| --- Log opened Wed Feb 13 00:00:45 2013 | ||
| -!- SCV__ [~FSCV@173.254.212.46] has quit [Quit: Leaving] | 01:05 | |
| -!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving] | 01:42 | |
| -!- sumit [75e210ac@gateway/web/freenode/ip.117.226.16.172] has joined #shogun | 03:02 | |
| -!- sumit [75e210ac@gateway/web/freenode/ip.117.226.16.172] has quit [Ping timeout: 245 seconds] | 03:11 | |
| -!- sumit [ca4eaca2@gateway/web/freenode/ip.202.78.172.162] has joined #shogun | 04:29 | |
| -!- sumit [ca4eaca2@gateway/web/freenode/ip.202.78.172.162] has quit [Ping timeout: 245 seconds] | 05:21 | |
| -!- sumit [ca4eaca2@gateway/web/freenode/ip.202.78.172.162] has joined #shogun | 05:38 | |
| -!- blackburn [~blackburn@83.234.54.219] has joined #shogun | 06:38 | |
| -!- sumit [ca4eaca2@gateway/web/freenode/ip.202.78.172.162] has quit [Ping timeout: 245 seconds] | 06:49 | |
| -!- blackburn [~blackburn@83.234.54.219] has quit [Quit: Leaving.] | 07:09 | |
| -!- sumit [ca4eaca2@gateway/web/freenode/ip.202.78.172.162] has joined #shogun | 08:11 | |
| -!- lambday [3ba2174d@gateway/web/freenode/ip.59.162.23.77] has joined #shogun | 08:54 | |
| -!- sumit [ca4eaca2@gateway/web/freenode/ip.202.78.172.162] has quit [Ping timeout: 245 seconds] | 09:13 | |
| -!- blackburn [~lisitsin@mxs.kg.ru] has joined #shogun | 09:43 | |
| -!- shogun-notifier- [~irker@7nn.de] has joined #shogun | 09:43 | |
| shogun-notifier- | shogun: iglesias :master * c4443b0 / src/shogun/converter/KernelLocallyLinearEmbedding.cpp: https://github.com/shogun-toolbox/shogun/commit/c4443b05e08274639d5053f4bc5dfdb357b7df71 | 09:43 |
|---|---|---|
| shogun-notifier- | shogun: small fix in KLLE to avoid valgrind errors | 09:43 |
| shogun-notifier- | shogun: iglesias :master * ded1104 / src/shogun/converter/DiffusionMaps.cpp: https://github.com/shogun-toolbox/shogun/commit/ded1104c19f0445fbf33f40f2508145b5606291d | 09:43 |
| shogun-notifier- | shogun: + Gaussian kernel width initialization in DiffusionMaps | 09:43 |
| shogun-notifier- | shogun: iglesias :master * 3f87a2e / src/shogun/converter/NeighborhoodPreservingEmbedding.cpp: https://github.com/shogun-toolbox/shogun/commit/3f87a2e0183aa42396be7994f6bfd7233544223a | 09:43 |
| shogun-notifier- | shogun: * minor fix in NPE to avoid valgrind complains | 09:43 |
| shogun-notifier- | shogun: Sergey Lisitsyn :master * db0a595 / src/shogun/converter/ (3 files): https://github.com/shogun-toolbox/shogun/commit/db0a595a2cc65032be3573b340c9d10548912c55 | 09:43 |
| shogun-notifier- | shogun: Merge pull request #862 from iglesias/tapkee | 09:43 |
| shogun-notifier- | shogun: | 09:43 |
| shogun-notifier- | shogun: [WIP] Small fixes in converters to avoid valgrind errors | 09:43 |
| shogun-buildbot_ | build #805 of deb1 - libshogun is complete: Failure [failed test] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/805 blamelist: iglesias <fernando.iglesiasg@gmail.com> | 09:55 |
| -!- n4nd0 [~nando@n158-p227.kthopen.kth.se] has joined #shogun | 09:56 | |
| blackburn | n4nd0: thanks for issuing a PR | 09:56 |
| blackburn | is everything fixed now? | 09:56 |
| n4nd0 | no, not everything yet | 09:56 |
| n4nd0 | I am working in the LLTSA errors now | 09:56 |
| n4nd0 | I tried with the usual path, setting m_nullspace_shift but it didn't solve it | 09:57 |
| blackburn | hmm | 09:57 |
| n4nd0 | I tried also a couple of things more but didn't succeed | 09:57 |
| n4nd0 | I will look to it now :) | 09:57 |
| blackburn | if you have some time for that | 09:58 |
| blackburn | I'd appreciate | 09:58 |
| n4nd0 | mmm SGVectorTest failed | 09:59 |
| n4nd0 | blackburn: btw I was thinking yesterday | 10:02 |
| n4nd0 | it could be good if we think of a small set of tasks | 10:02 |
| n4nd0 | and write them down somewhere | 10:02 |
| blackburn | yeah.. | 10:02 |
| blackburn | n4nd0: github issues | 10:02 |
| n4nd0 | blackburn: yes, it is a good idea to write them there | 10:03 |
| blackburn | n4nd0: I added 'entrance' label | 10:04 |
| blackburn | if you have any ideas please put it | 10:04 |
| n4nd0 | will do, thank you! | 10:04 |
| n4nd0 | blackburn: why is it that lltsa needs the features to be set both in parameters.kernel and parameters.features? | 10:08 |
| n4nd0 | https://github.com/shogun-toolbox/shogun/blob/master/src/shogun/converter/LinearLocalTangentSpaceAlignment.cpp#L41 | 10:08 |
| blackburn | n4nd0: it uses both kernel and feature access callbacks | 10:10 |
| n4nd0 | aham | 10:10 |
| blackburn | n4nd0: so I asked for help in the ML | 10:21 |
| n4nd0 | blackburn: yeah, just read it | 10:21 |
| n4nd0 | that's nice, let's see if we get more suggestions | 10:21 |
| blackburn | n4nd0: 3 already! :) | 10:26 |
| n4nd0 | :) | 10:27 |
| lambday | hello, this is soumyajit. I want to get involved with shogun.. I've already cloned the source code and built it.. (cmdline static and python interfaces).. But I am not sure how to get started with such a huge codebase.. could you please suggest me something? | 10:32 |
| n4nd0 | lambday: you have come in a good moment :) | 10:33 |
| lambday | :) | 10:33 |
| n4nd0 | blackburn has just created an entrance section on github issues | 10:33 |
| n4nd0 | https://github.com/shogun-toolbox/shogun/issues?labels=entrance&page=1&state=open | 10:33 |
| lambday | okay. i'm checking.. | 10:33 |
| n4nd0 | it was announced in the mailing list a few minutes ago too | 10:34 |
| -!- proKrammer [73fe7104@gateway/web/freenode/ip.115.254.113.4] has joined #shogun | 10:34 | |
| lambday | okay.. I'm not yet registered to the mailing list.. will do | 10:35 |
| lambday | and also, is there any kernel related development task currently within the scope? I'm doing some reading on string kernels and I'm really interested to work with one | 10:40 |
| n4nd0 | lambday: you could check if we already have implemented that kernel | 10:40 |
| n4nd0 | if it is not, you are welcome to include it | 10:41 |
