogun20:01
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun20:09
-!- mode/#shogun [+o iglesiasg] by ChanServ20:09
gsomixgood evening20:16
gsomixsonney2k, there?20:17
-!- botton [~willie@router.isis.poly.edu] has joined #shogun20:37
-!- botton [~willie@router.isis.poly.edu] has left #shogun []20:37
gsomixsonney2k, I have problem with linking R modular interface. are you faced with this problems? http://pastebin.com/V3P6SgCS20:41
gsomixmaybe I forgot smth20:42
-!- sumit [73f91219@gateway/web/freenode/ip.115.249.18.25] has quit [Ping timeout: 245 seconds]20:43
-!- hoijui [~hoijui@dslb-088-074-124-102.pools.arcor-ip.net] has joined #shogun20:45
shogun-notifier-shogun: Soeren Sonnenburg :develop * 866f885 / / (13 files): https://github.com/shogun-toolbox/shogun/commit/866f88502d3b6e64b7816e0442e39033b1548ced20:55
shogun-notifier-shogun: drop flaky subgradientsvm / lpm20:55
lisitsynsonney2k: I like that!20:56
@iglesiasglisitsyn: what was wrong with them?20:57
lisitsyniglesiasg: no idea sonney2k knows better20:57
lisitsynbut I like dropping the code20:57
@iglesiasghaha ok20:57
-!- vikram360 [~vikram360@117.216.150.54] has joined #shogun21:05
-!- travis-ci [~travis-ci@ec2-50-16-8-109.compute-1.amazonaws.com] has joined #shogun21:06
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/642375221:06
-!- travis-ci [~travis-ci@ec2-50-16-8-109.compute-1.amazonaws.com] has left #shogun []21:06
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has quit [Ping timeout: 252 seconds]21:09
-!- vikram360 [~vikram360@117.216.150.54] has quit [Ping timeout: 245 seconds]21:12
-!- vikram360 [~vikram360@117.192.165.125] has joined #shogun21:13
shogun-notifier-shogun: Soeren Sonnenburg :develop * b74e53c / src/Makefile.template: https://github.com/shogun-toolbox/shogun/commit/b74e53c719edf225ce1c2a4a3bf576ced8d26ff021:21
shogun-notifier-shogun: don't fail on b0rken R modular tests / examples21:21
shogun-notifier-shogun: Soeren Sonnenburg :develop * 0f6c5a9 / .travis.yml: https://github.com/shogun-toolbox/shogun/commit/0f6c5a9c30d15f1ceaa2a96534f6645d2001719d21:21
shogun-notifier-shogun: require ruby 1.9.1 on travis21:21
-!- travis-ci [~travis-ci@ec2-50-16-8-109.compute-1.amazonaws.com] has joined #shogun21:37
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/642459221:37
-!- travis-ci [~travis-ci@ec2-50-16-8-109.compute-1.amazonaws.com] has left #shogun []21:37
vikram360Hi, I just wanted to know if Shogun has a style guide for development21:47
lisitsynvikram360: yes there is some infromation in README.developer21:47
vikram360lisitsyn Thank you21:48
shogun-notifier-shogun: Soeren Sonnenburg :develop * f08c329 / src/shogun/machine/Machine.h,src/shogun/ui/GUIClassifier.cpp: https://github.com/shogun-toolbox/shogun/commit/f08c32968d6cb43ec52e7072254df3d3325d920a21:48
shogun-notifier-shogun: drop subgradient svm/lpm from ui21:48
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun21:56
-!- mode/#shogun [+o iglesiasg] by ChanServ21:56
-!- travis-ci [~travis-ci@ec2-50-19-174-217.compute-1.amazonaws.com] has joined #shogun21:59
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/642521421:59
-!- travis-ci [~travis-ci@ec2-50-19-174-217.compute-1.amazonaws.com] has left #shogun []21:59
-!- Asix3 [452f9490@gateway/web/freenode/ip.69.47.148.144] has joined #shogun22:27
wikingsonne|work: here?22:29
-!- lisitsyn [~blackburn@109-226-120-160.clients.tlt.100megabit.ru] has quit [Quit: Leaving.]22:31
-!- lisitsyn [~blackburn@109-226-120-160.clients.tlt.100megabit.ru] has joined #shogun22:31
-!- mode/#shogun [+o lisitsyn] by ChanServ22:32
Asix3Greetings! I'm interested in working for Shogun through GSOC. I have a little background in both machine learning and C++, but not much else -- including git and working on large scale projects. I'm looking at Large Scale Learning & Dimension reduction using hashing.22:44
@lisitsynAsix3: hey22:44
Asix3@lisitsyn: Hi! I didn't expect a response so quickly. I'm basically wondering if I'd be a good candidate given my background. I've nearly worked through stanfords free machine learning course, now available on coursera -- found it challenging, but also interesting and completely doable. Aside from that I've taken a few courses in C++, two were introductory and one was on algorithms and data structures22:47
@lisitsynAsix3: I'd suggest start learning the project structure and try to improve anything (you may find some issues at github)22:47
Asix3@lisitsyn: Will do! Thank you!22:47
@lisitsynAsix3: we are searching for motivated and interested students :)22:48
@lisitsynso background matters but not that much22:48
Asix3I'm definitely both... Thanks again. You will be seeing more of me on this channel as I try to set up shogun and submit work on one of your entry tasks22:49
@lisitsynAsix3: would be nice!22:50
-!- hoijui [~hoijui@dslb-088-074-124-102.pools.arcor-ip.net] has quit [Ping timeout: 255 seconds]23:01
--- Log closed Thu Apr 18 00:00:28 2013