Open in new window / Try shogun cloud
--- Log opened Wed Mar 13 00:00:11 2013
blackburnI was in high school when house md first seasons were coming out00:01
n4nd0yeah me too00:02
blackburnas I was doing nothing in school I watched it00:02
n4nd0I watched house while in high school too00:02
blackburnI finished school in 2008 IIRC00:03
blackburn:D00:03
blackburnn4nd0: when did you?00:03
n4nd0mmmmm00:03
n4nd0I finished high school 6 years ago00:03
blackburn2007 then I guess00:03
n4nd02007 probably00:03
n4nd0yeah00:04
n4nd0I started uni when I was 1700:04
n4nd0now I am 2300:04
n4nd0so yeah 2007, indeed00:04
blackburnI am still 21 but 3 days lef00:04
blackburnt00:04
n4nd0happy birthday!!00:05
blackburnhah00:05
n4nd0haha00:05
blackburnmy brother has b-day today00:05
n4nd0happy birthday to him!00:05
blackburnyeah I'll let him know, thanks :)00:05
n4nd0Saturday 16th yours right?00:05
blackburnyeah00:05
blackburnour parents were precise00:05
n4nd0haha yeah00:05
n4nd0older brother?00:06
blackburn13 and 16 march but 7 years00:06
blackburnyes, 2900:06
blackburnn4nd0: https://dl.dropbox.com/u/10139213/share/image2.png00:09
blackburnhttps://dl.dropbox.com/u/10139213/share/image.png00:09
blackburnhttps://dl.dropbox.com/u/10139213/shogun/lle.png00:10
blackburnn4nd0: ^ something like that in d3.js would be ideal00:10
n4nd0yeah00:10
n4nd0I have to see how to do it00:10
blackburnhttps://dl.dropbox.com/u/10139213/shogun/pics.png00:10
blackburnn4nd0: http://bl.ocks.org/mbostock/95064200:12
blackburnn4nd0: images as nodes00:13
n4nd0cool00:21
blackburnn4nd0: http://bl.ocks.org/mbostock/1804919 that's an example of graph layout with 'suggested' position00:22
blackburnn4nd0: so you could change gravity function, add image nodes and that's it00:23
blackburntime to sleep00:24
n4nd0ok00:27
n4nd0gn!00:27
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving]00:41
-!- FSCV_ [~FSCV@65.19.131.247] has quit [Quit: Leaving]01:42
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout]02:10
shogun-buildbotbuild #276 of nightly_none is complete: Failure [failed compile]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/nightly_none/builds/27603:01
-!- zxtx [~zv@cpe-98-148-115-57.socal.res.rr.com] has joined #shogun03:50
-!- zxtx [~zv@cpe-98-148-115-57.socal.res.rr.com] has quit [Ping timeout: 258 seconds]05:48
-!- blackburn [~blackburn@188.168.13.118] has quit [Ping timeout: 264 seconds]07:10
-!- blackburn [~blackburn@188.168.3.95] has joined #shogun07:25
-!- zxtx [~zv@76.91.87.166] has joined #shogun07:46
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun07:59
shogun-notifier-shogun: Sergey Lisitsyn :master * 3b6afdd / src/shogun/lib/tapkee/ (3 files): https://github.com/shogun-toolbox/shogun/commit/3b6afdd8ed5266523fff13bf0b061c944cac25c407:59
shogun-notifier-shogun: Compilation fix for tapkee code which fails w/o ARPACK07:59
-!- zxtx [~zv@76.91.87.166] has quit [Ping timeout: 250 seconds]08:46
-!- blackburn [~blackburn@188.168.3.95] has left #shogun []08:51
-!- n4nd0 [~nando@n177-p170.kthopen.kth.se] has joined #shogun08:59
sonne|workshogun-buildbot: force build nightly_none09:09
shogun-buildbotbuild forced [ETA 15m26s]09:09
shogun-buildbotI'll give a shout when the build finishes09:09
-!- blackburn [~lisitsin@mxs.kg.ru] has joined #shogun09:22
shogun-buildbotbuild #277 of nightly_none is complete: Success [build successful]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/nightly_none/builds/27709:24
-!- zxtx [~zv@cpe-75-83-151-252.socal.res.rr.com] has joined #shogun09:38
-!- lambday [b613104d@gateway/web/freenode/ip.182.19.16.77] has joined #shogun10:05
lambdayhi.. I've been writing a string kernel and I want to normalize the it... how do I do it?10:07
lambdayI've set the normalizer in the constructor as set_normalizer(new CSqrtDiagKernelNormalizer());10:07