| n4nd0 | lambday: see src/shogun/kernel/string | 10:41 |
| lambday | n4nd0: thanks.. I'm checking | 10:42 |
| n4nd0 | what kernel is it btw the one you are interested to work with? | 10:42 |
| lambday | n4nd0: well, I was reading this paper "Fast Kernels for String and Tree matching" by Vishwanathan et al. as a part of my coursework | 10:45 |
| lambday | http://books.nips.cc/papers/files/nips15/AA11.pdf | 10:47 |
| blackburn | I'd recommend Soeren's thesis http://sonnenburgs.de/soeren/publications/Son08.pdf | 10:48 |
| n4nd0 | blackburn: everything in there is probably already in shogun though | 10:50 |
| n4nd0 | I am not sure, just guessing | 10:50 |
| lambday | blackburn: are these the kernels that are already implemented in shogun? | 10:50 |
| blackburn | lambday: I think so | 10:51 |
| lambday | oh I see the subsequence kernel | 10:51 |
| lambday | but it implemented the paper by lodhi.. | 10:51 |
| lambday | blackburn: well, can you suggest me something then, any development related ideas (better if kernel related).. I'll do some reading and will be in touch... | 10:55 |
| lambday | and also, how should I go about understanding the architecture of shogun? | 10:56 |
| blackburn | lambday: we are now collecting ideas https://github.com/shogun-toolbox/shogun/issues?labels=entrance&state=open | 10:56 |
| blackburn | there should be some kernel related for sure - we just started | 10:56 |
| n4nd0 | lambday: get to know the code superficially, browse different classes and so on | 10:57 |
| lambday | blackburn: okay.. thanks :) | 10:57 |
| blackburn | it is a difficult question about architecture - I have nothing general about that | 10:57 |
| lambday | n4nd0: alright.. I'll keep in touch.. meanwhile, I'll go through the thesis by Soeren | 11:00 |
| n4nd0 | wiking: haha ok ok | 11:14 |
| n4nd0 | I don't think that in this case to compile with -lpthread -pthread would make a mistake anyway | 11:15 |
| n4nd0 | wiking: who can take a look to configure script? | 11:15 |
| wiking | n4nd0: i did now... it's a bit problematic | 11:15 |
| wiking | i mean the configure script for -pthread or -lpthread is quite easy | 11:16 |
| wiking | the problem is more about that i actually do not understand | 11:16 |
| wiking | why -pthread is not working for unit test compilation | 11:16 |
| n4nd0 | yeah, I reckon that feeling when looking at configure :) | 11:16 |
| wiking | but works fine for libshogun | 11:16 |
| sonne|work | n4nd0: regarding this list - is LDA multiclass already? | 11:17 |
| n4nd0 | sonne|work: LDA multiclass and change QDA to eigen were in my head | 11:17 |
| n4nd0 | I am going to write them soon | 11:18 |
| wiking | mmm | 11:18 |
| n4nd0 | in the entrance issues I mean | 11:18 |
| sonne|work | n4nd0: yes please thx | 11:18 |
| wiking | n4nd0: what is the exact compilation that does not work for you for unit test? | 11:18 |
| n4nd0 | wiking: give me a couple of minutes to check | 11:18 |
| wiking | n4nd0: ok | 11:19 |
| n4nd0 | sonne|work: did you talk with c-base people? | 11:19 |
| n4nd0 | I don't exactly recall when was the exact | 11:19 |
| n4nd0 | around 15 IIRC | 11:19 |
| n4nd0 | exact date* | 11:19 |
| blackburn | 14 I think | 11:19 |
| blackburn | love meeting | 11:19 |
| blackburn | lol | 11:19 |
| n4nd0 | haha yeah | 11:19 |
| sonne|work | 14th but it looks like I don't have time that evening :/ | 11:20 |
| sonne|work | and gunnar == /dev/null | 11:20 |
| n4nd0 | :( | 11:20 |
| n4nd0 | wiking: https://gist.github.com/iglesias/4943650 | 11:22 |
| wiking | ah shit the linking line is actually not there :( | 11:23 |
| wiking | n4nd0: can u edit the Makefile under tests/unit | 11:24 |
| wiking | and there at line 28 | 11:24 |
| n4nd0 | wiking: sure | 11:25 |
| wiking | remove the @ in front of @$(LINK) | 11:25 |
| n4nd0 | it seems that sonne|work has found an answer though hehe | 11:25 |
| wiking | n4nd0: so just remove that @ | 11:25 |
| wiking | and do make again | 11:25 |
| wiking | and copy paste the linking line | 11:25 |
| wiking | the last c++ line | 11:25 |
| wiking | sonne|work: afaik there's a better fix i just need to see the linking line | 11:26 |
| wiking | i think the order of -pthread matters | 11:26 |
| proKrammer | hi n4nd0 , I have spoken to you yesterday regarding the installation, i am done with it, but what about the GUI part? | 11:26 |
| sonne|work | wiking: true | 11:26 |
| wiking | n4nd0: u have it? | 11:26 |
| wiking | :) | 11:26 |
| n4nd0 | c++ -o shogun-unit-test ./classifier/svm/SVMOcas_unittest.cc.o ./multiclass/MulticlassOCAS_unittest.cc.o ./base/main_unittest.cc.o ./latent/LatentSVM_unittest.cc.o ./lib/SGVector_unittest.cc.o -Xlinker --no-undefined -L../../src/shogun -lshogun libgmock.a | 11:27 |
| n4nd0 | /usr/bin/ld: libgmock.a(gtest-all.o): undefined reference to symbol 'pthread_key_delete@@GLIBC_2.2.5' | 11:27 |
| wiking | ok | 11:27 |
| n4nd0 | /usr/bin/ld: note: 'pthread_key_delete@@GLIBC_2.2.5' is defined in DSO /usr/lib/libpthread.so.0 so try adding it to the linker command line | 11:27 |
| n4nd0 | /usr/lib/libpthread.so.0: could not read symbols: Invalid operation | 11:27 |
| n4nd0 | collect2: error: ld returned 1 exit status | 11:27 |
| n4nd0 | make: *** [shogun-unit-test] Error | 11:27 |
| wiking | mmmm | 11:27 |
| wiking | i wonder why there's no pthread there | 11:27 |
| wiking | ok just a sec | 11:27 |
| wiking | sonne|work: shouldn't $(LINKFLAGS) contain -pthread? | 11:28 |
| wiking | sonne|work: POSTLINKFLAGS | 11:28 |
| wiking | that should be it?! | 11:28 |
| wiking | n4nd0: can u add to that line 28 this variable | 11:28 |
| wiking | $(POSTLINKFLAGS) | 11:28 |
| wiking | $(LINK) -o shogun-unit-test $(TESTOBJFILES) $(LINKFLAGS) -L$(SHOGUNSRCTOP)/shogun -lshogun libgmock.a $(LINKFLAGS_GMOCK) $(POSTLINKFLAGS) | 11:29 |
| wiking | so it should look like something like this ^ | 11:29 |
| n4nd0 | ok | 11:29 |
| wiking | and run make again | 11:29 |
