Open in new window / Try shogun cloud
--- Log opened Sat Mar 30 00:00:18 2013
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout]01:21
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving]01:28
-!- hoijui [~hoijui@dslb-088-075-046-171.pools.arcor-ip.net] has quit [Ping timeout: 260 seconds]02:11
shogun-buildbotbuild #338 of nightly_default is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/nightly_default/builds/33803:56
-!- sumit [73f91219@gateway/web/freenode/ip.115.249.18.25] has joined #shogun07:58
-!- sumit [73f91219@gateway/web/freenode/ip.115.249.18.25] has quit [Ping timeout: 245 seconds]08:36
-!- hoijui [~hoijui@dslb-088-075-046-171.pools.arcor-ip.net] has joined #shogun09:08
-!- wiking [~wiking@huwico/staff/wiking] has joined #shogun09:39
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun10:18
n4nd0sonney2k: good morning, around?10:19
n4nd0blackburn: ah you are also here :)10:19
blackburnn4nd0: yes10:19
n4nd0blackburn: I started taking a look yesterday to these tests that fail10:19
n4nd0I ma not really sure about a couple of things10:20
n4nd0blackburn: so the likeliest is that the errors are inside the data files I understand, the ones that are written in matlab10:20
blackburnn4nd0: most of errors are due to wrong parameter setting or introduced bugs10:21
n4nd0blackburn: and to find out if the data used makes any sense at all10:22
n4nd0one can plot it I guess10:23
blackburnn4nd0: I don't think the data is broken10:23
blackburnit is here from the very beginning10:23
n4nd0mm I understand10:23
@sonney2kn4nd0, blackburn good morning!10:26
@sonney2kabout the failing string kernel tests:10:27
@sonney2kit would already help to see if the kernels do sth reasonable10:27
@sonney2kthis could be done by generating some simple sequences say all contain letter 'A'10:27
@sonney2kand are of lenght say 5010:27
n4nd0ok10:28
@sonney2kconsider you have 20 sequences like this10:28
@sonney2kyou could introduce in 10 of them a couple of 'C's in the middle (or somewhere)10:28
@sonney2kand then just compute the kernel matrix10:28
@sonney2kyou should see a block structure10:28
@sonney2kwith largest values on the diagonal10:29
-!- lambday [0e8b614d@gateway/web/freenode/ip.14.139.97.77] has joined #shogun10:32
-!- hoijui [~hoijui@dslb-088-075-046-171.pools.arcor-ip.net] has quit [Quit: Leaving]10:33
lambdayfor the log-determinant of sparse matrix, shall I take eigen's sparse matrix as input? I thought of using SGSparseMatrix but it seems like an adjacency list and I don't know how to get the dimensions from it11:39
n4nd0sonney2k: http://pastebin.com/L1iaNB2011:39
n4nd0that is the test you have suggested above11:40
n4nd0I see block structure but no large values on the diagonal11:40
n4nd0lambday: SGSparseMatrix::num_vector SGSparseMatrix::num_features maybe help you to get dimensions??11:41
n4nd0num_vectors*11:41
lambdayn4nd0: ummm... I think it would be tough.. say, size of a sparse matrix can be 100x100 and say only 10 columns have non-zero entries, so SGSparseMatrix would have only 10 num_vector entries I am guessing..11:44
n4nd0lambday: no, I don't think so11:45
n4nd0for me it makes sense that num_vectors and num_features would both be 10011:45
n4nd0then there will be only stored 10 numbers11:45
n4nd0it is like if you do size of a sparse matrix in matlab11:46
n4nd0I mean the other elements exist, they are just zero so they are not explit. stored in memory11:47
n4nd0anyway, I am not testing right now what I am telling you - so don't trust me 100% :)11:47
lambdayn4nd0: I'11:48
lambdayI'm testing... will let you know :)11:48
@sonney2kn4nd0, could you put the 'A' strings all as the first 10 and the ones with C to the end?12:02
@sonney2kand then do12:02
@sonney2kfrom pylab import *12:02
@sonney2kpcolor(kmatrix)12:02