lambdayand tried to use return normalizer->normalize(total, idx_a, idx_b); in the compute function but got a segmentation fault10:08
blackburnlambday: normalizer is being called in base CKernel class, no need to call it once more10:12
lambdaybut compute function returns the unnormalized kernel value :(10:12
blackburnlambday:https://github.com/shogun-toolbox/shogun/blob/master/src/shogun/kernel/Kernel.h#L23410:16
blackburnlambday: compute should return unnormalized value as kernel() computes final value10:16
lambdayblackburn: oh... I get it now..10:17
lambdaythanks a lot10:17
sonne|worklambday: just make sure that you call init_normalizer() in kernels' init(a,b) function10:25
lambdaysonne|work: yes... I did and its working fine now :)10:26
sonne|workk10:26
lambdayone more thing... I want to add this SSKStringKernel to modular interface... what exactly should I do? blackburn told me few days back that it needs to be added to interfaces/modular/Kernel.i, and Kernel_includes.i.10:26
lambdayI see a few renames and includes in Kernel.i10:27
blackburnlambday: that's the only thing10:27
blackburnyes, rename and include in Kernel.i10:27
blackburnand just include in Kernel_includes.i10:27
lambdayblackburn: okay... i'm checking10:28
-!- Netsplit *.net <-> *.split quits: sonne|work10:36
-!- Netsplit over, joins: sonne|work10:46
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout]10:59
lambdayblackburn: thanks man... its working perfectly... the kernel matrix is symmetric.. tested with some data..11:32
sonne|worklambday: now it only has to be positive :-]11:33
lambdayhow do i test its psd?11:33
sonne|worklambday: empirically? check if eigenvalues are positive11:36
lambdaysonne|work: I gotta check it manually? :(11:39
sonne|worklambday: yeah proof it!11:40
lambdaysonne|work: okay...11:40
sonne|worklambday: but don't worry if it is not - we have some 'kernels' in there that are not by theory but empirically everything is good11:42
lambdaysonne|work: that sounds relaxing :)11:45
lambdayI gotta check its performance on some protein test data... I should compare its performance against other string kernels... let's see :)11:47
lambdaythanks a lot :)11:47
-!- blackburn [~lisitsin@mxs.kg.ru] has quit [Quit: Leaving.]11:58
-!- blackburn [~lisitsin@mxs.kg.ru] has joined #shogun12:20
-!- heiko [~heiko@pat-191-250.internal.eduroam.ucl.ac.uk] has joined #shogun12:35
-!- sumit [73f91219@gateway/web/freenode/ip.115.249.18.25] has joined #shogun12:36
-!- sumit [73f91219@gateway/web/freenode/ip.115.249.18.25] has quit [Ping timeout: 245 seconds]12:47
blackburnheiko: I have fixed something12:50
heikoblackburn:  nice :)(12:50
heikoI am scanning the gpr framework12:51
heikofound some bugs12:51
heikostill results are nonsense12:51
heikohope to finish this today12:51
heikoat least for regression12:51
heikobtw while looking through this, I think I will write a project description for GP classification for gsoc12:51
heikospent whole yesterday evening with this gpr stuff, we really have to make people write tests12:53
blackburnheiko: I don't get a thing about GP - that's a lot of code I am quite lost within12:57
heikoblackburn:  I know it should have been  better documented/organised12:58
heikothe variable names are very misleading12:58
heikobut I am on it12:58
blackburnheiko: that's "-" of part-time gsoc development12:58
blackburn"+" is that we get some code though12:58
heikoblackburn:  I think that is our fault12:58
heikonot the student's12:58
heikosince we did not make them12:58
heikoso lets do this differently this year12:58
blackburnheiko: exactly12:59
heikoI wont merge any new code without a unit test12:59
blackburnheiko: that's always management12:59
heikoand without loads of comments12:59