| wiking | and paste here the c++ line | 11:29 |
| n4nd0 | wiking: that made it | 11:30 |
| wiking | ok how does the line look like? | 11:30 |
| n4nd0 | c++ -o shogun-unit-test ./classifier/svm/SVMOcas_unittest.cc.o ./multiclass/MulticlassOCAS_unittest.cc.o ./base/main_unittest.cc.o ./latent/LatentSVM_unittest.cc.o ./lib/SGVector_unittest.cc.o -Xlinker --no-undefined -L../../src/shogun -lshogun libgmock.a -lm -pthread -ljson -lxml2 -llapack -lcblas -larpack -lz -lbz2 -llzma | 11:30 |
| wiking | ok | 11:31 |
| n4nd0 | -pthread is in da house now | 11:31 |
| wiking | and it did not complain? | 11:31 |
| n4nd0 | no | 11:31 |
| wiking | ran fine? | 11:31 |
| n4nd0 | yeah | 11:31 |
| wiking | cool | 11:31 |
| n4nd0 | all tests passed | 11:31 |
| wiking | commit and PR it | 11:31 |
| n4nd0 | just a few warnings | 11:31 |
| wiking | i'll merge it | 11:31 |
| wiking | that's ok that's because of MOCK | 11:31 |
| n4nd0 | should I put back the @ in front of @(LINK)? | 11:31 |
| wiking | yes | 11:31 |
| wiking | plz | 11:31 |
| wiking | :) | 11:31 |
| n4nd0 | compiling again with @ to ensure | 11:32 |
| wiking | yeah that just silences the compile line | 11:32 |
| wiking | so that it doesn't get printed on stdout | 11:32 |
| n4nd0 | aham | 11:32 |
| wiking | but should work fine still | 11:32 |
| n4nd0 | yeah it did | 11:32 |
| wiking | ok commit + pr :> | 11:32 |
| wiking | plz | 11:32 |
| n4nd0 | wiking: the pull request is a bit ugly because of the previous commit and pull request I had issued :S | 11:39 |
| n4nd0 | wiking: take a look | 11:39 |
| n4nd0 | three commits in there | 11:40 |
| wiking | aah yeah ok | 11:40 |
| wiking | i got it why :) | 11:40 |
| n4nd0 | was there any way to fix that | 11:40 |
| n4nd0 | ? | 11:40 |
| wiking | can u do --amend? | 11:40 |
| n4nd0 | I should always work from a branch | 11:40 |
| wiking | git commit --amend . | 11:40 |
| n4nd0 | it was already pushed in my repo, since I did the PR | 11:41 |
| wiking | ah yeah | 11:41 |
| wiking | true | 11:41 |
| n4nd0 | git commit --amend would make a problem? | 11:41 |
| wiking | i think then it would break your branch | 11:41 |
| wiking | :)))) | 11:41 |
| wiking | but that's ok we can merge it like this no sonne|work ? | 11:41 |
| wiking | i mean yeah 3 commits for 1 fix but still it's workng :> | 11:41 |
| shogun-notifier- | shogun: iglesias :master * 021237f / tests/unit/Makefile: https://github.com/shogun-toolbox/shogun/commit/021237f32332ba15b6b0baf337b4ae3e5293c0d4 | 11:42 |
| shogun-notifier- | shogun: add -lpthread flag to build tests correctly | 11:42 |
| shogun-notifier- | shogun: iglesias :master * a954c4f / tests/unit/Makefile: https://github.com/shogun-toolbox/shogun/commit/a954c4f2912d5c736a9c521750e48bf9082f1707 | 11:42 |
| shogun-notifier- | shogun: minor fix: add POSTLINKFLAGS in linking gtest | 11:42 |
| shogun-notifier- | shogun: iglesias :master * c553667 / : https://github.com/shogun-toolbox/shogun/commit/c5536673621c26ebbaa7cdb496eb733a66ce5a16 | 11:42 |
| shogun-notifier- | shogun: Merge branch 'master' of github.com:iglesias/shogun | 11:42 |
| shogun-notifier- | shogun: | 11:42 |
| shogun-notifier- | shogun: Conflicts: | 11:42 |
| wiking | done | 11:42 |
| shogun-notifier- | shogun: tests/unit/Makefile | 11:42 |
| shogun-notifier- | shogun: Viktor Gal :master * da53770 / tests/unit/Makefile: https://github.com/shogun-toolbox/shogun/commit/da53770c936c847376f96449e78f2545fc7269ef | 11:42 |
| shogun-notifier- | shogun: Merge pull request #867 from iglesias/master | 11:42 |
| shogun-notifier- | shogun: | 11:42 |
| shogun-notifier- | shogun: minor fix: add POSTLINKFLAGS in linking gtest | 11:42 |
| sonne|work | why 3 commits not one? | 11:42 |
| n4nd0 | wiking: thank you for finding out the error :) | 11:42 |
| wiking | nw | 11:42 |
| wiking | sonne|work: see ^ | 11:43 |
| n4nd0 | sonne|work: becuase of the previous PR with one commit | 11:43 |
| n4nd0 | I did that from master :S | 11:43 |
| n4nd0 | I shouldn't have done it, my bad | 11:43 |
| sonne|work | n4nd0, wiking now please get shogun in release shape and we are good | 11:43 |
| n4nd0 | sonne|work: I am currently taking a look to converter errors detected by valgrind | 11:44 |
| sonne|work | thanks | 11:44 |
| blackburn | fixing my mistakes hah | 11:47 |
| shogun-buildbot_ | build #807 of deb1 - libshogun is complete: Failure [failed test] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/807 blamelist: iglesias <fernando.iglesiasg@gmail.com> | 11:48 |
| n4nd0 | well you are actually telling me where the errors are most of the time :D | 11:48 |
| wiking | why did u faaaail | 11:48 |
| wiking | ah ok | 11:49 |
| n4nd0 | wiking: I think the test was already broken before | 11:49 |
| wiking | SGVectorTest.misc failed | 11:49 |
| n4nd0 | yeah | 11:49 |
| wiking | ok yeah we need to use EXPECT_DOUBLE_EQ :> | 11:50 |
| blackburn | wiking: if you want to get more tests for shogun and have ideas - please add them to 'entrance' issues at github | 11:50 |
| wiking | blackburn: oh good idea! | 11:51 |
| blackburn | but it would need to describe what exactly to test | 11:54 |
| shogun-buildbot_ | build #808 of deb1 - libshogun is complete: Failure [failed test] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/808 blamelist: Viktor Gal <vigsterkr@gmail.com>, iglesias <fernando.iglesiasg@gmail.com> | 11:54 |
| blackburn | our candidates could get stucked with concepts that are well-known for us | 11:55 |
| blackburn | that's probably normal but we should keep that in mind :) | 11:56 |
| wiking | blackburn: yeah i'll try to make a module list and put them up one by one | 11:56 |
| n4nd0 | blackburn: hey | 12:43 |
| n4nd0 | maybe this makes sense to you | 12:43 |
| n4nd0 | in lltsa to set parameters.eigenshift equal to m_nullspace_shift (default -1e-9) | 12:44 |
| n4nd0 | gives more valgrind errors thatn | 12:44 |