@sonney2kcolorbar()12:02
n4nd0sonney2k: ok, doing right now12:05
@sonney2kn4nd0, I am changing the code to use the sqrtdiagnormalizer by default12:05
@sonney2kI think that makes a lot of sense12:06
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun12:07
shogun-notifier-shogun: Soeren Sonnenburg :master * d07ddb4 / src/shogun/kernel/PolyKernel.cpp: https://github.com/shogun-toolbox/shogun/commit/d07ddb485b8c4fbde7bef821791fbdb88c87a77a12:07
shogun-notifier-shogun: move normalizer assignment to init() to avoid code duplication12:07
shogun-notifier-shogun: Soeren Sonnenburg :master * f33e4f3 / src/shogun/kernel/string/LocalAlignmentStringKernel.cpp: https://github.com/shogun-toolbox/shogun/commit/f33e4f35d5786227dc7493988f803db75208ecb012:07
shogun-notifier-shogun: normalize LA kernel to have norm 1 on the diagonal12:07
-!- travis-ci [~travis-ci@ec2-174-129-144-237.compute-1.amazonaws.com] has joined #shogun12:11
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/591877912:11
-!- travis-ci [~travis-ci@ec2-174-129-144-237.compute-1.amazonaws.com] has left #shogun []12:11
shogun-buildbotbuild #847 of deb2 - static_interfaces is complete: Failure [failed test python_static]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/847  blamelist: Soeren Sonnenburg <sonne@debian.org>12:12
n4nd0sonney2k: https://dl.dropbox.com/u/11020840/km_train.png12:13
shogun-buildbotbuild #1030 of deb1 - libshogun is complete: Failure [failed compile]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/1030  blamelist: Soeren Sonnenburg <sonne@debian.org>12:14
n4nd0sonney2k: I have no real idea how it should work but the blue part looks sort of weird, maybe it should be more uniform like the red??12:16
@sonney2kn4nd0, look at https://dl.dropbox.com/u/31480471/website-cbase.png12:17
n4nd0sonney2k: oh looking cool :)12:18
n4nd0but I guess we have to decide where to put it12:18
n4nd0we won't have the logo twice12:18
@sonney2kn4nd0, regarding the inhomogenity of the blue part12:18
@sonney2kcould you put the 'C' just at one fixed position in the sequence12:19
@sonney2kthen it should go away12:19
@sonney2k(I hope...)12:19
n4nd0sonney2k: for all the sequences?12:19
@sonney2kn4nd0, for the second half only12:19
n4nd0yes12:19
@sonney2kthe other are still just 'A's12:19
n4nd0I mean if for all the sequences with two 'C'12:19
n4nd0these two 'C' in the same position among all those sequences12:20
@sonney2kerrm?12:20
@sonney2kyes12:20
@sonney2kn4nd0, regarding the website12:21
@sonney2kmaybe we could put c-base in the middle12:21
@sonney2kand gsoc announcement as the long thing?12:21
n4nd0sonney2k: yeah, it sounds good12:21
@sonney2konly issue is we don't have much room for a workshop advertisement12:22
n4nd0sonney2k: why not to advert the workshop in the long thing?12:22
n4nd0on the rhs of the logo on top12:22
@sonney2kn4nd0, or that yes12:23
n4nd0sonney2k: fixing the 'C' ends with inhomogenity, I see only three colors now12:23
@sonney2kand then gsoc12:23
@sonney2kn4nd0, this kernel is really weird12:25
@sonney2klooks like it is not positive12:25
@sonney2kbut the result makes sense12:25
@sonney2kI assume that the red block are just the 'A's12:25
n4nd0yes12:25
@sonney2kand the blue block where the ones with A and C12:25
n4nd0yes12:26
@sonney2kand since it computes all possible local alignments between to sequences12:26
n4nd0and now there is a new block with darker blue that corresponds to only 'C's12:26
@sonney2kwe did have a quite inhomogenious blue block12:26
@sonney2kthe problem really is that 'AAA's align very well12:26
@sonney2kso they will have the highest score12:27
n4nd0sonney2k: how do see that it doesn't look positive? (I guess positive semi-definite)12:27
@sonney2kand there are not that many alignments of sequences that have a C12:27
