Open in new window / Try shogun cloud
--- Log opened Mon Aug 26 00:00:31 2013
shogun-buildbotbuild #1310 of cyg1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/1310  blamelist: Soeren Sonnenburg <sonne@debian.org>, Kevin <kevinhughes27@gmail.com>00:02
shogun-buildbotbuild #1521 of bsd1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/1521  blamelist: Kevin <kevinhughes27@gmail.com>00:03
shogun-buildbotbuild #49 of osx1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/osx1%20-%20libshogun/builds/49  blamelist: Kevin <kevinhughes27@gmail.com>00:09
shogun-buildbotbuild #1491 of deb2 - static_interfaces is complete: Failure [failed compile cmdline_static]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/1491  blamelist: Kevin <kevinhughes27@gmail.com>00:18
-!- iglesiasg [~iglesias@s83-179-44-135.cust.tele2.se] has quit [Quit: Ex-Chat]00:21
-!- travis-ci [~travis-ci@ec2-23-23-19-157.compute-1.amazonaws.com] has joined #shogun00:22
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/1060400300:22
-!- travis-ci [~travis-ci@ec2-23-23-19-157.compute-1.amazonaws.com] has left #shogun []00:22
shogun-buildbotbuild #987 of rpm1 - libshogun is complete: Failure [failed compile]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/rpm1%20-%20libshogun/builds/987  blamelist: Kevin <kevinhughes27@gmail.com>00:26
shogun-buildbotbuild #1492 of deb2 - static_interfaces is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/1492  blamelist: Soeren Sonnenburg <sonne@debian.org>00:42
shogun-buildbotbuild #1606 of deb3 - modular_interfaces is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/1606  blamelist: Soeren Sonnenburg <sonne@debian.org>00:43
shogun-buildbotbuild #1311 of cyg1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/1311  blamelist: Kevin <kevinhughes27@gmail.com>00:46
shogun-buildbotbuild #1522 of bsd1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/1522  blamelist: Soeren Sonnenburg <sonne@debian.org>00:49
shogun-buildbotbuild #50 of osx1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/osx1%20-%20libshogun/builds/50  blamelist: Soeren Sonnenburg <sonne@debian.org>00:50
shogun-buildbotbuild #988 of rpm1 - libshogun is complete: Failure [failed compile]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/rpm1%20-%20libshogun/builds/988  blamelist: Soeren Sonnenburg <sonne@debian.org>01:23
shogun-buildbotbuild #1312 of cyg1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/1312  blamelist: Soeren Sonnenburg <sonne@debian.org>01:29
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout]02:15
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has left #shogun ["WeeChat 0.4.0"]02:32
-!- pickle27 [~kevin@d67-193-243-174.home3.cgocable.net] has joined #shogun03:06
-!- pickle27 [~kevin@d67-193-243-174.home3.cgocable.net] has left #shogun []03:08
-!- sonne|osx_ [~sonne@g225080152.adsl.alicedsl.de] has joined #shogun03:53
-!- sonne|osx [~sonne@f053044032.adsl.alicedsl.de] has quit [Ping timeout: 264 seconds]03:55
-!- sonne|osx_ is now known as sonne|osx03:55
-!- lisitsyn [~lisitsyn@109-226-90-47.clients.tlt.100megabit.ru] has left #shogun []04:59
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun07:21
shogun-notifier-shogun: Soeren Sonnenburg :develop * 4d0b496 / .travis.yml: https://github.com/shogun-toolbox/shogun/commit/4d0b496966cc1738613b19d125cd5f66c89d5a8b07:21
shogun-notifier-shogun: use octave 3.6 ppa07:21
shogun-notifier-shogun: Soeren Sonnenburg :develop * d3541f5 / .travis.yml: https://github.com/shogun-toolbox/shogun/commit/d3541f565be1caad7f9d41ffdebfe2adb546c21207:45
shogun-notifier-shogun: add ppa to defaults07:45
shogun-notifier-shogun: Soeren Sonnenburg :develop * a2f7d25 / .travis.yml: https://github.com/shogun-toolbox/shogun/commit/a2f7d256385d71753f1fe99a02d7874b26df633f07:54
shogun-notifier-shogun: answer "yes"07:54
-!- sonne|osx [~sonne@g225080152.adsl.alicedsl.de] has quit [Quit: sonne|osx]08:01
shogun-buildbotbuild #1523 of bsd1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/1523  blamelist: Soeren Sonnenburg <sonne@debian.org>08:06
-!- travis-ci [~travis-ci@ec2-54-234-140-21.compute-1.amazonaws.com] has joined #shogun08:07
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/1061190508:07
-!- travis-ci [~travis-ci@ec2-54-234-140-21.compute-1.amazonaws.com] has left #shogun []08:07
shogun-buildbotbuild #1493 of deb2 - static_interfaces is complete: Success [build successful]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/149308:10
shogun-buildbotbuild #1524 of bsd1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/1524  blamelist: Soeren Sonnenburg <sonne@debian.org>08:11
shogun-buildbotbuild #1494 of deb2 - static_interfaces is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/1494  blamelist: Soeren Sonnenburg <sonne@debian.org>08:12
shogun-buildbotbuild #1495 of deb2 - static_interfaces is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/1495  blamelist: Soeren Sonnenburg <sonne@debian.org>08:12
shogun-buildbotbuild #1608 of deb3 - modular_interfaces is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/1608  blamelist: Soeren Sonnenburg <sonne@debian.org>08:13
shogun-buildbotbuild #989 of rpm1 - libshogun is complete: Success [build successful]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/rpm1%20-%20libshogun/builds/98908:13
shogun-buildbotbuild #1525 of bsd1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/1525  blamelist: Soeren Sonnenburg <sonne@debian.org>08:17
-!- sonne|osx [~sonne@89.204.139.45] has joined #shogun08:17
-!- travis-ci [~travis-ci@ec2-23-23-19-157.compute-1.amazonaws.com] has joined #shogun08:34
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/1061190508:34
-!- travis-ci [~travis-ci@ec2-23-23-19-157.compute-1.amazonaws.com] has left #shogun []08:34
shogun-buildbotbuild #1313 of cyg1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/1313  blamelist: Soeren Sonnenburg <sonne@debian.org>08:44
-!- sonne|osx [~sonne@89.204.139.45] has quit [Quit: sonne|osx]08:45
-!- travis-ci [~travis-ci@ec2-23-23-71-239.compute-1.amazonaws.com] has joined #shogun09:06
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/1061238809:06
-!- travis-ci [~travis-ci@ec2-23-23-71-239.compute-1.amazonaws.com] has left #shogun []09:06
-!- sonne|work [~sonnenbu@91-64-72-127-dynip.superkabel.de] has joined #shogun09:08
shogun-buildbotbuild #1607 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/160709:09
-!- lisitsyn [~lisitsin@mxs.kg.ru] has joined #shogun09:19
-!- mode/#shogun [+o lisitsyn] by ChanServ09:19
shogun-notifier-shogun: Soeren Sonnenburg :develop * 5dc2eb7 / .travis.yml: https://github.com/shogun-toolbox/shogun/commit/5dc2eb773f5d30ddc20dc2cfd9d216c14b79a5ed09:21
shogun-notifier-shogun: add another ppa for octave09:21
sonne|workohh man09:24
sonne|workanyone on ubuntu 12.04 LTS here?09:24
@lisitsynnot anymore09:25
sonne|workI wish travis was not using ubuntu LTS09:25
sonne|workstuff is too old09:25
@lisitsynsonne|work: I can install it to VM09:26
@lisitsynsonne|work: should I?09:27
sonne|worklisitsyn: I am just trying to install octave 3.6 w/ shogun now: http://blogs.bu.edu/mhirsch/2012/08/octave-3-6-on-ubuntu-12-04/09:27
sonne|workand it fails...09:27
@lisitsynsonne|work: what fails?09:27
sonne|worklisitsyn: well travis09:27
@lisitsynokay checking09:27
sonne|worklisitsyn: no idea why it doesn't work09:28
@lisitsynValueError: No JSON object could be decoded09:28
@lisitsynthis one?09:28
shogun-buildbotbuild #1314 of cyg1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/1314  blamelist: Soeren Sonnenburg <sonne@debian.org>09:28
shogun-buildbotbuild #1609 of deb3 - modular_interfaces is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/1609  blamelist: Soeren Sonnenburg <sonne@debian.org>09:28
@lisitsynsonne|work: I think ppa is not working09:29
@lisitsynsonne|work: https://launchpad.net/~mvanderkolff/+archive/octave-3.6 what about that ppa?09:29
shogun-buildbotbuild #1496 of deb2 - static_interfaces is complete: Success [build successful]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/149609:30
sonne|worklisitsyn: well no idea - it is trial and error09:31
shogun-buildbotbuild #1526 of bsd1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/1526  blamelist: Soeren Sonnenburg <sonne@debian.org>09:33
-!- thoralf [~thoralf@enki.zib.de] has joined #shogun09:41
thoralfHey.09:42
sonne|workthoralf: you ubuntu 12.04 LTS?09:44
thoralfOnly on my old laptop, which isn't available before tomorrow.09:45
-!- travis-ci [~travis-ci@ec2-23-23-19-157.compute-1.amazonaws.com] has joined #shogun10:01
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/1061405710:01
-!- travis-ci [~travis-ci@ec2-23-23-19-157.compute-1.amazonaws.com] has left #shogun []10:01
-!- iglesiasg [~iglesias@2001:6b0:1:1da0:19a:206d:eb6b:cdad] has joined #shogun10:03
-!- mode/#shogun [+o iglesiasg] by ChanServ10:04
@lisitsynsonne|work: tapkee is in jmlr vol. 14 pages 2287-2291 :)10:09
@lisitsynshould be ready soon we made final corrections10:10
shogun-buildbotbuild #1315 of cyg1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/1315  blamelist: Soeren Sonnenburg <sonne@debian.org>10:11
@lisitsynHeikoS: https://github.com/shogun-toolbox/shogun/issues/126510:41
sonne|worklisitsyn: congrats10:43
@lisitsynsonne|work: thanks10:43
@lisitsynsonne|work: please take a look https://github.com/shogun-toolbox/shogun/issues/126510:44
@lisitsynsonne|work: I do not suggest to use it yet as it looks messy but it is a way10:44
sonne|worklisitsyn: I think we are in trouble with swig - we would need to create %template for every getter10:44
@lisitsynsonne|work: well we will have like n_of_types getters10:45
@lisitsynlets say we have 10 types10:46
@lisitsynso we have 10 getters and 10 setters for all possible parameters10:46