| n4nd0 | to set it to m_reconstruction_shift (default 1e-3) | 12:44 |
| -!- n4nd0 [~nando@n158-p227.kthopen.kth.se] has quit [Ping timeout: 272 seconds] | 13:08 | |
| -!- n4nd0 [~nando@n158-p227.kthopen.kth.se] has joined #shogun | 13:19 | |
| blackburn | n4nd0: that's out of sense for me | 13:24 |
| n4nd0 | too bad | 13:26 |
| n4nd0 | I will look at it again later today or tom | 13:28 |
| blackburn | I hope I can take a look tonight | 13:30 |
| n4nd0 | ok | 13:30 |
| sonne|work | wiking: urgs -have you seen the ocas test failing? | 13:51 |
| wiking | sonne|work: noup | 13:54 |
| wiking | sonne|work: where? | 13:54 |
| sonne|work | http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/808/steps/test/logs/stdio | 13:55 |
| sonne|work | wiking: don't you get the buildbot emails ? | 13:55 |
| -!- proKrammer [73fe7104@gateway/web/freenode/ip.115.254.113.4] has quit [Quit: Page closed] | 13:55 | |
| wiking | i do but not always i check them ;P | 13:55 |
| wiking | uff | 13:55 |
| wiking | sonne|work: i reckon there's some problem there with an SGVector function | 13:56 |
| wiking | sonne|work: since we had the exact same errors when the ::max was bogus | 13:56 |
| sonne|work | wiking: no idea but please check. if we don't address the issues tests are useless... | 13:58 |
| wiking | sonne|work: working on it | 13:59 |
| blackburn | 7 entrance tasks so far! | 14:02 |
| wiking | sonne|work: btw what is the proper method of right away applying patches of mine into shogun? or do i should go still with the commit local repot -> pr -> merge ? | 14:02 |
| n4nd0 | blackburn: we need now 7 students! hehe | 14:03 |
| wiking | blackburn: ^ | 14:03 |
| blackburn | wiking: my position is that for small changes/bugfixes it is ok to commit directly | 14:03 |
| wiking | blackburn: how? | 14:03 |
| blackburn | wiking: hmm you can't? | 14:04 |
| wiking | blackburn: no i dont know how :) | 14:04 |
| blackburn | wiking: I think you need to change remote | 14:04 |
| blackburn | to git@github.com:shogun-toolbox/shogun.git | 14:05 |
| sonne|work | wiking: yeah that is best, but merge yourself | 14:05 |
| blackburn | you probably have git://github.com/shogun-toolbox/shogun.git | 14:05 |
| sonne|work | wiking: I will switch to doing that too | 14:05 |
| blackburn | as a remote | 14:05 |
| sonne|work | don't directly work on shogun master | 14:05 |
| blackburn | I do not think we should do PRs for every change we make | 14:06 |
| sonne|work | blackburn: and I do :) | 14:06 |
| wiking | mm this is interesting | 14:06 |
| wiking | http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/807/steps/test/logs/stdio | 14:06 |
| wiking | sonne|work: any ideas ^ | 14:06 |
| sonne|work | but we had that discussion already. | 14:06 |
| wiking | we use EXPECT_DOUBLE_EQ in both cases | 14:06 |
| wiking | but still it complains | 14:06 |
| blackburn | sonne|work: you know once I get POWER and AUTHORITY.. | 14:07 |
| sonne|work | wiking: then maybe we need a slightly higher epsilon | 14:07 |
| sonne|work | float stuff is certainly scary | 14:08 |
| sonne|work | blackburn: you will I am sure but it is better that everyone's damage to the code might be looked at | 14:09 |
| blackburn | sonne|work: well we actually should not work in master branch | 14:10 |
| blackburn | I do not feel the difference between one commit + PR and one commit | 14:10 |
| blackburn | both can be looked at | 14:10 |
| blackburn | and the damage is done | 14:10 |
| sonne|work | blackburn: yeah dev branches might be a good idea | 14:11 |
| sonne|work | once we have a release we should attempt to switch | 14:11 |
| wiking | lolface | 14:19 |
| wiking | fucking google test blog answers the same shit | 14:19 |
| n4nd0 | wiking: link? | 14:19 |
| wiking | http://googletesting.blogspot.hu/2008/10/tott-floating-point-comparison.html | 14:19 |
| wiking | sonne|work: can u enable unit testing on freebsd bot plz? | 14:21 |
| wiking | sonne|work: ah ok it is | 14:21 |
| wiking | sorry | 14:21 |
| sonne|work | wiking: so why not use say 1e-12 as precision? | 14:24 |
| wiking | mmmm | 14:26 |
| wiking | sonne|work: no i was wrong | 14:26 |
| wiking | what is interesting though | 14:26 |
| wiking | that the same bot | 14:26 |
| wiking | has some weird things | 14:26 |
| wiking | http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/807/steps/test/logs/stdio | 14:26 |
| wiking | so there the SGVector.misc fails | 14:27 |
| wiking | but then in the next build | 14:27 |
| wiking | http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/808/steps/test/logs/stdio | 14:27 |
| wiking | it is all good | 14:27 |
| wiking | but the multiocas fails | 14:27 |
| wiking | can we force a new build on this bot | 14:27 |
| wiking | just out of curiosity | 14:27 |
| sonne|work | wiking: sure just do it | 14:28 |
| wiking | sonne|work: what is the command? :D | 14:28 |
| sonne|work | wiking: I guess it is due to parallelization... | 14:28 |
| wiking | shogun-buildbot_: force build deb1 | 14:29 |
| sonne|work | shogun-buildbot_: force build "deb1 - libshogun" | 14:29 |
| shogun-buildbot_ | no such builder 'deb1' | 14:29 |
| shogun-buildbot_ | build forced [ETA 1m27s] | 14:29 |
| shogun-buildbot_ | I'll give a shout when the build finishes | 14:29 |
| wiking | almost :) | 14:29 |
| shogun-buildbot_ | build #809 of deb1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/809 | 14:29 |
| wiking | hohoho | 14:30 |
| wiking | this is mega insteresting | 14:30 |
| wiking | as now it succeeded | 14:31 |
| sonne|work | ^ | 14:32 |
| sonne|work | (02:28:20 PM) sonne|work: wiking: I guess it is due to parallelization... | 14:32 |
| wiking | yeps seems so | 14:32 |
| -!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout] | 14:42 | |
| -!- FSCV [~FSCV@204.45.110.163] has joined #shogun | 14:53 | |
| sonne|work | wiking: so then just limit it to use a single core | 15:31 |