blackburnheiko: that's actually my PM who should be blamed I am reading about C++11 instead of doing something useful :D12:59
blackburnso yes, I agree12:59
blackburnheiko: it would be nice if you declare such principle on the mailing list13:00
blackburnI could send a scan of my sign :D13:00
blackburnheiko: there is an anecdote about russian diplomates13:01
heikoblackburn:  yes? :)13:01
blackburnheiko: something like that: teachers asked future diplomates to write a letter to somalian pirates13:01
blackburnwhich would force to set hostages free13:01
heikohaha :)13:02
blackburnweek later - teacher "nice, but assholes should be written without a space and you made some mistakes in word f*cking"13:02
blackburnheiko: that's a lame translation but I hope you get it right :)13:03
heikoI get it13:03
heikowell, we have the power to merge :)13:03
blackburnheiko: in russian it sounds funnier as we have infinite amounts of various blames13:04
blackburnheiko: so when you have some time and will, could you please write a manifest about that?13:05
heikoblackburn: I will, we can put it on the GSoC page13:05
heikolike guielines13:05
heikoon developing code13:05
blackburnheiko: yeap13:06
heikoblackburn: yeah I fixed the bug13:12
heikoblackburn: would have been easily catched by a test13:12
heikocaused by something that was developed after this first thing was checked13:12
heiko#but since not test, no detection that old breaks when new is added13:12
blackburnheiko: that's quite cool we all get importance of unit-tests13:13
heikoblackburn: sorry :) I am too keen on this13:13
blackburnheiko: no that's ok13:13
blackburnheiko: ideal case is first tests then code13:15
heikoblackburn: yes, code a module/method, add test, then use it in code13:15
blackburnheiko: it won't work with software you don't know how to write but for established stuff it is the only way13:15
heikoblackburn:  well ideally one things of the stuff you gonna write before you do it right? :)13:16
blackburnheiko: well interface should be know first - so interface tests should come before real coding13:16
blackburn(ideal)13:16
blackburnheiko: one thing sklearn rocks with is that they have ->100% test coverage13:17
heikoi know13:18
heikobut we will increase ours !13:19
-!- n4nd0 [~nando@n177-p170.kthopen.kth.se] has quit [Ping timeout: 246 seconds]13:45
sonne|workheiko: please don't document code!14:17
sonne|work(I am serious here)14:17
heikosonne|work:  not doing14:17
blackburnsonne|work: that's interesting - motivate ;)14:17
heikofixing14:17
sonne|workcomments are not a good idea14:18
heiko?14:18
sonne|workonly for doxygen api use14:18
blackburnyeah that's true, code should be selfcommenting14:18
sonne|workheiko: if you need comments your code is not readable14:18
heikosonne|work: I do  not agree14:18
heikoI think it is very good to explain what one is doing14:18
sonne|workheiko: no14:18
sonne|workheiko: give the function a proper name14:18
sonne|workinstead of calling it do_stuff14:19
sonne|workcall it compute_variance etc14:19
sonne|work-> no comment necessary14:19
heikoits way easier to read "compute cholesky" than14:19
heikocblas_dsymm(CblasColMajor, CblasLeft, CblasUpper,14:19
heikom_ktrtr.num_rows, temp2.num_cols, (m_scale*m_scale)/(sigma*sigma),14:19
heikom_ktrtr.matrix, m_ktrtr.num_cols,14:19
heikotemp2.matrix, temp2.num_cols, 1.0,14:19
heikotemp1.matrix, temp1.num_cols);14:19
heikomemcpy(m_kern_with_noise.matrix, temp1.matrix,14:19
heikotemp1.num_cols*temp1.num_rows*sizeof(float64_t));14:19
heikoclapack_dpotrf(CblasColMajor, CblasUpper,14:19
heikotemp1.num_rows, temp1.matrix, temp1.num_cols);14:19
heikom_L = SGMatrix<float64_t>(temp1.num_rows, temp1.num_cols);14:19
blackburnheiko: eigen3 :)14:19
heikoblackburn: I did not write that! :)14:19