@sonney2kn4nd0, it's diagonal should have the largest values12:28
@sonney2kin any case IMHO sanity check passed12:29
@sonney2kjust do the same with the other 2 string kernels12:29
n4nd0sonney2k: mmm OligoKernel gives me a completely uniform matrix12:33
n4nd0all values equal to 112:33
@sonney2kinvestigating12:34
@sonney2kn4nd0, do the other in the meantime12:34
n4nd0size_cache=10, k=3, width=1.212:34
n4nd0PolyKernel is the other one I guess12:34
n4nd0ah no12:34
n4nd0SimpleLocalityImproved12:34
@sonney2ktry a much larger width12:35
@sonney2ksay sth like length of the strings12:35
n4nd0still the same with width 5012:36
@sonney2kthat's weird12:38
shogun-buildbotbuild #974 of deb3 - modular_interfaces is complete: Failure [failed test python_modular]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/974  blamelist: Soeren Sonnenburg <sonne@debian.org>12:41
n4nd0sonney2k: and the last one looks pretty weird too12:41
n4nd0maybe I set nonsense parameters: length=5, inner_degree=5, outer_degree=112:42
shogun-notifier-shogun: Soeren Sonnenburg :master * dc0b603 / src/shogun/kernel/string/LocalAlignmentStringKernel.cpp: https://github.com/shogun-toolbox/shogun/commit/dc0b60389353019bbb2454a4b287346dbf22b83f12:43
shogun-notifier-shogun: add missing include to fix compile error12:43
@sonney2kn4nd0, image?12:43
@sonney2kcould you try inner_degree=112:44
n4nd0https://dl.dropbox.com/u/11020840/2013-03-30-124548_1600x900_scrot.png12:44
n4nd0that is with inner_degree=5, let me change12:44
@sonney2kthis certainly needs some normalization too12:46
n4nd0https://dl.dropbox.com/u/11020840/2013-03-30-124805_1600x900_scrot.png12:47
n4nd0with inner degree 1, ^12:47
shogun-buildbotbuild #1031 of deb1 - libshogun is complete: Success [build successful]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/103112:49
@sonney2kn4nd0, strange12:49
-!- travis-ci [~travis-ci@ec2-107-21-142-105.compute-1.amazonaws.com] has joined #shogun12:50
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/591908512:50
-!- travis-ci [~travis-ci@ec2-107-21-142-105.compute-1.amazonaws.com] has left #shogun []12:50
@sonney2kn4nd0, please email me the code I will have to investigate12:51
n4nd0sonney2k: ok12:51
n4nd0sent12:54
shogun-buildbotbuild #848 of deb2 - static_interfaces is complete: Failure [failed test python_static]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/848  blamelist: Soeren Sonnenburg <sonne@debian.org>12:59
shogun-buildbotbuild #975 of deb3 - modular_interfaces is complete: Failure [failed test python_modular]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/975  blamelist: Soeren Sonnenburg <sonne@debian.org>13:18
n4nd0ok, I've got a plane to take :) see you on Tuesday again13:23
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving]13:23
wikingsonney2k: can u plz change fbsd bot to clang13:48
wikingit should be fine now...13:48
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout]15:43
-!- lambday [0e8b614d@gateway/web/freenode/ip.14.139.97.77] has quit []16:39
-!- wiking [~wiking@huwico/staff/wiking] has quit [Read error: Connection reset by peer]16:42
-!- shogun-buildbot [~shogun-bu@7nn.de] has quit [Quit: buildmaster reconfigured: bot disconnecting]19:41
-!- shogun-buildbot [~shogun-bu@7nn.de] has joined #shogun19:42
@sonney2kshogun-buildbot, force build "bsd1 - libshogun"19:42
shogun-buildbotbuild forced [ETA 5m38s]19:42
shogun-buildbotI'll give a shout when the build finishes19:42
shogun-buildbotHey! build bsd1 - libshogun #843 is complete: Success [build successful]19:45
shogun-buildbotBuild details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/84319:45
@sonney2kwiking ^ all good20:05
blackburnsonney2k: did you notice anything wrong with these kernels?20:09
@sonney2kblackburn, LocalAlignment kernel looked ok20:10