| n4nd0 | sonne|work, wiking : why does parallelization make it fail? | 15:36 |
| sonne|work | n4nd0: not sure if it fails or just finds some other valid solution (given objective w/i epsilon) | 15:37 |
| wiking | lol now i have a failure with: | 15:44 |
| wiking | Value of: SGVector<float64_t>::arg_max_abs(a.vector, 1, a.vlen, __null) Actual: 0 | 15:44 |
| wiking | Expected: arg_max_abs | 15:44 |
| wiking | Which is: 9 | 15:44 |
| wiking | same with max_abs | 15:44 |
| sonne|work | wiking: valgrind.... | 15:46 |
| wiking | ok i see why | 15:46 |
| wiking | :))0 | 15:46 |
| wiking | it's an actual bug indeed | 15:47 |
| wiking | btw where's github bot? | 15:50 |
| -!- shogun-notifier- [~irker@7nn.de] has joined #shogun | 15:51 | |
| shogun-notifier- | shogun: Viktor Gal :master * 59e15f5 / tests/unit/ (2 files): https://github.com/shogun-toolbox/shogun/commit/59e15f5087e05618b4f29faa533b3d8002571ec0 | 15:51 |
| shogun-notifier- | shogun: Fix unit tests for SGVector::max_abs and MOCASSVM | 15:51 |
| shogun-notifier- | shogun: there was a bug when calculating max_abs in an SGVector | 15:51 |
| shogun-notifier- | shogun: and MulticlassOCAS has been set to use only 1 thread | 15:51 |
| shogun-notifier- | shogun: Viktor Gal :master * 18ea338 / tests/unit/ (2 files): https://github.com/shogun-toolbox/shogun/commit/18ea3388b7b64b08df95f7d7f34c7fad484c9851 | 15:51 |
| shogun-notifier- | shogun: Merge pull request #872 from vigsterkr/master | 15:51 |
| shogun-notifier- | shogun: | 15:51 |
| shogun-notifier- | shogun: Fix unit tests for SGVector::max_abs and MOCASSVM | 15:51 |
| -!- blackburn [~lisitsin@mxs.kg.ru] has quit [Quit: Leaving.] | 15:51 | |
| wiking | interesting that until now that max_abs error has not surfaced... there's no fool proof unit-testing as this one proves it well ;) | 15:54 |
| sonne|work | wiking: what is the github bot? | 15:54 |
| wiking | sonne|work: we used to have github hook for commits and pull requests | 15:55 |
| wiking | that was set up with a hook | 15:55 |
| shogun-buildbot_ | build #811 of deb1 - libshogun is complete: Failure [failed test] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/811 blamelist: Viktor Gal <vigsterkr@gmail.com> | 16:00 |
| wiking | mmm | 16:00 |
| wiking | seems 1 thread did not help | 16:00 |
| wiking | :( | 16:00 |
| sonne|work | wiking: no longer working with github :/ | 16:01 |
| wiking | really? :) i guess then it's a bug of github that we should report | 16:03 |
| -!- tom__ [2eda6d52@gateway/web/freenode/ip.46.218.109.82] has joined #shogun | 16:04 | |
| tom__ | hi! | 16:04 |
| sonne|work | wiking: or they shoudl just install irker on their side... | 16:04 |
| sonne|work | wiking: about the build fail - could you please print the objective value in the test? | 16:05 |
| tom__ | Quesion about saving/loading mklmachine : I use 'CSerializableAsciiFile' but when loaded from file, mkl->getsubkernelweights() return 0. Is there code sample for that ? | 16:06 |
| sonne|work | tom__: please give us a minimal example to reproduce | 16:07 |
| sonne|work | then we can create a test out of this and fix it! | 16:07 |
| tom__ | sonne|work: ok | 16:08 |
| tom__ | sonne|work: here it is http://pastebin.com/ZZKNRzqj | 16:15 |
| tom__ | my first question is: does the mkl saved in file contain all kernels, machine params and weights ? | 16:16 |
| -!- n4nd0 [~nando@n158-p227.kthopen.kth.se] has quit [Quit: leaving] | 16:33 | |
| tom__ | sonne|work: I just re-use the mkl example from shogun examples and add save and load function. Ths for help | 16:52 |
| sonne|work | tom__: yes it should | 16:53 |
| sonne|work | tom__: it should be self containing | 16:53 |
| tom__ | then no need to save kernels and weights separatly ? | 16:53 |
| tom__ | ok | 16:53 |
| sonne|work | no | 16:53 |
| sonne|work | just save the core mklmachine | 16:53 |
| sonne|work | then all kernels/labels/features/parameters *should* be saved too | 16:54 |
| sonne|work | if not -> bug | 16:54 |
| tom__ | that's what I am doing in the example I gave you. it is c++, maybe I misunderstood smthg | 16:54 |
| tom__ | how to access to combined kernel and weights from mkl pointer ? | 16:55 |
| sonne|work | for example the multiclass classifier for the hand written digit recognition is stored exactly like that (the one from here http://shogun-toolbox.org/page/documentation/samples/ ) | 16:55 |
| sonne|work | tom__: gtg now might have a look later | 16:55 |
| sonne|work | cu | 16:55 |
| tom__ | ok thx anyway i wil lokk into samples | 16:56 |
| -!- tom__ [2eda6d52@gateway/web/freenode/ip.46.218.109.82] has quit [Quit: Page closed] | 16:56 | |
| -!- KMcQuisten [d8338942@gateway/web/freenode/ip.216.51.137.66] has joined #shogun | 17:35 | |
| KMcQuisten | Hi folks | 17:36 |
| -!- KMcQuisten [d8338942@gateway/web/freenode/ip.216.51.137.66] has quit [Quit: Page closed] | 17:41 | |
| -!- ipondering [~iponderin@li568-69.members.linode.com] has joined #shogun | 18:07 | |
| -!- ipondering [~iponderin@li568-69.members.linode.com] has left #shogun [] | 18:08 | |
| -!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout] | 18:51 | |
| -!- blackburn [~blackburn@188.168.2.253] has joined #shogun | 18:51 | |
| blackburn | sonney2k: around? | 19:26 |
| -!- heiko1 [~heiko@nat-166-57.internal.eduroam.ucl.ac.uk] has joined #shogun | 20:02 | |
| -!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 20:16 | |
| blackburn | heiko1: really? | 20:19 |
| blackburn | :) | 20:19 |
| blackburn | NP-hard graph colouring? | 20:19 |
| heiko1 | blackburn, greedy version allowed :) | 20:21 |
| heiko1 | blackburn, updated | 20:22 |
| blackburn | heiko1: what do you need it for btw? | 20:22 |
| heiko1 | just need *a* graph colouring | 20:22 |
| heiko1 | for estimating the trace of a matrix | 20:22 |
| heiko1 | as | 20:23 |
| heiko1 | tr A = E_s [s^T A s] where s is a random variable with zero mean and identity covariance | 20:23 |