blackburntrue14:19
sonne|workheiko: yes but call it compute_cholesky14:19
sonne|workand put it in a function14:20
heikosonne|work: also sometimes its good to describe what the next section will be doing14:20
sonne|workno code needed14:20
sonne|workheiko: no!14:20
sonne|workexactly not!14:20
sonne|workif you want to comment what the next section does14:20
blackburnheiko: matrix.ldlt().solve(rhs)14:20
sonne|workput it in a function and give it a self-describing name!14:20
blackburnheiko: ^ eigen3 way14:20
sonne|workheiko: btw what is the state about this GP stuff?14:20
heikosonne|work: I do not agree, of course this is what one should do, but still, sometimes comments are needed.14:21
heikosonne|work, I solved the bug that made the basic GPR return nonsense14:21
sonne|workvery extremely rarely14:21
sonne|worknice!14:21
heikoand wrote many tests to check14:21
heikobut now14:21
heikoit crashes depending on size of test data14:21
heiko(it did that before)14:21
heikoso investigating ....14:21
heikoat least the regression works if it terminates14:22
heiko blackburn: I know, but I am not going to change all these calls :)  (as long as they work)14:23
blackburnheiko: just persuading you to join the dark side14:24
heikohey, I use eigen3 for my current project! :)14:24
heikobtw exactly this call you mentioned :)14:24
blackburnheiko: then I am helping you with motivation to not left the dark side :)14:25
heikohehe14:25
heikoblackburn: maybe I should use eigen3, there is an error in the cholesky for certain matrix sizes14:27
heikoblackburn, whats the best way to convert a shogun matrix into an eigen3 one?14:28
heikoI think thats the problem, there is some uninitialised memory14:29
blackburnheiko: Map class14:29
blackburnheiko: Map<MatrixXd>(pointer,cols,rows)14:30
blackburnheiko: act just like with typical matrix afterwards14:31
heikoMap class of eigen3?14:31
heikoso this returns a 2d array?14:31
blackburnheiko: yes, just like MatrixXf14:33
blackburnor MatrixXd14:33
heikonice, what about K^T * x , blackburn?14:39
heikohow do I do that?14:39
heikowill get lunch now, see you in a bit14:40
blackburnheiko: K.transpose()*x14:43
-!- n4nd0 [~nando@n177-p170.kthopen.kth.se] has joined #shogun15:01
-!- FSCV [~FSCV@204.45.132.11] has joined #shogun15:02
heikoblackburn, does that work in place?15:09
heikoI guess so15:09
blackburnheiko: no, y = K*transpose()*x will form it inplcae15:12
heikoblackburn:  I see thanks15:12
blackburnbut just K.transpose()*x does nothing15:12
heikoblackburn, LLT15:13
blackburnheiko: mutable stuff is in array mainly15:13
heikodoes it change the given matrix?15:13
blackburnheiko: solveInPlace()15:13
blackburnI think you need it if you want to change it15:13
heikoblackburn: is there any example eigen code in shogun that I could borrow some things from?15:16
blackburnheiko: argh wait that's not true15:16
blackburnheiko: LLT computes a copy15:16
blackburnyes15:16
blackburnheiko: https://github.com/shogun-toolbox/shogun/tree/master/src/shogun/lib/malsar15:16
blackburnthese algorithms are made of eigen15:17
heikoMap<VectorXd> Mzp_DiagSigz(x,n_tasks);Mzp_Pz = eigensolver.eigenvectors().real();15:17
blackburnheiko: yeah I had to keep naming to track with original implementation15:18
heikono worries, just asking whether this map is then the 2d array?15:18
blackburnheiko: yes, a vector of variable length (X) and (d)ouble type15:19
blackburnheiko: this code adhere Mzp_DiagSigz to pointer x15:19
blackburnand then steal eigenvectors I guess15:19
blackburnI don't remember exactly how can that work though15:20
blackburn:D15:20
blackburneigenvectors must be a matrix..15:20
heikoblackburn:15:21
heikosorry I always have problems with new frameworks15:22
heikowhat is the code15:22
heikoI have a shogun matrix , I want to turn it into MatrixXd15:22