@lisitsynoops10:49
@lisitsyn:D10:49
@iglesiasglisitsyn, *minima* concrete example :P?10:49
@iglesiasgminimal*10:49
@lisitsyniglesiasg: yes it is like what is the minimal music style to music10:50
@lisitsynI mean too minimal10:50
@lisitsyn:D10:50
@iglesiasghehe10:50
@lisitsynHeikoS: iglesiasg: sonne|work: please see the latest comment https://github.com/shogun-toolbox/shogun/issues/126510:54
@lisitsynplease ask me if you didn't get the idea or see so major drawback I missed10:56
@lisitsyns/so/a/10:57
@iglesiasglisitsyn, let me see10:57
-!- van51 [~van51@ppp-94-66-12-14.home.otenet.gr] has joined #shogun10:57
sonne|worklisitsyn: the get/set you implement in SGObject right?10:57
@lisitsynsonne|work: yes10:57
sonne|workvan51: hey there! How is it goind?10:57
@lisitsynget(Keyword<int>), get(Keyword<float>), get(Keyword<float-moat>)10:58
van51sonne|work: hello!10:58
@lisitsynwhatever types10:58
van51sonne|work: first of all I 've got some results on the lang_detection10:58
@iglesiasglisitsyn, what if you try to do some_object.get/set(keywords.parameter, ...)10:58
sonne|worklisitsyn: what do we do when we have say a combined kernel and want kernels as input?10:58
van51sonne|work: turns out  with a larger hash size, the quadratics perform better10:58
@iglesiasglisitsyn, and that some_object doesn't have that member10:58
sonne|workvan51: excellent!10:58
van51sonne|work: I also checked a bit the misclassifications10:58
@iglesiasglisitsyn, with member I mean the keywords.parameter10:59
@lisitsyntoo many messages :D10:59
van51sonne|work: and some of the sentences were trash, because they were generated randomly10:59
@lisitsynsonne|work: Keyword<SGObject*>10:59
@lisitsynsonne|work: I don't see the problem could you please elaborate?10:59
@lisitsyniglesiasg: runtime error11:00
van51sonne|work: I've also started working on RandomFourierDotFeatures11:00
@iglesiasglisitsyn, it would be nicer compile time error I think, but I can live with that too11:00
sonne|worklisitsyn: ok so we would do the type checking underneath? I mean if we want a CKernel we would need to do KeyWord<CKernel*> or do SGObject & checks?11:01
@iglesiasgmaybe what I just said doesn't make sense since this is to be used from python or so...11:01
@lisitsynsonne|work: I'd add checks but we can add keywords for ckernel11:01
@lisitsynsonne|work: runtime error is ok too I think but there is no trouble to add a keyword that needs CKernel*11:02
@lisitsynI would not add keywords for CMulticlassOCAS or whatever to reduce the number of exported keyword classes11:03
@lisitsynbut it is totally ok to add Keyword<CKernel*> kernel; I believe11:03
@lisitsyniglesiasg: no it is possible11:03
@lisitsyniglesiasg: you'd need a new class for every keyword though11:04
@lisitsyn:D11:04
@lisitsynbut you are right compile time doesn't make a lot sense as python just checks the syntax11:04
sonne|worklisitsyn: so what is the overhead - I mean we currently have a lot of stuff doing e.g. range checks or so in a for loop from local parameters11:07
@lisitsynsonne|work: it looks for me it is possible to make it quite fast11:08
@lisitsynwith hash map11:08
@lisitsynwe can ensure each keyword has unique hash11:08
@lisitsynmethod dispatching is fast (compile time in java for example)11:09
@lisitsynso we just have to look up for the value in map11:09
@iglesiasglisitsyn, good that hash tables are already in the C++ standard11:14
sonne|workvan51: sounds nice. I would be interested in actual results :)11:19
van51sonne|work: do you mean for the random fourier feats or for the lang detection?11:21
sonne|workvan51: lang11:22
@lisitsyniglesiasg: finally in 2011..11:23
@lisitsyniglesiasg: a bit late ain't it?11:24
@lisitsyn;)11:24
@iglesiasglisitsyn, what?11:24
@iglesiasgtapkee?11:24
@lisitsyniglesiasg: hash tables I mean11:25
@lisitsyn;)11:25
@iglesiasglisitsyn, aaah ok11:25
@iglesiasgyeah11:25
@lisitsynthere were in java in 90s11:25
sonne|worklisitsyn: seems like there are huge differences http://preshing.com/20110603/hash-table-performance-tests11:25
@lisitsynsonne|work: between what and what?11:26
sonne|workhash speed11:26
sonne|worktable11:26
@lisitsynsonne|work: what we shouldn't use is std::map :)11:26
@iglesiasgindeed, unordered_map is the thing11:26
@iglesiasgbut I mean it doesn't make sense to compare std::map and a hash table, they are just not for the same things11:27
sonne|workiglesiasg: what is the difference?11:28
@lisitsynsonne|work: O(ln N) vs O(1)11:28
@iglesiasgsonne|work, std::maps are balanced binary search trees and unordered_map is a hash table11:28
@lisitsynred-black tree vs hash tree11:28
@lisitsynerr11:29
@lisitsynhash table :D11:29
@lisitsynhash tree haha11:29
@iglesiasg:)11:29
@lisitsyniglesiasg: how does hash table work in java?11:29
@lisitsyniglesiasg: interview question time11:29
@iglesiasglisitsyn, haha11:29
van51sonney2k: https://gist.github.com/van51/633962811:29
@lisitsyniglesiasg: what happens if the hash function is not that good? a != b but hash(a) == hash(b)?11:30
@iglesiasglisitsyn, linked list11:30
@lisitsyngood11:30
@lisitsyn;)11:30
@lisitsyniglesiasg: how to break a hash map in java?11:30
@iglesiasgnot that good because I don't remember what that is called11:30
@iglesiasgarrch11:30
@iglesiasganyway11:30
@iglesiasglisitsyn, break?11:30
@lisitsyniglesiasg: yeah a bit, make it incostistent may be11:31
@iglesiasglisitsyn, what I find weird of hash tables in java is that you don't need to define your hash function11:32
@lisitsyniglesiasg: you do need ;)11:33
sonne|workvan51: what is combinations again?11:33
@iglesiasglisitsyn, mmm, are you sure?11:33
@iglesiasgIIRC there is something done by default11:33
@lisitsyniglesiasg: well they are usually auto-generated11:33
van51sonne|work: how many tokens/ngrams to combin11:33
@iglesiasglisitsyn, exactly, that is what I mean. IMHO, that is not good, at all11:33
sonne|workvan51: and n-gram length?11:34
@lisitsyniglesiasg: why? you write some POJO and it generates hash for you11:34
van51sonne|work: for these results, 311:34
sonne|workvan51: in any case ~90% is pretty okish already11:34
@lisitsyniglesiasg: I had a problem with entity graph though :D11:34
@lisitsynthere were cycles so hash was like %infty time11:34
van51sonne|work: also some sentences were trash so on that specific dataset that I've created I don't think it can go higher11:34
@iglesiasglisitsyn, I have no idea how is the hash function autogenerated. But for things that are not primitive data types, I think that should not be automatically made11:35
sonne|workvan51: well OK it is already nice. how big is the data set?11:35
@iglesiasglisitsyn, you just told me an example why :)11:35
sonne|workvan51: I mean if we .xz compress it - could it be stored in shogun-data?11:35
@lisitsyniglesiasg: that's why in the west coast enterprise yo systems you always have some base class11:36
@lisitsynI just use UUID most of the time11:36
@iglesiasglisitsyn, that sounds reasonable11:36
van51sonne|work: it's ~246M11:36
van51sonne|work: w/o compression11:36
van51sonne|work: how much is acceptable?11:36
@lisitsyniglesiasg: it is tiresome to implement these hashes and equals's11:37
@iglesiasglisitsyn, I tried once for my search engines course hash tables that extend to secondary disk. Cool stuff11:37
@lisitsyniglesiasg: some B-tree I guess11:37
@lisitsynahh11:37
@lisitsynhash table11:37
@iglesiasglisitsyn, no idea how was implemented internally. I should have been curious about that back then11:37
@lisitsyniglesiasg: people like B-trees when it comes to disk IO11:38
@iglesiasglisitsyn, databases are commonly b-trees, right?11:38
@lisitsyniglesiasg: they were (books say that) but I don't know what about now :)11:39
@iglesiasggood point :)11:40
@lisitsynthere are two guys in the same city with me that develop mysql11:40
@lisitsynno idea what makes them stay here but they kind of working remotely11:40
sonne|workvan51: well xz -9 compress them :)11:41
sonne|workvan51: it should be <100MB for sure11:41
van51sonne|work: compressing! I'll let you know11:42
@iglesiasglisitsyn, Eigen question11:44
@iglesiasglisitsyn, sqrt of a matrix. A.array().sqrt() or A.cwiseSqrt() but afterwards A remains unchanged, why?11:45
@lisitsyniglesiasg: you expect some eigenanswer11:45
@lisitsynhmm I think it shouldn't change anything11:45
@lisitsyniglesiasg: A = A.array.sqrt()11:46
@lisitsynyou can try that11:46
@iglesiasglisitsyn, I think I tried already, but let me check11:47
@iglesiasglisitsyn, it doesn't compile11:47
@lisitsynA.array()11:47
@lisitsyniglesiasg: A.array() = A.array().sqrt()?11:48
@lisitsynwhat about that/11:48
@iglesiasgmmm that I didn't come up with11:48
@iglesiasglet me see11:48
@iglesiasglisitsyn, yep, that made it11:49
@iglesiasglisitsyn, I got some problems with the data type though11:50
@iglesiasglisitsyn, it was not working before that I was using Matrix3i or Matrix3d, but I changed to MatrixXd and it worked11:50
@iglesiasglet me show you what I mean11:50
@lisitsyniglesiasg: Matrix3d is kind of special case I guess11:50
sonne|workvan51: and?11:51
van51sonne|work: it's still going11:51
sonne|workvan51: how big is it already then?11:52
van51sonne|work: 52M11:52
@iglesiasglisitsyn, but doing A.array() = A.array().sqrt() is efficiently doing the stuff in place?11:52
@iglesiasglisitsyn, I got the feeling that just A.array().sqrt() (wihtout the operator=) seems more efficient11:52
van51sonne|work: it's done.. 57M11:55
sonne|workvan51: ok so that is good enough IMHO11:55
sonne|worklisitsyn: heh http://www.sanmayce.com/Fastest_Hash/index.html#Yoshimura11:59
sonne|workYoshimura instead of murmur3 ;)11:59
@iglesiasgomg what is that12:00
sonne|workiglesiasg: looks like an assembly code optimized hash function :)12:02
@iglesiasgpeople are too hardcore12:03