| heiko1 | the distribution of s determines the variance of this estimator | 20:23 |
| heiko1 | and there is a clever way to choose it | 20:24 |
| heiko1 | based on a graph colouring of the matrix | 20:24 |
| blackburn | interesting | 20:24 |
| heiko1 | blackburn, its a bit weird, but gives really cool results | 20:24 |
| heiko1 | Google Hutchison estimator for the basic ideas | 20:24 |
| heiko1 | (without graph colouring) | 20:24 |
| blackburn | okay anyway I don't mind :) | 20:24 |
| heiko1 | hey blackburn a question | 20:25 |
| blackburn | yes? | 20:25 |
| heiko1 | c++ /swig /python | 20:25 |
| heiko1 | I have a class which inherits from a templated class | 20:25 |
| heiko1 | so class B: A<float64_t> | 20:26 |
| heiko1 | fixing the template arugment | 20:26 |
| heiko1 | I can call any methods from B<T> in C++ | 20:26 |
| heiko1 | but not from python modular | 20:26 |
| heiko1 | says they dont exist | 20:26 |
| heiko1 | blackburn, any idea how to solve that? | 20:27 |
| heiko1 | I could make B also a template and then pass the parameter, create a bunch of names for pyhton for all different types | 20:27 |
| heiko1 | this works | 20:27 |
| heiko1 | but I dont want to do all the template hell since not needed anyway | 20:27 |
| blackburn | wait wait let me understand | 20:27 |
| blackburn | nicht schiessen | 20:28 |
| heiko1 | I will commit this then you can try it out | 20:28 |
| blackburn | ahh | 20:28 |
| blackburn | that sounds strange for me actually | 20:28 |
| blackburn | didn't know about that | 20:29 |
| heiko1 | maybe I did a mistake | 20:29 |
| blackburn | heiko1: what if A->B->C | 20:30 |
| blackburn | and B is templated | 20:30 |
| blackburn | while A and C not | 20:30 |
| blackburn | does C have methods of A? | 20:30 |
| heiko1 | dont know | 20:31 |
| heiko1 | but thats not my problem :D | 20:31 |
| blackburn | wait wait | 20:31 |
| blackburn | we have such example | 20:31 |
| blackburn | heiko1: CPNorm | 20:31 |
| blackburn | heiko1: I think it works nice - nothing lost | 20:32 |
| -!- shogun-notifier- [~irker@7nn.de] has joined #shogun | 20:33 | |
| shogun-notifier- | shogun: Heiko Strathmann :master * 7e6bc9c / src/shogun/features/streaming/generators/ (2 files): https://github.com/shogun-toolbox/shogun/commit/7e6bc9c9fd797cd0120735760dc42393c06b63ab | 20:33 |
| shogun-notifier- | shogun: fixed typo | 20:33 |
| shogun-notifier- | shogun: Heiko Strathmann :master * 822d40a / src/shogun/features/streaming/generators/ (2 files): https://github.com/shogun-toolbox/shogun/commit/822d40a6534b801ea2c83fede17d9c165dde846c | 20:33 |
| shogun-notifier- | shogun: added new streaming data generator class for creating Gaussian blobs as in NIPS paper for optimal kernel choice for linear time MMD | 20:33 |
| shogun-notifier- | shogun: Heiko Strathmann :master * 773664d / examples/undocumented/libshogun/streaming_generators.cpp: https://github.com/shogun-toolbox/shogun/commit/773664da6de95fa85580eccd5aa549cf9f94d923 | 20:33 |
| shogun-notifier- | shogun: added example for new gaussian blobs data generator | 20:33 |
| shogun-notifier- | shogun: Heiko Strathmann :master * d9bb5b8 / / (7 files): https://github.com/shogun-toolbox/shogun/commit/d9bb5b825203d4e52c7e1399a3b3f2a56dca23ca | 20:33 |
| shogun-notifier- | shogun: made CMeanShiftDataGenerator not be a template anymore since useless and annoying | 20:33 |
| shogun-notifier- | shogun: Heiko Strathmann :master * 0a21ab6 / / (9 files): https://github.com/shogun-toolbox/shogun/commit/0a21ab68ecbae93c2f31dd3962b6512b9e12a3b6 | 20:33 |
| shogun-notifier- | shogun: Merge pull request #874 from karlnapf/master | 20:33 |
| shogun-notifier- | shogun: | 20:33 |
| shogun-notifier- | shogun: New class for data generation | 20:33 |
| heiko1 | blackburn, try in python to call the method get_streamed_features of CMeanShiftDataGenerator | 20:34 |
| heiko1 | probably just me being stupid :) | 20:34 |
| heiko1 | ah just read a hint in compile warnings | 20:35 |
| blackburn | heiko1: what is it? | 20:35 |
| heiko1 | ../../shogun/features/streaming/generators/MeanShiftDataGenerator.h:28: Warning 401: Nothing known about base class 'CStreamingDenseFeatures< float64_t >'. Ignored. | 20:36 |
| heiko1 | ../../shogun/features/streaming/generators/MeanShiftDataGenerator.h:28: Warning 401: Maybe you forgot to instantiate 'CStreamingDenseFeatures< float64_t >' using %template. | 20:36 |
| blackburn | ahh | 20:36 |
| blackburn | so here you go then :) | 20:36 |
| heiko1 | yep | 20:36 |
| heiko1 | this was not in the interface | 20:36 |
| heiko1 | mmh | 20:37 |
| heiko1 | it IS in there | 20:37 |
| heiko1 | #ifdef USE_FLOAT64 | 20:37 |
| heiko1 | %template(StreamingRealFeatures) CStreamingDenseFeatures<float64_t>; | 20:37 |
| heiko1 | #endif | 20:37 |
| heiko1 | I think the problem is that the wrong class name is used | 20:37 |
| heiko1 | when compiling the interface | 20:37 |
| heiko1 | any ideas how to fix that? | 20:38 |
| blackburn | hmm | 20:38 |
| shogun-buildbot_ | build #812 of deb1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/812 | 20:38 |
| heiko1 | this is the class | 20:38 |
| heiko1 | class CMeanShiftDataGenerator: public CStreamingDenseFeatures<float64_t> | 20:38 |
| heiko1 | I think fixing the base class here is the problem | 20:38 |
| heiko1 | python thinks that CStreamingDenseFeatures<float64_t> should be "real"densefeaturesblabla | 20:38 |
| shogun-buildbot_ | build #814 of deb1 - libshogun is complete: Failure [failed compile] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/814 blamelist: Heiko Strathmann <heiko.strathmann@gmail.com> | 20:38 |
| blackburn | hmm | 20:39 |
| blackburn | well | 20:39 |
| blackburn | two ideas | 20:39 |
| blackburn | idea one - forward declaration of that class in meanshift...h | 20:39 |
| blackburn | like template <class T> CStreamingDenseFeatures<T> | 20:39 |
| blackburn | template <class T> CStreamingDenseFeatures<T>; | 20:39 |