@sonney2kI still have to check the others20:10
blackburnso what to do with it then?20:10
@sonney2kI've changed the default kernel normalizer to something I consider more reasonable - just generate new data for the la-kernel and good20:11
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun20:17
shogun-notifier-shogun: Soeren Sonnenburg :master * 640ab7f / src/shogun/kernel/string/SimpleLocalityImprovedStringKernel.cpp: https://github.com/shogun-toolbox/shogun/commit/640ab7f4a1bf93d0c073e662ae59babac77d5c2d20:17
shogun-notifier-shogun: normalize diagonal to one for LIK by default20:17
-!- travis-ci [~travis-ci@ec2-54-234-19-227.compute-1.amazonaws.com] has joined #shogun20:20
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/592572820:20
-!- travis-ci [~travis-ci@ec2-54-234-19-227.compute-1.amazonaws.com] has left #shogun []20:20
shogun-buildbotbuild #1032 of deb1 - libshogun is complete: Failure [failed compile]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/1032  blamelist: Soeren Sonnenburg <sonne@debian.org>20:28
shogun-notifier-shogun: Soeren Sonnenburg :master * 39c115e / src/shogun/kernel/string/SimpleLocalityImprovedStringKernel.cpp: https://github.com/shogun-toolbox/shogun/commit/39c115eaf37ce691968d8e4a6808ef15d30acfdd20:28
shogun-notifier-shogun: include normalizer after kernel header file to fix compile error20:28
-!- travis-ci [~travis-ci@ec2-23-20-101-65.compute-1.amazonaws.com] has joined #shogun20:35
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/592592620:35
-!- travis-ci [~travis-ci@ec2-23-20-101-65.compute-1.amazonaws.com] has left #shogun []20:35
shogun-buildbotbuild #1033 of deb1 - libshogun is complete: Success [build successful]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/103320:40
@sonney2kblackburn, I've checked the LIK kernel20:43
@sonney2kit also works as I expect it to work20:43
blackburnsonney2k: quite strange it fails..20:43
@sonney2kit is just that parameter choice is crucial20:43
shogun-buildbotbuild #849 of deb2 - static_interfaces is complete: Failure [failed test python_static]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/849  blamelist: Soeren Sonnenburg <sonne@debian.org>20:43
@sonney2kblackburn, Oligokernel is weird. I get only 1's in the km20:53
@sonney2kblackburn, do you know how to see the git log when a file was moved?20:53
blackburnsonney2k: no not really20:53
@sonney2kblackburn, git log --follow  shogun/kernel/string/OligoStringKernel.cpp20:58
blackburnwell that's just history?20:59
@sonney2kblackburn, exactly what I want20:59
blackburndirk messages in the bottom21:00
blackburnsonney2k: did you fire him?21:00
blackburn:D21:00
@sonney2kblackburn, no university kicked him out because he failed to enroll in time21:01
blackburnoh he writes poetry in commits21:01
@sonney2kbut we helped him to get his state back (our secretary went with him to the right people and all solved)21:02
@sonney2kblackburn, I think the bug is in oligokernel for 5 years or so21:03
blackburnha which means21:03
blackburnnobody is using it21:03
@sonney2kyes implementation in shogun is too slow anyway and the WD kernels are just much much more fast21:03
@sonney2kin 2009 I 'slightly sped up the oligokernel' haha21:04
@sonney2kmaybe it is broken since then21:04
shogun-buildbotbuild #976 of deb3 - modular_interfaces is complete: Failure [failed test python_modular]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/976  blamelist: Soeren Sonnenburg <sonne@debian.org>21:19
blackburnargh another day of dental thing is setting me to sleep21:20
blackburnsonney2k: I am now spending each saturday at dentist :D21:21
blackburnsonney2k: I'll take a look at sgd again tomorrow21:22
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout]23:28
--- Log closed Sun Mar 31 00:00:18 2013