van51sonne|work: should I place it in a new dir?12:03
sonne|workvan51: yes sure it is a real application..12:04
@wikingfuckiiiing JNI12:07
van51sonne|work: got a warning from github that file exceeds 50M12:12
sonne|workvan51: well it is only a warning or?12:13
van51sonne|work: but from what I read it should be ok, unless it needs an update12:13
van51sonne|work: it pointed me here : https://help.github.com/articles/working-with-large-files12:13
sonne|workvan51: you could then put your application under applications/language_detection12:15
van51sonne|work: ok, cool! I'll clean the code a bit first and push later12:18
sonne|workvan51: another hash http://burtleburtle.net/bob/hash/spooky.html12:20
sonne|workvan51: or https://code.google.com/p/cityhash/ :)12:21
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout]12:21
van51sonne|work: cityhash looks nice idd12:26
@lisitsynsonne|work: that's totally crazy12:30
@lisitsynunordered_map should be ok for 99% purposes12:30
@lisitsynI mean its default hashing function12:30
@lisitsynhowever you are doing some fancy things may be it would be ok here12:32
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has joined #shogun12:40
@iglesiasglisitsyn, sonne|work, HeikoS: about merging the ipython notebooks. How would you like to do it? One pull request at the end once they are finished or several as we go adding more stuff12:45
@iglesiasgsince I think this cannot really be revised as normal code I guess that maybe one pull request at the end with it is better12:45
@lisitsyniglesiasg: I have no opinion on that :)12:55
-!- van51 [~van51@ppp-94-66-12-14.home.otenet.gr] has quit [Quit: Leaving.]13:01
-!- HeikoS1 [~heiko@nat-188-24.internal.eduroam.ucl.ac.uk] has joined #shogun13:02
@lisitsynHeikoS1: !13:05
@lisitsynHeikoS1: HeikoS: need comments from your side on the 'keywords' idea13:06
HeikoS1lisitsyn: hi!13:12
HeikoS1its bank holiday here, excuse my late appearance :)13:13
-!- foulwall [~user@2001:da8:215:6901:11d1:4711:4290:8f68] has joined #shogun13:14
foulwall/usr/local/include/c++/4.8.1/atomic:1115:57: error: expected initializer before 'noexcept'13:17
foulwall     atomic_fetch_sub(atomic<_ITp*>* __a, ptrdiff_t __d) noexcept13:17
foulwall                                                         ^13:17
foulwallIn file included from /usr/local/include/shogun/base/SGObject.h:20:0,13:17
foulwall                 from /usr/local/include/shogun/io/File.h:16,13:17
foulwall                 from /usr/local/include/shogun/io/HDF5File.h:18,13:17
foulwall                 from library_hdf5.cpp:5:13:17
foulwall/usr/local/include/shogun/lib/RefCount.h: In constructor 'shogun::RefCount::RefCount(int32_t)':13:17
foulwall/usr/local/include/shogun/lib/RefCount.h:23:46: error: no matching function for call to 'std::atomic<int>::atomic(int32_t&)'13:17
foulwall  RefCount(int32_t ref_start=0) : rc(ref_start) {}13:17
foulwall                                              ^13:17
foulwall/usr/local/include/shogun/lib/RefCount.h:23:46: note: candidate is:13:17
HeikoS1foulwall: thats c++ issue13:17
foulwallIn file included from /usr/local/include/shogun/lib/RefCount.h:2:0,13:17
foulwall                 from /usr/local/include/shogun/base/SGObject.h:20,13:17
HeikoS1you have to compile with the latest c++13:17
foulwall                 from /usr/local/include/shogun/io/File.h:16,13:17
foulwall                 from /usr/local/include/shogun/io/HDF5File.h:18,13:18
foulwall                 from library_hdf5.cpp:5:13:18
foulwall/usr/local/include/c++/4.8.1/atomic:599:7: note: std::atomic<int>::atomic()13:18
foulwall       atomic() noexcept = default;13:18
foulwall       ^13:18
HeikoS1-std=c++1113:18
foulwall/usr/local/include/c++/4.8.1/atomic:599:7: note:   candidate expects 0 arguments, 1 provided13:18
foulwallmake: *** [library_hdf5] Error 113:18
foulwallI compile it with gcc 4.8.113:18
foulwallok thanks HeikoS113:18
HeikoS1foulwall: add this flag13:18
HeikoS1foulwall: if this is your own program, that should do it13:18
HeikoS1if this is from cmake, no idea whats causing it13:18
foulwallfrom examples/undocumented/libshogun/library_hdf5.cpp13:19
HeikoS1foulwall: yes then compile with the flag13:20
HeikoS1lisitsyn: cool stuff13:20
foulwallthanks HeikoS:)13:20
HeikoS1lisitsyn: but the global keywords are unreaslistic13:20
HeikoS1could that be done with namespaces?13:20
HeikoS1lisitsyn: what do you think on13:21
HeikoS1https://github.com/shogun-toolbox/shogun/issues/1460#issuecomment-2323579313:21
votjakovrHeikoS1: hi! The difference between our EP in combination with probit likelihood and EP from GPML is no more than 1e-3. Should i send a PR?13:25
HeikoS1votjakovr: hi ah finally we meet13:25
@lisitsynHeikoS1: lets talk about that13:25
HeikoS1votjakovr: that should be fine, how is it against GPstuff?13:25
@lisitsynHeikoS1: I need to clarify - what's unrealistic in global keywords?13:25
HeikoS1lisitsyn: imagine the parameter keywork "k"13:26
HeikoS1that will be used everywhere13:26
votjakovrHeikoS1: i think that this is because of random13:26
@lisitsynHeikoS1: you mean name clash with variables?13:26
HeikoS1votjakovr: could you elaborate a bit?13:26
HeikoS1lisitsyn: yes13:26
@lisitsynHeikoS1: we can put them to inner scope13:26
@lisitsynKNN.k13:26
@lisitsynwe can put them to any namespace13:27
HeikoS1lisitsyn: is the KNN generated automatically?13:27
HeikoS1or do people have to add that?13:27
votjakovrHeikoS1: since difference between negative log marginal likelihoods is something about 1e-1013:27
@lisitsynno you'd have to add that13:27
HeikoS1lisitsyn: how do you see that?13:27
HeikoS1is there a way to check that?13:27
@lisitsynHeikoS1: to check what?13:27
HeikoS1lisitsyn: so in the optimal world, devs would just have to register the parameter, and then everything else is generated automatically13:28
HeikoS1votjakovr: ok, but you are also integrating differently right?13:28
@lisitsynHeikoS1: this way you'd need to add a keyword if it is not here13:29
HeikoS1votjakovr: maybe test another dataset to make sure that the difference is low everywhere13:29
votjakovrHeikoS1: the order of the local likelihood approximations is random13:30
HeikoS1votjakovr: I see13:30
@lisitsynHeikoS1: I don't get the uniqueness point13:30
HeikoS1lisitsyn: maybe I get it wrong13:30
HeikoS1lisitsyn: could you tell me what I would have to do on writing a new class with parameter "k" ?13:30
votjakovrHeikoS1: and GPML has a difference between calls on the same dataset13:30
HeikoS1votjakovr: ah ok, well then this is fine13:31
@lisitsynHeikoS1: define `static const Keyword k;`13:31
@lisitsynand mark somehow that you will use it13:31
HeikoS1lisitsyn: what if another class also has a parameter k?13:31
votjakovrHeikoS1: ok, i'll send a PR13:31
HeikoS1votjakovr: nice!13:31
@lisitsynHeikoS1: totally ok13:31
@lisitsynHeikoS1: the keyword doesn't store any value13:31
@lisitsynit is just used to dispatch the type13:32
HeikoS1lisitsyn: ok13:32
HeikoS1lisitsyn: float k, int k, both ok?13:32
@lisitsynahh13:32
@lisitsyndifferent types?13:32
HeikoS1yes13:32
@lisitsynyes that's a problem13:32
HeikoS1lisitsyn: also keep in mind what our goal was: That one registers a parameter once, and then getters/setters are automatically generated13:33
@lisitsynwell that's just name clash13:33
@lisitsynHeikoS1: yes that fits the goal13:33
HeikoS1so it would be best if the SG_ADD, or a similar line would do that13:33
@lisitsynHeikoS1: use(keyword.k);13:33
@lisitsynHeikoS1: that's how I see it13:33
sonne|workiglesiasg: I prefer incremental contributions wrt notebooks13:33
@iglesiasgsonne|work, all right13:34
@lisitsynHeikoS1: static keyword instance is just a replacement for string13:34
@lisitsynbut it can be used to infer the type13:34
@lisitsynthat's the only difference13:34
sonne|worklisitsyn: sure this is more about what van51 is doing :)13:34
@lisitsynsonne|work: yeah then makes sense with such a massive hashing13:34
HeikoS1I am confused ;)13:35
@lisitsynHeikoS1: but do you think it is a common case with float k and int k?13:35
@lisitsynI mean it is much better to call it n_neighbors13:35
HeikoS1lisitsyn:  I just think we should not sacrifice flexibility13:35
HeikoS1lisitsyn: because currently we can do that13:35
HeikoS1lisitsyn: but say this is not a problem13:35
@lisitsynHeikoS1: well it is just a suggestion :)13:36
sonne|worklisitsyn: do you have gcc 4.8?13:36
@lisitsynsonne|work: not really13:36
sonne|workfoulwall: hmmhh is this the latest clone on what arch?13:37
HeikoS1lisitsyn: I like it, just not sure whether I understand it13:37
@lisitsynHeikoS1: what is unclear? let me explain :)13:37
@lisitsynI am afraid I am losing some point so I'd like to discuss it13:37
HeikoS1lisitsyn: yes cool13:37
HeikoS1lisitsyn: I guess an example would be cool13:38
HeikoS1where a parameter is registered13:38
HeikoS1and then how to do get/set13:38
@lisitsynHeikoS1:well it is not really about parameter registration13:38
foulwallsonne|work: latest clone and on debian, in examples/undocument/libshogun, I add -std=C++11 in libshogun's Makefile and now it works fine13:39
HeikoS1lisitsyn: I know, but thats what I dont get yet13:39
@lisitsynHeikoS1: in sgobject we'd have a map<Keyword,any>13:39
@lisitsynHeikoS1: so once you register parameter you add something to the map13:39
sonne|workwiking: are we compiling examples with same flags like c++11 etc like libshogun?13:40
@lisitsynsay you call float get(Keyword<float>)13:40
@lisitsynit finds the keyword, casts the any value to float13:40
@lisitsynand returns it13:40
foulwallsonne|work: there's no c++11 flags in libshogun's makefile13:40
sonne|workfoulwall: yeah but you build with cmake now right?13:40
HeikoS1lisitsyn: ok13:40
votjakovrHeikoS1: i've just sent it :)13:41
HeikoS1lisitsyn: should we start a feature branch with that?13:41