| blackburn | just declaration | 20:40 |
| @sonney2k | blackburn, yes %template is missing | 20:40 |
| heiko1 | in fact, it all compiled on my machine | 20:40 |
| blackburn | sonney2k: where? | 20:40 |
| blackburn | sonney2k: who could lead vowpal wabbit? | 20:41 |
| blackburn | I put it to myself for onw | 20:41 |
| blackburn | it is of great priority IMO | 20:42 |
| @sonney2k | why? | 20:42 |
| shogun-notifier- | shogun: Heiko Strathmann :master * e15edc3 / src/shogun/features/streaming/generators/MeanShiftDataGenerator.cpp: https://github.com/shogun-toolbox/shogun/commit/e15edc3eb2ac0c81b875801d22d9877eea75e396 | 20:42 |
| shogun-notifier- | shogun: added forgotten template arugment in base class | 20:42 |
| shogun-notifier- | shogun: Heiko Strathmann :master * 8d72b29 / src/shogun/features/streaming/generators/MeanShiftDataGenerator.cpp: https://github.com/shogun-toolbox/shogun/commit/8d72b2949edc7427ff33414b914239745d7e9a83 | 20:42 |
| shogun-notifier- | shogun: Merge pull request #875 from karlnapf/master | 20:42 |
| shogun-notifier- | shogun: | 20:42 |
| shogun-notifier- | shogun: quickfix | 20:42 |
| blackburn | sonney2k: it is something like true large scale | 20:43 |
| blackburn | heiko1: whoa! | 20:43 |
| blackburn | how did that compile :D | 20:43 |
| heiko1 | blackburn ? | 20:43 |
| heiko1 | no idea :) | 20:43 |
| blackburn | heiko1: wonderful world of C++ | 20:43 |
| heiko1 | sonney2k, blackburn, how to fix this template issue? | 20:44 |
| heiko1 | blackburn, yes :) | 20:44 |
| blackburn | heiko1: http://abstrusegoose.com/strips/ars_longa_vita_brevis.png | 20:44 |
| shogun-buildbot_ | build #813 of deb1 - libshogun is complete: Failure [failed test] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/813 blamelist: Heiko Strathmann <heiko.strathmann@gmail.com> | 20:44 |
| heiko1 | hehe , great comic blackburn :) | 20:45 |
| heiko1 | not responsible for the above fail | 20:45 |
| -!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Ping timeout: 245 seconds] | 20:46 | |
| @sonney2k | blackburn, I don't know - it is overhyped | 20:46 |
| @sonney2k | it is just sgd | 20:46 |
| blackburn | sonney2k: really? I don't know for sure | 20:46 |
| @sonney2k | but yes it can run on kiloclusters | 20:46 |
| @sonney2k | that is the cool part about it | 20:47 |
| @sonney2k | IMHO any truely largescale stuff is just (distributed) sgd | 20:47 |
| blackburn | sonney2k: I added ideas page - feel free to update if you have some horsepower | 20:48 |
| @sonney2k | no more horses left :( | 20:48 |
| blackburn | I think I do not so will continue tomorrow | 20:49 |
| blackburn | :) | 20:49 |
| @sonney2k | heiko1, what did you do? you derived from a templated class without specifying the type? | 20:49 |
| @sonney2k | you crazy? | 20:50 |
| heiko1 | nono | 20:50 |
| heiko1 | sonney2k, class CMeanShiftDataGenerator: public CStreamingDenseFeatures<float64_t> | 20:50 |
| heiko1 | works fine in C++ | 20:50 |
| heiko1 | I think its an issue of renaming CStreamingDenseFeatures<float64_t> to StreamingRealFeatures | 20:51 |
| heiko1 | when processing Features.i | 20:51 |
| heiko1 | it does work when I do | 20:52 |
| heiko1 | template <class T> class CMeanShiftDataGenerator: public CStreamingDenseFeatures<T> | 20:52 |
| heiko1 | but then I have to create all these names for the different types | 20:53 |
| heiko1 | which I would like to avoid | 20:53 |
| shogun-buildbot_ | build #815 of deb1 - libshogun is complete: Failure [failed compile] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/815 blamelist: Heiko Strathmann <heiko.strathmann@gmail.com> | 20:57 |
| @sonney2k | heiko1, I have to see the error | 20:57 |
| blackburn | see you tomorrow | 20:58 |
| -!- blackburn [~blackburn@188.168.2.253] has quit [Quit: Leaving.] | 20:58 | |
| -!- lambday [3ba2174d@gateway/web/freenode/ip.59.162.23.77] has quit [Ping timeout: 245 seconds] | 20:58 | |
| heiko1 | sonney2k, ok will show soon, first have to fix this thing here, which is strange since compiles on my machine | 20:59 |
| heiko1 | gotta go now, will look after the build in a few hours when home | 21:02 |
| -!- FSCV_ [~FSCV@173.254.212.46] has joined #shogun | 21:04 | |
| -!- FSCV [~FSCV@204.45.110.163] has quit [Ping timeout: 245 seconds] | 21:06 | |
| -!- syst3mw0rm [~quassel@aws.aamirkhan.co.in] has quit [Ping timeout: 255 seconds] | 21:06 | |
| -!- syst3mw0rm [~quassel@aws.aamirkhan.co.in] has joined #shogun | 21:09 | |
| shogun-notifier- | shogun: Soeren Sonnenburg :master * b80a07c / src/ (3 files): https://github.com/shogun-toolbox/shogun/commit/b80a07c63b7123d23c4b1f4682c000daf404182d | 21:12 |
| shogun-notifier- | shogun: fix %template warning and potential compile error | 21:12 |
| @sonney2k | wiking, so how do I write a test | 21:13 |
| @sonney2k | wiking, I want to write one for MKL | 21:13 |
| @sonney2k | err CombinedKernel | 21:13 |
| @sonney2k | shogun/kernel/CombinedKernel.cpp | 21:13 |
| shogun-buildbot_ | build #816 of deb1 - libshogun is complete: Failure [failed test] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/816 blamelist: Soeren Sonnenburg <sonne@debian.org> | 21:24 |
| wiking | sonney2k: well follow my lead :) | 21:25 |
| -!- heiko1 [~heiko@nat-166-57.internal.eduroam.ucl.ac.uk] has left #shogun [] | 21:25 | |
| wiking | see latentsvm for example | 21:25 |
| -!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 21:27 | |
| shogun-notifier- | shogun: Soeren Sonnenburg :master * f509af5 / tests/unit/classifier/svm/SVMOcas_unittest.cc: https://github.com/shogun-toolbox/shogun/commit/f509af518f34862e792b86b65b3831fd13203976 | 21:30 |
| shogun-notifier- | shogun: add objective test for ocas | 21:30 |
| @sonney2k | wiking, done :) | 21:30 |
| shogun-notifier- | shogun: Soeren Sonnenburg :master * 9580b2d / tests/unit/kernel/CombinedKernel_unittest.cc: https://github.com/shogun-toolbox/shogun/commit/9580b2de01e236f9bbc411ee066f40f41179ee6c | 21:30 |