heikoor similar15:22
blackburnheiko: map is the thing you need15:22
heikohow15:23
heikoMap<MatrixXd> K(m_kern_with_noise.matrix, m_ktrtr.num_rows,15:23
heikom_ktrtr.num_cols);15:23
heikoI did this15:23
heikonow I want to add something to the diagonal, how do I do that?15:23
heikohow to get the MatrixXd instance?15:23
heikoblackburn: ah I finally found the documentation site :) nevermind15:24
-!- FSCV [~FSCV@204.45.132.11] has quit [Ping timeout: 245 seconds]15:25
blackburnheiko: sorry getting distracted15:27
blackburnheiko: asDiagonal() += vec15:27
blackburnK.asDiagonal().array() += vec15:27
blackburnsth like that IIRC15:27
blackburnheiko: but actually yes they have not that bad doc15:28
blackburnsome things are not clear though15:28
blackburnI asked about some thing once on their channel15:28
blackburnthe answer I got was 'check doc'15:28
blackburngo figure :D15:28
heikoha15:28
heikoso now I have a MatrixXd15:30
heikohow to get the 2d array?  :D15:30
blackburnheiko: what do you mean by array?15:32
sonne|workblackburn: double[]15:32
heikoyes15:32
blackburnheiko: K.data()15:32
heikothanks!15:33
blackburnthat's pointer for me15:33
blackburn:D15:33
heikoblackburn, what I am trying to do:15:34
heikoI have a shogun matrix15:34
heikoand I want another shogun matrix of the cholesky of it15:34
heikoso it seems like I have to copy the data twice15:34
heikoor once?15:34
heikoor at all=?15:34
blackburnheiko: first map15:36
blackburnexisting matrix15:36
blackburnyou get Map<MatrixXd>15:36
blackburnthen prepare llt()15:36
blackburnahh I see what you mean15:36
blackburnyes matrix would try to shoot your leg if you just steal the matrix15:37
blackburn* the array15:37
heikoyes, I need to store the llt suff15:37
heikoso I need to memcpy right?15:38
blackburnheiko: yes I think so15:38
blackburnheiko: but I'd avoid memcpy here15:38
heikoso what to do?15:39
blackburnheiko: I don't like that solution15:39
blackburnheiko: just copy values with for15:39
blackburnheiko: but again that feels wrong, let me check15:40
-!- sumit [73f91219@gateway/web/freenode/ip.115.249.18.25] has joined #shogun15:40
blackburnheiko: why do you need it btw?15:41
heikoblackburn:  because the GP code stored the L for late ruse15:41
blackburncan't you store LLT?15:41
blackburnahh15:41
heikoand I dont want to store an eigen object15:41
blackburnheiko: alright, copying is quite bad though15:42
heikowhy?  isnt it faster than for15:42
blackburnheiko: I mean in general I don't like that you've got to copy it twice15:43
heikoit is only once right?15:43
heikothe map should not copy?15:43
blackburnheiko: I think I got a solution15:43
heikoor does llt copy?15:43
blackburnokay15:43
blackburnyou have SGMatrix G15:43
heikolets call it K15:43
blackburnyou create Map<MatrixXd> K15:43
blackburnokay15:43
blackburnsgK15:43
blackburnand eigK15:43
blackburnyou also have matrix SGMatrix L15:44
heikoeigK is the ma?p15:44
blackburnwhich is cholesky factor15:44
blackburnyou add Map<MatrixXd> eigL15:44
blackburnand15:44
blackburneigL = eigK.llt().matrixL();15:44
heikoI see15:44
heikogood idea15:44
heikobtw how to I get upper?15:44
blackburnheiko: that sounds like a most efficient way to do that15:44
heikolike in matlab?15:44
blackburnheiko: matrixU but what is upper?15:44
blackburndo you need it?15:45
heikothe upper triangular part15:45
heikoyes15:45
blackburnheiko: alright then matrixL and matrixU15:45
heikook15:45
heikothanks15:45
heikoI will try ...15:45
blackburnheiko: these matrices are actually a views15:45
blackburnheiko: ah btw if your K is symmetric15:46
blackburnK.selfadjointView<Eigen::Upper>().llt().matrixU()15:47
blackburnthat can be more efficient15:47
-!- FSCV [~FSCV@vpnreactor-static44.den.icastcenter.com] has joined #shogun15:47