@lisitsynHeikoS1: I don't know13:41
HeikoS1votjakovr: reading :)13:41
@lisitsynHeikoS1: I can try13:41
foulwallsonne|work: Ah, for examples I built with make. I'll try cmake13:41
@lisitsynHeikoS1: it doesn't sound like a big amount of work actually13:41
HeikoS1lisitsyn: that would be even better :)13:41
@lisitsynHeikoS1: the issue that bothers me13:42
@lisitsynand you too probably13:42
@lisitsynis there to place there keywords or whatever13:42
@lisitsynthese keywords*13:42
@lisitsynwhere*13:42
@lisitsyn:D13:42
@lisitsynshit13:42
@lisitsynHeikoS1: so the class scope or namespace13:42
@lisitsynor globally13:42
@lisitsynI don't know13:43
@lisitsynclass scope is bad because you write the class name too many times13:43
@lisitsynnamespace scope is ok but a bit messy13:43
HeikoS1lisitsyn: ehm what?13:44
@lisitsynglobal scope can lead to name clashes as people do `from modshogun import *` sometimes13:44
@lisitsynHeikoS1: class scope is KNN.k13:44
HeikoS1lisitsyn:  I see13:44
@lisitsynby namespace scope I mean kw.k13:44
@lisitsynso modshogun.kw.k13:44
@lisitsynlike that13:44
HeikoS1no class scope should be better13:44
HeikoS1dont you think?13:44
@lisitsynwell depends13:44
@lisitsynHeikoS1: actually quite a few parameters are shared13:45
@lisitsynn_neighbors, n_iters13:45
@lisitsynwidth, etc13:45
HeikoS1yep13:46
@lisitsynHeikoS1: I expect it speeds up compilation13:46
HeikoS1lisitsyn: thats good13:46
HeikoS1lisitsyn: a question: will this allow to manually override getters/setters?13:47
@lisitsynHeikoS1: but I don't know how much13:47
HeikoS1lisitsyn: since the idea also was to use this for safer model-selection calls13:47
@lisitsynHeikoS1: there should be a way13:47
HeikoS1you know that for example one can make sure stter only accepts pos. values of an integer13:47
@lisitsynHeikoS1: I think we need some 'checkers' for that13:48
@lisitsynso you attach Positive checker to parameter13:48
@lisitsynit is a rare case you really need something custom13:48
-!- foulwall [~user@2001:da8:215:6901:11d1:4711:4290:8f68] has quit [Ping timeout: 264 seconds]13:48
@lisitsynHeikoS1: static const Keyword<float> width("Width of gaussian kernel", positive + non_zero);13:49
HeikoS1lisitsyn: no wait13:49
HeikoS1the point why I suggested that originally was13:49
HeikoS1that in model-selection, I can only modify the parameters directly13:50
HeikoS1but sometimes, one has some post methods that need to be called or statements executed13:50
@lisitsynHeikoS1: this is obvious observer pattern for me I think it can be done13:50
HeikoS1lisitsyn: lets thing of an example13:51
HeikoS1ah yes13:51
HeikoS1Gaussian kernel13:51
HeikoS1the diagonal is precomputed13:51
@lisitsynwe just add some NOP method that is called once parameter is changed13:51
HeikoS1lisitsyn: actually, the setter does not even update it13:52
HeikoS1argh13:52
HeikoS1lisitsyn: ok13:52
HeikoS1that would do it13:52
@lisitsynHeikoS1: yeah got it13:52
HeikoS1so currnetly GaussianKernel::init(lhs,rhs);GaussianKernel.set_width(other_value); GaussianKernel::get_kernel_matrix13:53
HeikoS1returns wrong results13:53
@lisitsynwhy?13:53
HeikoS1ah no13:53
HeikoS1precompute squared just computes dot products13:53
HeikoS1sorry13:53
@lisitsynHeikoS1: I think there is nothing infeasible here but good you mention it13:54
HeikoS1votjakovr: I will wait for travis and then merge, good work!13:54
HeikoS1lisitsyn: oh one more thing13:56
@lisitsynHeikoS1: yes?13:56
HeikoS1lisitsyn: is it possible for the model-selection framework to use that?13:56
votjakovrHeikoS1: good :)13:56
HeikoS1lisitsyn: since in there, one has only access to the registered parameters13:56
@lisitsynHeikoS1: too generic question ;)13:56
HeikoS1as shogun parameters13:56
HeikoS1so TParameter13:56
@lisitsynHeikoS1: it doesn't allow *unregistered* parameters13:56
@lisitsynyou just can't use it if it is not in map13:57
HeikoS1lisitsyn: no I mean13:57
HeikoS1from the source code, I need to call things13:57
HeikoS1let me show you how I do it currently13:57
HeikoS1ParameterCombination::apply_to_modsel_parameter13:58
HeikoS1lisitsyn:  and then this uses Parameter::set_from_parameters13:59
HeikoS1in there I even have to do a type distinction13:59
@lisitsynHeikoS1: I don't really get it14:00
HeikoS1the grid search iterates over all possible parameter combinations14:00
HeikoS1and then applies them to SGObjects14:00
@lisitsynyes14:00
HeikoS1this is currently done via just copying the memory14:00
HeikoS1or pointer etc14:00
HeikoS1will it be possible to do this with the setters/getters?14:01
@lisitsynHeikoS1: yes probably if I get what you mean14:01
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun14:02
shogun-notifier-shogun: Roman Votyakov :develop * fa3a7c9 / / (3 files): https://github.com/shogun-toolbox/shogun/commit/fa3a7c92bbd0c0e78b3b0c9d8d9f16fb6986929514:02
shogun-notifier-shogun: add expectation propagation inference method14:02
shogun-notifier-shogun: Heiko Strathmann :develop * 8aee21e / / (3 files): https://github.com/shogun-toolbox/shogun/commit/8aee21eed0b16d18b348976da5bb51e11fe6332514:02
shogun-notifier-shogun: Merge pull request #1462 from votjakovr/feature/gp_binary_classification14:02
shogun-notifier-shogun:14:02
shogun-notifier-shogun: Add expectation propagation inference method14:02
HeikoS1lisitsyn: though, as I said, I would totally drop grid-search14:02
HeikoS1lisitsyn: the effort of using our framework is larger than doing it by hand if cross-validation is available14:02
@lisitsynHeikoS1: yeah that's why I try to think how to do that better way14:02
HeikoS1lisitsyn:  still14:03
HeikoS1the amount of work to maintain that is just insane14:03
HeikoS1and it sooo easy to do  it by hand14:03
@lisitsynHeikoS1: I believe it is too low-level14:03
@lisitsynpointers etc14:03
@lisitsynindexes14:03
@lisitsynenums14:04
@lisitsynbut I am not sure14:04
HeikoS1lisitsyn: exactly14:04
@lisitsynif it could be done better14:04
HeikoS1the whole parameter tree14:04
HeikoS1paramter combination14:04
HeikoS1etc14:04
HeikoS1lisitsyn:  I dont really see another way to do it14:05
HeikoS1lisitsyn: the problem is the way how shogun parameters are registered14:05
@lisitsynyeah you would do it better if you knew :)14:05
HeikoS1and creating parameter data from "nowhere" is just hard14:05
HeikoS1lisitsyn: also, no debugging of model-selection14:05
HeikoS1remember this x-validation output class?14:06
@lisitsynyes14:06
HeikoS1we are lacking that for grid-search14:06
HeikoS1its easy to do, but again, its effort14:06
HeikoS1and if people do this by hand, its not14:06
HeikoS1but later14:07
HeikoS1ill fix the xml serialization for now :)14:07
@lisitsynHeikoS1: alright let me think about it in background daemon14:08
@lisitsynas usual14:08
HeikoS1ok :)14:08
HeikoS1lisitsyn: cool ideas you had though, I would love to see this happen14:08
@lisitsynyou may notice I disappeared for a month and then came with some idea14:08
@lisitsyn:D14:08
HeikoS1haha :)14:08
HeikoS1lisitsyn: btw do you have any idea why we use cstrings rather than string?14:08
@lisitsynHeikoS1: cstrings - const char*?14:09
HeikoS1yes14:09
@lisitsynno idea14:09
HeikoS1lisitsyn: just so dirty14:09
@lisitsynwe are in std haters club14:09
HeikoS1haha :)14:09
thoralf:D14:09
@lisitsynthe first rule14:09
@lisitsynis never use std::14:09
thoralfshogun is implemented in C with classes :)14:10
HeikoS1thoralf: indeed :)14:10
thoralflisitsyn: Rule 1: Whenever you can reinvent the wheel: Do it.14:10
@lisitsynHeikoS1: tech-wise we also need shared_ptr14:10
@lisitsynthoralf: yes true14:11
@lisitsynthe wheel is not round enough14:11
thoralflisitsyn: SGVector (or better: void *) instead of std::vector14:11
@lisitsynthoralf: sgvector is actually something else than std::vector now14:12
thoralflisitsyn: I know.14:12
@lisitsynI hate everything about it though14:12
@lisitsynall these cmaps and etc14:12
thoralflisitsyn: Elaborate?14:12
@lisitsyndynamicobjectarray14:12
@lisitsynor how it is called14:12
@lisitsynI wish we didn't had that code14:12
HeikoS1lisitsyn: go for it! kill it14:13
HeikoS1remove things step by step :)14:13
HeikoS1lisitsyn: we are currently on a good run to make things better, so this is another one on the list14:13
HeikoS1lisitsyn: and you are the expert ;)14:13
thoralflisitsyn: I'd be happy to use a thin std::vector wrapper instead of sgvector. :)14:13
@lisitsynthoralf: why?14:14
thoralflisitsyn: Just to simplify code.14:14
shogun-buildbotbuild #1527 of bsd1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/1527  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>14:14
@lisitsynthoralf: I'd like to see Eigen::VectorXd here14:14
thoralflisitsyn: Even better.14:15
thoralfStop.  No reinventing?  No!14:15
@lisitsynthoralf: opencv uses just Mat14:15
@lisitsynstupid naming but engineering wise it looks better14:16
HeikoS1lisitsyn: btw we should also have a wrapper class that makes a CSGObject out of SG Types14:17
HeikoS1to be able to use them in lists etc14:17
@lisitsynHeikoS1: what lists?14:17
HeikoS1like a list of vectors14:17
HeikoS1different sized vectors14:17
HeikoS1sorry14:18
HeikoS1not a list14:18
HeikoS1an array that dynamically grows14:18
@lisitsynstd::vector?14:18
@lisitsyn;)14:18
HeikoS1lisitsyn: yes14:18
HeikoS1but serializable14:18
HeikoS1lisitsyn: a shogun type14:18
@lisitsynahh I have no idea about it14:18
HeikoS1we could even define a wrapper class for std::vector14:18
thoralflisitsyn: Anyway, I think that SGSparse* is a good thing. :)14:19
HeikoS1which IS a CSGObject14:19
votjakovrHeikoS1: wow EP works much faster than Laplace!14:25
HeikoS1votjakovr: really?14:25
HeikoS1votjakovr: thats weird14:25
HeikoS1usually its more expensive14:25
HeikoS1I wonder why that is14:25
votjakovrHeikoS1: yep i know, but i've tried graphical example with EP and it is subjectively faster14:26