| shogun-notifier- | shogun: add test for combined kernel weights getter/setter | 21:30 |
| shogun-notifier- | shogun: Soeren Sonnenburg :master * bba7f87 / tests/unit/multiclass/MulticlassOCAS_unittest.cc: https://github.com/shogun-toolbox/shogun/commit/bba7f878e20f6440b221b564015637620d713c66 | 21:34 |
| shogun-notifier- | shogun: set #threads to one *before* training and epsilon to sth very low | 21:34 |
| shogun-buildbot_ | build #817 of deb1 - libshogun is complete: Failure [failed test] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/817 blamelist: Soeren Sonnenburg <sonne@debian.org> | 21:39 |
| n4nd0 | sonney2k: hey, what do you think about the project idea "Further design and implementation of an API for structured output learning"? | 21:40 |
| shogun-notifier- | shogun: Soeren Sonnenburg :master * d41b313 / tests/unit/kernel/CombinedKernel_unittest.cc: https://github.com/shogun-toolbox/shogun/commit/d41b3134895c3e95440fe9c8e446d6d8cddc69cd | 21:45 |
| shogun-notifier- | shogun: add serialization test for combined kernel | 21:45 |
| @sonney2k | wiking, gtest stuff is really a cool and convenient framework!! | 21:46 |
| n4nd0 | sonney2k: do you know anyone who could mentor that? | 21:46 |
| @sonney2k | n4nd0, nah needs a totally different name | 21:46 |
| @sonney2k | n4nd0, did you ask georg? | 21:46 |
| n4nd0 | sonney2k: aham, such as? | 21:46 |
| n4nd0 | sonney2k: yep | 21:46 |
| n4nd0 | he is going to contact us by this weekend with his ideas | 21:47 |
| @sonney2k | n4nd0, names that continue some work are not a good thing | 21:47 |
| n4nd0 | aham I understand | 21:47 |
| n4nd0 | and I think it makes sense | 21:47 |
| @sonney2k | so sth like implement CRFs, Markov Random Field, Conintuos variable SO would be good names :) | 21:48 |
| n4nd0 | got it | 21:48 |
| * sonney2k wonders what the issue was with _tom's problem - the tests work just fine... | 21:49 | |
| @sonney2k | who was it who said sparse features + serialization doesn't work? | 21:49 |
| n4nd0 | not me I think :) AFAIK SGNDArray + serialization doesn't work, other than that everything works :) | 21:50 |
| @sonney2k | n4nd0, https://github.com/shogun-toolbox/shogun/issues/851 | 21:58 |
| @sonney2k | this one | 21:58 |
| @sonney2k | I've just written a test | 21:58 |
| @sonney2k | and it indeed fails | 21:58 |
| @sonney2k | wiking, how can I run just a single test e.g. via valgrind? | 21:58 |
| @sonney2k | debugging? | 21:58 |
| @sonney2k | with gdb? | 21:58 |
| @sonney2k | argh | 22:05 |
| @sonney2k | it is the inplace destructor | 22:05 |
| -!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Ping timeout: 255 seconds] | 22:16 | |
| -!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 22:17 | |
| shogun-notifier- | shogun: Soeren Sonnenburg :master * 8cad60c / tests/unit/features/SparseFeatures_unittest.cc: https://github.com/shogun-toolbox/shogun/commit/8cad60cd802899b362aef78286d2f6da449586e1 | 22:18 |
| shogun-notifier- | shogun: add sparse features serialization test | 22:18 |
| shogun-notifier- | shogun: Soeren Sonnenburg :master * 2f80763 / tests/unit/Makefile: https://github.com/shogun-toolbox/shogun/commit/2f807630df1d2b18345701ba662d96e1108452fa | 22:18 |
| shogun-notifier- | shogun: add valgrind target | 22:18 |
| -!- cwidmer [8ca3fe9d@gateway/web/freenode/ip.140.163.254.157] has joined #shogun | 22:21 | |
| cwidmer | hi all | 22:21 |
| cwidmer | what, no blackburn? | 22:21 |
| cwidmer | ;) | 22:21 |
| n4nd0 | cwidmer: he said good night some time ago hehe | 22:22 |
| n4nd0 | I am surprised too! :) | 22:23 |
| -!- zxtx_ [~zv@cpe-75-83-151-252.socal.res.rr.com] has quit [Ping timeout: 244 seconds] | 22:23 | |
| cwidmer | n4nd0, wow, that seems unusual | 22:23 |
| cwidmer | just sent you an email | 22:23 |
| cwidmer | looking into SGMatrixList after all | 22:23 |
| n4nd0 | reading | 22:23 |
| cwidmer | https://github.com/cwidmer/shogun/blob/cc01bf5ca02aa0425be24159eae8b842f89c5a23/src/interfaces/python_modular/swig_typemaps.i | 22:24 |
| cwidmer | I did some typemaps some time ago | 22:24 |
| cwidmer | this one seems to be more involved, though | 22:24 |
| n4nd0 | then you know more about typemaps than I do :) | 22:24 |
| n4nd0 | but sonney2k seems very active this evening | 22:24 |
| n4nd0 | maybe he can help you? | 22:24 |
| n4nd0 | I am taking a look in any case | 22:25 |
| cwidmer | sonney2k, around? | 22:25 |
| cwidmer | trying to add typemap for SGMatrixList | 22:25 |
| n4nd0 | cwidmer: curiosity, what is it you didn't like of the SGNDArray interface? | 22:29 |
| n4nd0 | cwidmer: stupid question, it hits me that there is matrix_list_from_numpy but no matrix_list_to_numpy | 22:31 |
| n4nd0 | as for SGMatrix, SGVector, etc | 22:31 |
| cwidmer | that's right, its one directional at the moment | 22:32 |
| n4nd0 | not needed yet? | 22:32 |
| cwidmer | only an IN typemap | 22:32 |
| n4nd0 | ok | 22:32 |
| cwidmer | as for SGNDArray, it doesn't let me access the matrices as nicely using operator() | 22:32 |
| cwidmer | I mean, SGMatrix now has operator() | 22:33 |
| n4nd0 | aham | 22:33 |
| n4nd0 | SGNDArray is a T* actually, isn't it? | 22:33 |
| n4nd0 | IIRC | 22:33 |
| cwidmer | SGMatrixList seems to have operator[] | 22:33 |
| n4nd0 | yeah, it does | 22:34 |
| cwidmer | so it's just a matter of nicely accessing it | 22:34 |
| n4nd0 | I understand | 22:35 |
| n4nd0 | cwidmer: did you see SGNDArray::get_matrix()? | 22:38 |
| n4nd0 | I think it should be better done returning an SGMatrix instead of a T* but still | 22:38 |
| n4nd0 | good night then guys | 22:50 |
| -!- sumit [73f91219@gateway/web/freenode/ip.115.249.18.25] has joined #shogun | 22:51 | |
| -!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving] | 22:51 | |
| --- Log closed Thu Feb 14 00:00:45 2013 | ||
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!