HeikoS1votjakovr: ok, well thats great then14:27
HeikoS1votjakovr: we can always tune later14:27
HeikoS1votjakovr: did you see my pm?14:27
votjakovrHeikoS1: i think in future we could make a benchmark test and try to optimize bottlenecks14:28
votjakovrHeikoS1: pm?14:29
HeikoS1votjakovr: yep that would be awesome! also to have benchmarks which other people can use14:29
HeikoS1votjakovr: my private message in irc14:29
HeikoS1votjakovr: btw did you already start on the notebook?14:33
HeikoS1I liked your graphical examples a lot14:33
HeikoS1they should be in there definitely14:33
-!- lisitsyn [~lisitsin@mxs.kg.ru] has quit [Read error: Connection reset by peer]14:40
-!- travis-ci [~travis-ci@ec2-23-23-71-239.compute-1.amazonaws.com] has joined #shogun14:40
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/1062111214:40
-!- travis-ci [~travis-ci@ec2-23-23-71-239.compute-1.amazonaws.com] has left #shogun []14:40
@iglesiasgwhat's up with travis?14:47
@iglesiasgit seems that the ppa for octave change killed it14:47
HeikoS1yep14:48
shogun-buildbotbuild #1610 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/161014:48
-!- lisitsyn [~lisitsin@mxs.kg.ru] has joined #shogun14:50
shogun-buildbotbuild #992 of rpm1 - libshogun is complete: Failure [failed compile]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/rpm1%20-%20libshogun/builds/992  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>14:51
shogun-buildbotbuild #1611 of deb3 - modular_interfaces is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/1611  blamelist: Roman Votyakov <votjakovr@gmail.com>14:53
@iglesiasgHeikoS1: is someone trying to fix that?14:53
HeikoS1iglesiasg: not aware of :)14:53
HeikoS1I am fixing other things currently14:53
shogun-buildbotbuild #1316 of cyg1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/1316  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>14:54
shogun-buildbotbuild #1528 of bsd1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/1528  blamelist: Roman Votyakov <votjakovr@gmail.com>14:57
@iglesiasgsonne|work, why is the second ppa for octave 3.6 added?14:58
sonne|workiglesiasg: remove it - it didn't work and I cannot from work14:58
@iglesiasgsonne|work, ok14:59
@iglesiasgsonne|work,  it starts with sudo sudo add-apt instead of sudo add-apt14:59
@iglesiasgmaybe that is the problem14:59
sonne|workiglesiasg: don't think so but still a bug15:00
@iglesiasgyeah.. the trace looks unrelated to that, but I will try without sudo twice first just in case15:01
shogun-notifier-shogun: Fernando Iglesias :develop * ebf640f / .travis.yml: https://github.com/shogun-toolbox/shogun/commit/ebf640f341ee65b4996d8640f35570f5b3e3319615:03
shogun-notifier-shogun: Remove double sudo in travis octave 3.6 second ppa.15:03
@iglesiasgnah, it didn't fix it15:05
HeikoS1iglesiasg: please dont use "m_" as parameter names15:07
HeikoS1I will change that in you LMNN class15:07
shogun-notifier-shogun: Fernando Iglesias :develop * 765e145 / .travis.yml: https://github.com/shogun-toolbox/shogun/commit/765e145c23292f9cd8f8b678aaecffb330e71cf115:07
shogun-notifier-shogun: Remove octave 3.6 second ppa in travis.15:07
@iglesiasgHeikoS1, eeeh all right, but why not?15:07
@iglesiasgI think it is a good practice to diff the members somehow15:08
HeikoS1thats more of a style thing for the member variables15:08
HeikoS1but the variable name should be the parameter name15:08
@iglesiasgwhat do you mean?15:08
@iglesiasgI am just using it for the members, right?15:08
HeikoS1yep15:09
HeikoS1so CClass:m_k is registered as "k"15:09
HeikoS1not a big thing, I just saw it15:09
@iglesiasgaaaah ok15:09
@iglesiasgjust when doing SG_ADD15:09
HeikoS1yes :)15:09
@iglesiasgnot changing the name of attributes15:09
@iglesiasgsure :)15:09
@iglesiasgHeikoS1: out of curiosity, why? I mean, for what are used later these names?15:10
@iglesiasgI guess that for the SGObject clone is a reason15:10
HeikoS1iglesiasg: just whenever the parameter is accessed written or read15:11
HeikoS1also for modelselection15:12
HeikoS1so simpler is better there15:12
HeikoS1oh man15:12
HeikoS1iglesiasg: I justed looked into the PCA class15:13
HeikoS1thats pure horror15:13
@iglesiasgyeah, a little bit :)15:13
@iglesiasgit took me a little while to discover how to get the transform matrix when I wanted to use it for PCA15:13
@iglesiasgthere were some arguments to tune and so15:14
HeikoS1PCA is a three liner why is that so complicated ...grrrr15:14
@iglesiasghehe yeah15:14
HeikoS1but no eigen was around back then15:15
@iglesiasgI mean, it is good that it can be tunable, the PCA we have does extra stuff15:15
HeikoS1true15:15
@iglesiasgbut I think it could be way more readable using Eigen and still keeping those functionalities15:15
HeikoS1iglesiasg: feel free to change :) also a unit test might be nice for such a basic thing15:16
@iglesiasgbut at the end of the day, it works just fine so I am not sure to the point it pays off to do it15:16
@iglesiasgHeikoS1, unit test is a good idea indeed15:16
@iglesiasgHeikoS1: BTW, we should set up a shogun-team meeting for the NIPS workshop15:17
HeikoS1iglesiasg: yep totally15:17
HeikoS1wiking is the guy to do :)15:17
shogun-buildbotbuild #993 of rpm1 - libshogun is complete: Success [build successful]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/rpm1%20-%20libshogun/builds/99315:17
shogun-buildbotbuild #1529 of bsd1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/1529  blamelist: Fernando Iglesias <fernando.iglesiasg@gmail.com>15:18
@iglesiasgI don't mind setting up the doodle, I will do it this evening probably15:18
-!- pickle27 [~kevin@d67-193-243-174.home3.cgocable.net] has joined #shogun15:18
lisitsynHeikoS1: I thought you wanted to go? ;)15:19
lisitsynit can be any of us though, I know I am off15:20
-!- HeikoS1 [~heiko@nat-188-24.internal.eduroam.ucl.ac.uk] has quit [Ping timeout: 245 seconds]15:22
@wikingHeikoS: ?15:28
-!- HeikoS1 [~heiko@nat-188-24.internal.eduroam.ucl.ac.uk] has joined #shogun15:28
@wikingah yeah the meeting15:28
@iglesiasgyeah15:28
@wikingi'll send around a doodle today15:28
@iglesiasgcool!15:28
* wiking is a bit busy today with errands... no commits today :P15:29
@iglesiasgwiking, I am excited about this NIPS thing, it would be cool if we can do something for it15:29
@wikingyeah indeed15:29
@iglesiasgwiking, we have been mistreating travis :-O15:29
@wikingiglesiasg: heheh i saw...15:29
@wikingfixit :)15:29
@iglesiasgit seems to be fine again though15:30
@iglesiasgoh no fuck15:30
HeikoS1tada15:30
HeikoS1XML serialization is lossy15:30
@wikingiglesiasg: it's fine... i've restarted the job... it was just some randome15:30
@wikingerror ;P15:30
@iglesiasgwiking, yes, MulticlassOCASTest.train, it is not the first time it fails I think mm15:31
@wikingyeah15:32
@wikingthat unittest should be a bit modified + random seed should be fixed imho15:33
shogun-notifier-shogun: Heiko Strathmann :feature/SerialUTests * 2741b0c / src/shogun/ (7 files): https://github.com/shogun-toolbox/shogun/commit/2741b0c8c10e49012fa51a6485387f2f506ba51a15:37
shogun-notifier-shogun: removed spaces from parameter names that are registered15:37
shogun-notifier-shogun: Heiko Strathmann :feature/SerialUTests * e11517f / src/shogun/base/Parameter.cpp: https://github.com/shogun-toolbox/shogun/commit/e11517f0012428b96614094cd3a56ddcf20342af15:37
shogun-notifier-shogun: only allow alnum and underscore for registered parameter names15:37
shogun-notifier-shogun: Heiko Strathmann :feature/SerialUTests * 9759434 / tests/unit/io/SerializationXML_unittest.cc.jinja2: https://github.com/shogun-toolbox/shogun/commit/97594343003d94a63293737877a0135a2dbeee0115:37
shogun-notifier-shogun: introduce accuracy for xml serialization equal tests15:37
shogun-notifier-shogun: Heiko Strathmann :feature/SerialUTests * a022433 / tests/unit/CMakeLists.txt: https://github.com/shogun-toolbox/shogun/commit/a0224338eabf701264a66dfd9b0551de00b8871a15:37
shogun-notifier-shogun: activate XML serialization since it works15:37
HeikoS1wiking: XML works!15:37
HeikoS1wiking: json last one15:38
shogun-buildbotbuild #1317 of cyg1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/1317  blamelist: Roman Votyakov <votjakovr@gmail.com>15:38
-!- travis-ci [~travis-ci@ec2-23-23-19-157.compute-1.amazonaws.com] has joined #shogun15:42
travis-ci[travis-ci] it's Fernando Iglesias'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/1062277115:42
-!- travis-ci [~travis-ci@ec2-23-23-19-157.compute-1.amazonaws.com] has left #shogun []15:42
shogun-notifier-shogun: Roman Votyakov :develop * 3171861 / / (4 files): https://github.com/shogun-toolbox/shogun/commit/31718619c67e1c5543305c1add80b613fbdf68cf15:44
shogun-notifier-shogun: add expectation propagation to modular interfaces15:44
shogun-notifier-shogun: Heiko Strathmann :develop * a447d63 / / (4 files): https://github.com/shogun-toolbox/shogun/commit/a447d636d7cc5bd1233076d90cb6d8ebbc91211b15:44
shogun-notifier-shogun: Merge pull request #1463 from votjakovr/feature/gp_binary_classification15:44
shogun-notifier-shogun:15:44
shogun-notifier-shogun: Add expectation propagation to modular interfaces15:44
-!- pickle27 [~kevin@d67-193-243-174.home3.cgocable.net] has left #shogun []15:47
-!- thoralf [~thoralf@enki.zib.de] has quit [Read error: Connection reset by peer]15:47
-!- thoralf [~thoralf@enki.zib.de] has joined #shogun15:49
thoralfYo.15:49
HeikoS1yoyo15:50
thoralf:)15:50
thoralfyolo15:50
HeikoS1krkr ;)15:51
thoralfHeikoS1: I don't speak russian. ;)15:52
HeikoS1haha :D15:52
shogun-buildbotbuild #1530 of bsd1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/1530  blamelist: Fernando Iglesias <fernando.iglesiasg@gmail.com>15:57
votjakovrHeikoS1: what system is used on travis-ci? I'd like to enable nlopt package on travis-ci, since gradient model selection depends on it.15:58
HeikoS1votjakovr: best ask wiking for that, I think you can just sudo-apt get install the usual stuff15:58
HeikoS1have a look how we do it for eigen etc15:59
@wikingvotjakovr: is nlopt available as ubuntu package?15:59
@wikingvotjakovr: i'm afraid that it's only available in raring http://packages.ubuntu.com/raring/libnlopt-dev16:01
votjakovrwiking: yep, since 12.1016:01
@wikingvotjakovr: yeah and travis-ci is not 12.1016:01
@wikingit's like 12.0416:01
@wikingso you have to write a bundle script for it16:01
@wikingsee cmake/external/16:01
votjakovrwiking: ok, thanks, i'll do it like there :)16:02
@wikingand try to write a patch accordingly16:02
@wikingbundle it as a static lib16:02
@wikingso just write it send a PR tag me and i'll review it16:03
lisitsynthoralf: russian sounds a bit different ;)16:03
votjakovrwiking: ok, thanks16:03
thoralflisitsyn: Yeah.  From chinese perspective japanese is different.  From european it's not. ;)16:04
-!- lambday [67157f4c@gateway/web/freenode/ip.103.21.127.76] has joined #shogun16:06
lisitsynthoralf: iglesiasg had to listen russian quite a lot in berlin :D16:06
@iglesiasghaha not that much :P16:06
lisitsyniglesiasg: all the time you walked with us barbarians16:07
HeikoS1can anyone help me compiling with json?16:07
lisitsyniglesiasg: oh I remember the FBI thing16:07
@iglesiasglisitsyn, barbarians? I wouldn't say that taking into account the Russian embassy in Berlin hehe16:07
HeikoS1what do I need to pass to my program compiler if I used Cmake to fetch it?16:07
@iglesiasgHeikoS1, do you need to give an argument to cmake?16:08
-!- travis-ci [~travis-ci@ec2-23-23-71-239.compute-1.amazonaws.com] has joined #shogun16:08
travis-ci[travis-ci] it's Fernando Iglesias'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/1062292016:08
-!- travis-ci [~travis-ci@ec2-23-23-71-239.compute-1.amazonaws.com] has left #shogun []16:08
lisitsyniglesiasg: hahah yes it takes a half of under den linden16:08
@iglesiasgHeikoS1, in that case, have you tried taking a look using ccmake?16:08
HeikoS1so I told cmake to bundle json16:08
HeikoS1and now I have a little .cpp file that I would like to compile against shogun with json16:08
HeikoS1so the unit tests work with it16:09
HeikoS1but I would like to compile my own program16:10
@iglesiasg-l shogun?16:10
@iglesiasgmaybe you need to include a directory using -L16:10
@iglesiasgif I remember correctly I needed to use -L from the moment we switched to cmake16:10
shogun-buildbotbuild #1612 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/1612  blamelist: Fernando Iglesias <fernando.iglesiasg@gmail.com>16:10
HeikoS1so first weird thing is that HAVE_JSON is 016:11
HeikoS1so I manually set it to 116:11
HeikoS1then I added -I path/to/json16:11
HeikoS1and -ljson16:11
HeikoS1how do I again tell the compiler where the lib is?16:11
HeikoS1l L ?16:11
HeikoS1-L?16:11
@iglesiasg-L for the path to the dir where the lib is, yes16:11
@iglesiasgand then -l for the lib16:11
HeikoS1but it is in a std place16:12
HeikoS1:/16:12
@iglesiasgHeikoS1, but if HAVE_JSON is 0, it sounds like shogun didn't compile JSON stuff16:12
HeikoS1mmh16:12
HeikoS1lets see16:12
HeikoS1iglesiasg: the unit tests with json run16:12
@iglesiasgmmmm I think that is a bit weird :)16:13
@iglesiasgare they guarded with HAVE_JSON?16:13
HeikoS1maybe ill use my local json16:14
HeikoS1since i have it installed16:14
HeikoS1so recompiling16:14
shogun-buildbotbuild #1318 of cyg1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/1318  blamelist: Fernando Iglesias <fernando.iglesiasg@gmail.com>16:21
HeikoS1wiking: is cmake done with json?16:22
HeikoS1seems that there are some weird things going on16:22
HeikoS1mmh on the other hand the unit tests are running ...16:23
HeikoS1ah confused16:23
-!- foulwall [~user@2001:da8:215:c252:39e3:9dc6:650c:106a] has joined #shogun16:26
foulwallls16:27
-!- lisitsyn [~lisitsin@mxs.kg.ru] has quit [Quit: Leaving.]16:31
shogun-buildbotbuild #1501 of deb2 - static_interfaces is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/1501  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>, Roman Votyakov <votjakovr@gmail.com>16:43
shogun-buildbotbuild #1614 of deb3 - modular_interfaces is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/1614  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>, Roman Votyakov <votjakovr@gmail.com>16:44
shogun-buildbotbuild #1613 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/161316:47
shogun-buildbotbuild #1531 of bsd1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/1531  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>, Roman Votyakov <votjakovr@gmail.com>16:48
votjakovrHeikoS1: how accurately should be an estimate value of marginal likelihood with num_importance_samples=10000? i have the difference less than 1e-3 for EP. Is it ok?16:53
-!- thoralf [~thoralf@enki.zib.de] has quit [Quit: Konversation terminated!]16:55
votjakovrHeikoS1: in contrast difference for Laplace approximation is something about 1e-116:58
votjakovrHeikoS1: i played a little bit with your InferenceMethod.get_log_ml_estimate_binary_logit_laplace unit test :)16:59
shogun-buildbotbuild #1319 of cyg1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/1319  blamelist: Fernando Iglesias <fernando.iglesiasg@gmail.com>17:04
HeikoS1votjakovr: hi sorry I was out for some food17:05
HeikoS1votjakovr: the EP should converge faster than the laplace17:05
HeikoS1votjakovr: the number of samples one needs to get an accurate estimate depends on the dimension of the GP17:06
HeikoS1so for more data points, once needs a larger number17:06
HeikoS1votjakovr: its good to hear that the EP does a better job, this means less samples needed17:06
HeikoS1although the sampling should not be too expensive, if you look at the code, youll see it17:07
HeikoS1computing the approximation is more expensive17:07
HeikoS1votjakovr: btw17:08
HeikoS1there is *no* way of computing the exact marginal likelihood17:08
HeikoS1so no comparison17:08
HeikoS1you are comparing to an approximation17:08
votjakovrHeikoS1: yeah17:08
HeikoS1the sampling is the only way to get an exact estimate17:09
HeikoS1the cool thing it allows to sample from the marginal posterior over the hyperparameters17:09
HeikoS1which is useful for model-selection the Bayesian way (not single point estimates but averaging over distributions)17:09
votjakovrHeikoS1: yep, i know :)17:10
HeikoS1votjakovr: if you are interested, we can add methods for that to shogun later this year17:11
votjakovrHeikoS1: something like mcmc?17:12
HeikoS1votjakovr: well that is a lot of work (though I am also planning doing that=17:12
HeikoS1but for now I meant just given one has samples, average predictions over them17:12
HeikoS1so monte carlo integration17:13
HeikoS1then, sampling can work with that17:13
HeikoS1mcmc is a big thing with many options17:13
HeikoS1so not that easy to offer black-box methods there17:13
HeikoS1but it is possible17:13
HeikoS1mcmc might be a gsoc project next year17:14
votjakovrHeikoS1: cool :)17:15
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has quit [Quit: WeeChat 0.4.0]17:16
-!- pickle27 [~kevin@d67-193-243-174.home3.cgocable.net] has joined #shogun17:31
shogun-buildbotbuild #1320 of cyg1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/1320  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>, Roman Votyakov <votjakovr@gmail.com>17:46
-!- travis-ci [~travis-ci@ec2-174-129-166-170.compute-1.amazonaws.com] has joined #shogun17:48
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/1062421817:48
-!- travis-ci [~travis-ci@ec2-174-129-166-170.compute-1.amazonaws.com] has left #shogun []17:48
-!- foulwall` [~user@2001:da8:215:c252:3025:2d13:68db:4199] has joined #shogun17:55
lambdayHeikoS1: there?17:57
-!- foulwall [~user@2001:da8:215:c252:39e3:9dc6:650c:106a] has quit [Ping timeout: 264 seconds]17:57
HeikoS1lambday: yes, hi!17:58
lambdayjust got your mail17:58
lambdayumm...so far I remember, it was an warning, right?17:58
lambdayor I put error :-/17:58
HeikoS1lambday: ah yes, I am currently trying to fix json serialization tests17:58
lambdayHeikoS1: I am not too sure but I think we can't serialize for complex with json...had some trouble17:59
HeikoS1lambday: cant we just make two floats?17:59
HeikoS1I remember now :)17:59
lambdayyes17:59
lambdaybut it was left that time...17:59
HeikoS1I see17:59
lambdayas in, if someone needs this, then only we'll add support18:00
HeikoS1ok dont worry too much, just keep in mind to fix it18:00
HeikoS1it breaks serialization tests18:00
lambdayand now is the time18:00
HeikoS1yep :D18:00
lambdayalright18:00
lambday:D18:00
lambdayoh and matrix power's sparsity pattern works fine18:00
-!- van51 [~van51@ppp-94-66-12-14.home.otenet.gr] has joined #shogun18:00
lambdayjust added a unit test18:00
lambdaynow finishing the probing sampler unit test18:01
HeikoS1hey, cool!18:01
lambdaythen will send PR18:01
HeikoS1cool, how long does the graph colouring take?18:01
lambdayhaven't tested that yet :-/18:01
lambdayonly testing with small ones till now18:02
lambdaywith them its pretty fast18:02
HeikoS1ok18:02
HeikoS1really curious18:02
HeikoS1for the ozone matrix18:02
lambdayyes once I add the unit-test then will try with the ozone one18:03
HeikoS1I am excited :)18:03
lambdayHeikoS1: so, there are these distance-1 distance-2 colorings and various coloring methods, I am setting them as params18:04
lambdaydon't wanna make it fixed18:04
shogun-notifier-shogun: Heiko Strathmann :feature/SerialUTests * e48b2b4 / tests/unit/io/ (2 files): https://github.com/shogun-toolbox/shogun/commit/e48b2b43c7667f12e6538e9808e21aaeeeb6c8fd18:05
shogun-notifier-shogun: added accuracy for template classes18:05
shogun-notifier-shogun: Heiko Strathmann :feature/SerialUTests * 0610fb9 / tests/unit/io/SerializationJSON_unittest.cc.jinja2: https://github.com/shogun-toolbox/shogun/commit/0610fb952688155bcf938f386c45ac693e175ac018:05
shogun-notifier-shogun: json seems to have a low accuracy18:05
shogun-notifier-shogun: Heiko Strathmann :feature/SerialUTests * 92f01a9 / tests/unit/CMakeLists.txt: https://github.com/shogun-toolbox/shogun/commit/92f01a96b686120c2c158928b20326101df07e8f18:05
shogun-notifier-shogun: activate json since most things are working18:05
shogun-notifier-shogun: Heiko Strathmann :feature/SerialUTests * 3602ec7 / src/shogun/io/SerializableJsonFile.cpp: https://github.com/shogun-toolbox/shogun/commit/3602ec74749d86ec71d6af711a969d34b71f16c218:05
shogun-notifier-shogun: replace warning by error since otherwise shogun runs into uninitialized memory problems18:05
HeikoS1lambday: thats good18:06
HeikoS1lambday: we can experiment with that18:06
HeikoS1we basically just need *a* colouring18:06
HeikoS1in my experiments I used to compute it once and then even fix, even if the matrix changed18:06
lambdayalright18:07
lambdaywe'll have default values for them, but keeping the option to experiment later on might be good :)18:07
HeikoS1lambday: totally, we can experiment ourselves and then change the default values18:20
@wikingHeikoS1: i would ditch the json serialization18:29
HeikoS1wiking: why?18:29
@wikingHeikoS1: and go with msgpack or avro18:29
HeikoS1wiking: btw accuracy is 1e-618:29
@wikingHeikoS1: because that's some homemade hack18:29
@wikingHeikoS1: instead of using some standard stuf18:29
@wikingf18:29
HeikoS1wiking: up for that18:29
HeikoS1sonney2k wrote that not so long ago18:30
HeikoS1wiking: feel free to remove I dont object18:30
HeikoS1wiking: all other things are working now18:30
HeikoS1so branch ready to merge once this is done18:30
HeikoS1wiking: json is used by many languages though18:31
HeikoS1ipython for example18:31
@wikingHeikoS1: yes18:31
HeikoS1but yeah as said, I am happy to drop things18:31
@wikingHeikoS1: but it cannot be used as it is now18:31
@wikingin any other parts18:31
@wikingbecause one would need to know how to deserialize18:32
HeikoS1wiking: really? that sounds really nice18:32
HeikoS1haha,18:32
HeikoS1so then lets delete it18:32
@wikingi mean json is just a syntax18:32
@wikingone needs to know how to deserialize it18:32
@wikingand at least msgpack and avro is out there already for a while18:32
@wikingbeen used for a longer time now18:32
HeikoS1well for that we need to write another class18:32
@wikinghence we really dont need to test it if it works at all18:32
@wikingHeikoS1: as said in my comments18:33
@wikingi've already started msgpack18:33
HeikoS1wiking: ah well ok18:33
HeikoS1then go18:33
HeikoS1wiking: could you then delete the old things and update our feature branch and merge it?18:33
HeikoS1since rest works18:33
HeikoS1serialization though leaks memory sometimes18:33
HeikoS1json also does :)18:33
@wikingHeikoS1: hehe ok i'll check on it later today18:34
@wikingi gotta fix some shit still18:34
HeikoS1wiking: pretty cool with the automated tests now18:34
HeikoS1was a lot work work but I think its worth it18:34
@wikingyeah i mean serialization is necessary18:34
HeikoS1fixed a lot of bugs which would have scared of people18:34
HeikoS1wiking: what do you think about different versions of shogun files?18:35
HeikoS1should we support that?18:35
HeikoS1or only "latest" version18:35
@wikingHeikoS1: well... yeah that's a good question18:35
@wikingbut yeah let's first have one stable version ;)18:35
HeikoS1wiking: the migration framework in principle is able to translate older versions18:35
HeikoS1but its a horror to maintain, the code is hard to understand etc etc18:35
HeikoS1I would just go for "not supported" in this case18:36
HeikoS1and tell users to save their scripts so that they can re-generate their instances18:36
HeikoS1if we had more people that would be feasible18:36
HeikoS1I worked on that for a lot of hours and it doesnt even work now18:36
HeikoS1its hard to use, everyon ignores it etc18:36
HeikoS1so rather drop18:36
@wikingheheh indeed :)18:37
HeikoS1ok thats next step after your msgpack is there18:37
HeikoS1ok cool, I will stop shogunning now and do something else, let me know once the serailU brauch is merged18:37
@wikingHeikoS1: will do18:37
-!- foulwall` [~user@2001:da8:215:c252:3025:2d13:68db:4199] has quit [Ping timeout: 264 seconds]18:56
-!- sonne|osx [~sonne@89.204.135.161] has joined #shogun19:06
-!- lisitsyn [~lisitsyn@fb2-lo1.global63.net] has joined #shogun19:10
-!- mode/#shogun [+o lisitsyn] by ChanServ19:10
sonne|osxiglesiasg: so the travis stuff works - hmhh weird but excellent :)19:11
@iglesiasgsonne|osx, I just removed the second ppa, so it is basically as you left it before that last commit19:11
@iglesiasgas you said, the sudo sudo thing didn't trigger the problem19:11
sonne|osxiglesiasg: yeah but it wasn't building before19:12
sonne|osxno idea19:12
@iglesiasgmmm I think it was, let me show you19:12
@iglesiasgsonne|osx, https://travis-ci.org/shogun-toolbox/shogun/builds19:12
@iglesiasgafter the answer "yes" it was all good19:13
-!- travis-ci [~travis-ci@ec2-174-129-166-170.compute-1.amazonaws.com] has joined #shogun19:14
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/1062996019:14
-!- travis-ci [~travis-ci@ec2-174-129-166-170.compute-1.amazonaws.com] has left #shogun []19:14
@iglesiasgbut now we are bad again xD19:15
@iglesiasgthere is something wrong factor graph19:15
@iglesiasglet's see if hushell comes and can fix it some time soon19:15
shogun-notifier-shogun: Soeren Sonnenburg :develop * 64b8e3e / CMakeLists.txt: https://github.com/shogun-toolbox/shogun/commit/64b8e3e9152179404ff2dea48799636807c421b919:24
shogun-notifier-shogun: improve cmake summary19:24
sonne|osxiglesiasg: ohh well :)19:25
-!- sonne|osx [~sonne@89.204.135.161] has quit [Quit: sonne|osx]19:27
shogun-buildbotbuild #1502 of deb2 - static_interfaces is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/1502  blamelist: Soeren Sonnenburg <sonne@debian.org>19:28
-!- foulwall` [~user@2001:da8:215:c252:ad2d:8912:7f6d:5445] has joined #shogun19:30
shogun-buildbotbuild #1532 of bsd1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/1532  blamelist: Soeren Sonnenburg <sonne@debian.org>19:33
lambdaysorry I was away :(19:40
shogun-notifier-shogun-data: van51 :master * f3c9732 / lang_detection/lang_dataset.dat.xz,lang_detection/lang_label.txt.xz: https://github.com/shogun-toolbox/shogun-data/commit/f3c9732a63e57a60e525d43268e1d5a6f8f874e619:49
shogun-notifier-shogun-data: Added lang_detection data19:49
shogun-notifier-shogun-data: Soeren Sonnenburg :master * a48aa9c / lang_detection/lang_dataset.dat.xz,lang_detection/lang_label.txt.xz: https://github.com/shogun-toolbox/shogun-data/commit/a48aa9c34b3661a1a302dea9230feac62617ce1319:49
shogun-notifier-shogun-data: Merge pull request #13 from van51/develop19:49
shogun-notifier-shogun-data:19:49
shogun-notifier-shogun-data: Added lang_detection data19:49
shogun-notifier-shogun: Kevin :develop * cc30a5a / src/interfaces/octave_modular/swig_typemaps.i: https://github.com/shogun-toolbox/shogun/commit/cc30a5ac885fdd52ffc6228ed03fda47ea40a71919:51
shogun-notifier-shogun: SGNDArray octave in typemap fix19:51
shogun-notifier-shogun: Soeren Sonnenburg :develop * 2611cb1 / src/interfaces/octave_modular/swig_typemaps.i: https://github.com/shogun-toolbox/shogun/commit/2611cb1d63cc417917e81db4a75344c667ba1a7519:51
shogun-notifier-shogun: Merge pull request #1465 from pickle27/develop19:51
shogun-notifier-shogun:19:51
shogun-notifier-shogun: SGNDArray octave in typemap fix19:51
-!- iglesiasg [~iglesias@2001:6b0:1:1da0:19a:206d:eb6b:cdad] has quit [Quit: Ex-Chat]19:56
-!- foulwall` [~user@2001:da8:215:c252:ad2d:8912:7f6d:5445] has quit [Ping timeout: 245 seconds]19:58
@sonney2kvan51, once you have cleaned up your stuff you should talk to foulwall about how to do that in a demo :)19:59
van51sonney2k: sure :)19:59
van51sonney2k: I tried to start it on my own, but I got lost a bit19:59
-!- HeikoS1 [~heiko@nat-188-24.internal.eduroam.ucl.ac.uk] has quit [Quit: Leaving.]20:01
shogun-buildbotbuild #1321 of cyg1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/1321  blamelist: Soeren Sonnenburg <sonne@debian.org>20:12
shogun-buildbotbuild #1615 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/161520:15
van51sonney2k: I 've made a PR about random fourier feats20:16
van51sonney2k: start looking into it when you have time!20:16
@wikingHeikoS: gsoc meeting or shogun meeting?20:23
@wikingah gsoc shogun20:23
@wikingok20:23
@lisitsynwiking: wedding again? ;)20:29
@wikingyeah20:30
@wikingnot mine :)20:30
shogun-buildbotbuild #1616 of deb3 - modular_interfaces is complete: Failure [failed configure]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/1616  blamelist: Soeren Sonnenburg <sonne@debian.org>, Kevin <kevinhughes27@gmail.com>20:42
shogun-buildbotbuild #1503 of deb2 - static_interfaces is complete: Success [build successful]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/150320:45
shogun-buildbotbuild #1533 of bsd1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/1533  blamelist: Soeren Sonnenburg <sonne@debian.org>, Kevin <kevinhughes27@gmail.com>20:47
-!- thoralf [~thoralf@91-64-2-67-dynip.superkabel.de] has joined #shogun20:54
thoralfHeyho.20:55
-!- van51 [~van51@ppp-94-66-12-14.home.otenet.gr] has quit [Quit: Leaving.]21:07
@sonney2kwiking, you serious?21:13
@sonney2kwiking, yours?21:13
@sonney2kahh ok good :)21:13
@lisitsynsonney2k: he had his in 2012 ;)21:17
shogun-notifier-shogun: Thoralf Klein :develop * 6112665 / tests/unit/lib/SGSparseVector_unittest.cc: https://github.com/shogun-toolbox/shogun/commit/61126659534bf8c185f318b0c8cfc74b0d55510421:17
shogun-notifier-shogun: Unit test for future changes on SGSparseVector.21:17
shogun-notifier-shogun: Thoralf Klein :develop * b1e20ea / src/shogun/ (4 files): https://github.com/shogun-toolbox/shogun/commit/b1e20ea6253a02b35653dd6d2fc829ac6f0bbb1f21:17
shogun-notifier-shogun: Avoid copy/malloc in SGSparseVector::sort_feature() :21:17
shogun-notifier-shogun: * changed qsort_index to sort inline21:17
shogun-notifier-shogun: * changed compression to work in line (and filter zero-entries)21:17
shogun-notifier-shogun: * added switch "stable_pointers" to prevent re-alloc21:17
shogun-notifier-shogun: * added unit-tests for new behaviour (in previous commit)21:17
shogun-notifier-shogun: Soeren Sonnenburg :develop * 41d7a03 / / (5 files): https://github.com/shogun-toolbox/shogun/commit/41d7a0319758b64a507eb011e778c49c7f861ec621:17
shogun-notifier-shogun: Merge pull request #1467 from tklein23/sort_features_refactoring21:17
shogun-notifier-shogun:21:17
shogun-notifier-shogun: Last bit of sort_features() refactoring21:17
@sonney2klisitsyn, yeah that is why :D21:18
@sonney2klisitsyn, was thinking he is a serial marriage type of guy21:18
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun21:19
-!- mode/#shogun [+o iglesiasg] by ChanServ21:19
shogun-buildbotbuild #1898 of deb1 - libshogun is complete: Failure [failed compile]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/1898  blamelist: Thoralf Klein <thoralf.klein@zib.de>21:22
@wikingsonney2k: heheheh was thinking? :))21:23
* wiking has a deadline :S21:26
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has quit [Read error: Connection reset by peer]21:26
shogun-buildbotbuild #1322 of cyg1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/1322  blamelist: Soeren Sonnenburg <sonne@debian.org>, Kevin <kevinhughes27@gmail.com>21:26
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun21:27
-!- mode/#shogun [+o iglesiasg] by ChanServ21:27
@wikingiglesiasg lisitsyn  HeikoS up for tomorrow's early meeting? :)21:27
@lisitsynearly is?21:27
@wikingmmmm check doodle21:27
@lisitsyn12 UTC?21:28
@wiking10-2 CET21:28
@iglesiasgwiking, was going to check right now21:28
@lisitsynwaaat?21:28
@iglesiasgbut yeah, sure21:28
@wikinglisitsyn: not that long21:28
@wikinglisitsyn: anytime between that21:28
@lisitsynwhat is CET21:28
@lisitsyn;D21:28
@wikingcentral european time21:28
@wikingGMT+2 atm21:28
@lisitsynGMT+1?21:28
@wikingi thnk21:28
@lisitsynagh shit21:28
@wikingmmm CEST21:28
@wikingit's now powersaving mode21:29
@wikingso i guess it's GMT+221:29
@lisitsynI hate time zones21:29
@lisitsynhate hate21:29
@wikinglisitsyn: UTC/GMT +2 hours21:29
@lisitsynso it should be 12-14 for me21:29
@wikingshould be 12-4 yeah21:30
@wikinganytime between that21:30
@lisitsynokay I am ok21:30
@wikingput on doodle which interval is good for ya21:30
@lisitsynlet me put it21:30
@sonney2klisitsyn, GMT+2 I know that for sure. I only ever work w/ UTC21:30
@lisitsynsonney2k: I jump from gmt+4 to gmt+5 back and forth all the time21:31
@lisitsyndue to mind disabled government21:31
@lisitsyn:D21:31
@lisitsynso this makes me hate it more and more21:32
-!- hushell [~hushell@c-98-232-178-161.hsd1.or.comcast.net] has joined #shogun21:32
@wikingsonney2k: server status? comment on msgpack? :D21:33
@sonney2kwho is batman?21:35
@sonney2kwiking, msgpack?21:35
@sonney2kwiking, server still no news.. waiting for action21:35
-!- iglesiasg is now known as batman_21:37
@sonney2kwiking, messagepack? sure feel free to write another reader/writer. All you have to do is supply a SerializableReader/Writer21:39
@wikingsonney2k: i know ... i was just thnking that since we have some problems with json itself... why not switch to some more standard lib...21:41
@wikingand drop the native json serialization21:41
@sonney2kwiking, well we are not forced to use json and it is good to have21:41
@sonney2kwiking, well json is a standard many people use21:42
@sonney2kyes it has flaws21:42
@sonney2kbut java/.js web-people use it21:42
@wikingyeah but that's why i'm saying that let's use a lib for that21:43
@wikinglike msgpack21:43
@wikingor avro21:43
-!- batman_ is now known as iglesiasg21:45
@sonney2kwiking, not sure avro would work? Doesn't it need static schemas?21:47
@wikingsonney2k: autogen them? :)21:49
shogun-buildbotbuild #1899 of deb1 - libshogun is complete: Failure [failed compile]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/1899  blamelist: Soeren Sonnenburg <sonne@debian.org>, Thoralf Klein <thoralf.klein@zib.de>21:58
thoralfsonney2k: Damn.  I'm already using c++11 in my tests, but this buildbot isn't.21:59
thoralfsonney2k: How do you handle this usually?  ifndef xxx?21:59
@sonney2kthoralf, yes22:03
-!- lambday [67157f4c@gateway/web/freenode/ip.103.21.127.76] has quit [Ping timeout: 250 seconds]22:04
hushellDid you get this error? tests/unit/lib/SGSparseVector_unittest.cc:215:73: warning: '>>' operator will be treated as two right angle brackets in C++0x [-Wc++0x-compat]22:08
thoralfhushell: Its my fault.22:09
thoralfhushell: Your compiler does not know c++09/c++11?22:09
hushellthoralf: I think it should support, I am using g++4.622:10
hushellthoralf: I haven't fetch for few days. It works fine before22:11
thoralfhushell: Do a checkout of something before the last merge. ;)22:12
thoralfsonney2k, wiking: How can I disable c++11 while building?  I'd like to test my changes locally...22:16
@wikingthoralf: :D22:21
hushellthoralf: Maybe you edit SET(HAVE_CXX11 0) in CMakeLists.txt22:21
@sonney2kthoralf, ask wiking he is the god of hell fire22:25
thoralfhushell: Hmm... but then I also would need to remove the appended options from CXX_FLAGS.22:25
* thoralf removed the whole block from CMakeLists.txt22:26
thoralfShould work for now. :D22:26
@sonney2kthoralf, heh22:26
@sonney2kI totally don't get why some stuff in cmake cannot be -DTURNED_OFF=TRUE!22:27
thoralfsonney2k: god of hell fire?  sounds like Prodigy or sth similar?22:27
@lisitsynF-G INSTIGATOR22:30
@lisitsynI AM FIRESTARTER22:30
@lisitsynthoralf: the songs of my childhood!22:32
thoralfProdigy: Fire/jericho: "I am the god of Hell fire, and I bring you (fire)" :D22:33
@sonney2kthoralf, can you fix the test quickly or shall I?22:33
thoralfsonney2k: I'm on it.  Will take a big unless I fix it by nesting two tests with #ifndef ...22:34
thoralfs/bit/bit/22:34
@sonney2kthoralf, well you can just do > > (space!) and all good22:36
@sonney2kno need to ifdef22:37
thoralfsonney2k: No, it's also about the use of auto and "extended initializers".22:37
thoralfsonney2k: I tried insering spaces and it raised many new errors. ;)22:38
@iglesiasgpickle27, did you find many attacks to the English language in the notebook? :)22:38
pickle27iglesiasg: I don't think so, I read mainly for content, do you want me to read it again for language?22:39
@sonney2kthoralf, you!22:39
@sonney2khmmhh22:39
@sonney2kthe cygwin stuff runs fine from cmdline22:39
@sonney2kno idea why it crashes w/ the buildbot22:39
@sonney2kwiking, running the installer for visual studio express on the bb now22:43
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has quit [Ping timeout: 268 seconds]22:43
@sonney2kwhy does this need 5GB?22:45
thoralfsonney2k: https://github.com/shogun-toolbox/shogun/pull/1468 - but we should wait for travis.  I removed the tests, but I'm working on a real fix.22:45
* sonney2k is going to regret this22:45
* thoralf fell in love with c++09/11's "auto" type declaration - and now has to remove it.22:54
thoralfWhat a pity!22:54
-!- travis-ci [~travis-ci@ec2-107-22-100-33.compute-1.amazonaws.com] has joined #shogun23:03
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/1063628123:03
-!- travis-ci [~travis-ci@ec2-107-22-100-33.compute-1.amazonaws.com] has left #shogun []23:03
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has joined #shogun23:13
@sonney2kmicrosoft visual studio express is loading user settings ... this may take a few minutes23:18
@sonney2kexpress!23:18
@sonney2kman once installed it wants to update itself23:19
@sonney2kbrilliant technology23:19
@sonney2kwiking, now this crap is installed - so now what?23:19
@lisitsynhahahaha23:21
@lisitsynsonney2k: I guess you are cmaking stuff - just generate cmake project23:23
@lisitsynerr23:23
@lisitsynVS project I mean23:23
@HeikoSwiking, around23:27
@HeikoSwiking, I can do until 1130 cest tomorrow23:28
@HeikoSwiking, so 10 start would be good for me. 9 is too early23:28
@HeikoS1030 even better23:29
@HeikoSSo exactly the hour in between23:30
@HeikoSThe two existing ones23:30
shogun-notifier-shogun: Thoralf Klein :develop * a01097d / tests/unit/lib/SGSparseVector_unittest.cc: https://github.com/shogun-toolbox/shogun/commit/a01097da3f42b4e2a0f8e8bddc78d14c6b7b3ed023:31
shogun-notifier-shogun: Removed two big unit tests to fix the build.  (Used too many c++09/11 features like "auto" and extended initializers.)23:31
shogun-notifier-shogun: Soeren Sonnenburg :develop * 4e3362c / tests/unit/lib/SGSparseVector_unittest.cc: https://github.com/shogun-toolbox/shogun/commit/4e3362c82d68b9c7fafea6879db2c1a7da9bf63d23:31
shogun-notifier-shogun: Merge pull request #1468 from tklein23/build_hotfix23:31
shogun-notifier-shogun:23:31
shogun-notifier-shogun: Removed two unit tests to fix the build...23:31
shogun-buildbotbuild #1534 of bsd1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/1534  blamelist: Soeren Sonnenburg <sonne@debian.org>23:53
-!- iglesiasg [~Fernando@s83-179-44-135.cust.tele2.se] has joined #shogun23:54
-!- mode/#shogun [+o iglesiasg] by ChanServ23:54
@iglesiasgpickle27: nah, checking for language sounds too boring, I guess it is at least understandable which is the idea :)23:55
--- Log closed Tue Aug 27 00:00:32 2013