heikowhat does it do?15:48
blackburnheiko: it tells eigen that your matrix can be treated as selfadjoint15:49
blackburnthus all operations that can improve from that would improve from that15:50
blackburnheiko: for example I use it in PCA-like algorithms15:50
blackburncovariance is C.selfadjointView<Eigen::Upper>().rankUpdate(v,v,1.0);15:50
blackburnrank update is C <- C + alpha*x*x' - it is done much faster then you know C is symmetric15:51
blackburnwhen*15:51
-!- sumit [73f91219@gateway/web/freenode/ip.115.249.18.25] has quit [Ping timeout: 245 seconds]15:55
heikoblackburn: ha you know what16:03
heikothe cholesky is stored16:03
heikobut not used for solving16:03
blackburnheiko: are you talking about the bug?16:06
heikoblackburn: plural, but no16:06
heikothe thing about storing the factor16:06
heikosay is there a way to solve a system with an existing cholesky?16:06
heikollt.solve(b)16:06
heikoonly does it for the computed one16:07
blackburnheiko: ahh yes that's a problem16:07
blackburnheiko: they didn't expect such thing I guess :)16:09
heikoblackburn: mmh that is annoying16:09
heikoI mean16:09
heikocurrently the GP framework solves again anyway16:09
heikobut it would be good not to do that16:09
heikouuuh16:10
heikothere are so many bugs16:11
heikoits not possible to deal with all of them16:11
blackburnheiko: okay I'll get back to you in a hour or a little more16:11
heikook see you16:11
blackburnswitching place16:11
-!- FSCV_ [~FSCV@85.17.20.88] has joined #shogun16:14
-!- blackburn [~lisitsin@mxs.kg.ru] has quit [Quit: Leaving.]16:14
-!- FSCV [~FSCV@vpnreactor-static44.den.icastcenter.com] has quit [Ping timeout: 248 seconds]16:16
-!- FSCV_ [~FSCV@85.17.20.88] has quit [Ping timeout: 260 seconds]16:38
-!- FSCV_ [~FSCV@76.73.18.131] has joined #shogun16:38
-!- FSCV_ [~FSCV@76.73.18.131] has quit [Ping timeout: 264 seconds]16:46
-!- FSCV_ [~FSCV@65.19.131.247] has joined #shogun16:51
-!- n4nd0 [~nando@n177-p170.kthopen.kth.se] has quit [Quit: leaving]17:08
-!- blackburn [~blackburn@83.234.54.160] has joined #shogun17:11
-!- sonne|work [~sonnenbu@sams-office-nat.tomtomgroup.com] has quit [Quit: Leaving.]17:48
blackburnheiko: I'll have some 2-3 hours tonight for shogun stuff I think, let me know what can I fix17:54
heikoblackburn:  cool will do18:07
heikogpr is about to be working again18:07
heikojust had a break18:07
heikoblackburn: heureca, it works!18:31
heikoblackburn: multiclass ocas test fails18:45
blackburnheiko: alright18:49
blackburnheiko: mc ocas then18:49
blackburnheiko: what was wrong with gps?18:51
heikoblackburn: two things:18:51
heikoa wrong feature object was used at a place so the thing was doing a different thing that it should have18:52
heikoand a lapack method was wrongly called, causing memory errors/crashes18:52
heikoseems to work quite nice now18:52
heikobut only tested the basic GPR18:52
blackburnI am going to update tapkee in shougn now18:52
heikowhich is only a small part of the framework18:52
blackburnand then check mc ocas18:52
heikoblackburn:  no18:52
heikolets do bugs18:52
heikoI suggest18:52
blackburnheiko: that's just one commit18:52
heikoblackburn: ok :)18:52
blackburnheiko: I feel uncomfortable when tapkee is out of date in shogun18:53
blackburnheiko: I had to write a script that updates all includes to match that <shogun/lib/tapkee/...> :D18:54
heikohere we go!18:56
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun18:57
shogun-notifier-shogun: Heiko Strathmann :master * 1a024d4 / / (5 files): https://github.com/shogun-toolbox/shogun/commit/1a024d413ab76903a24c0876f12b9e2bbd15be9918:57
shogun-notifier-shogun: added additional parameter to mean shift data generator (select dimension of shift) and make sure that tests/examples still work18:57
shogun-notifier-shogun: Heiko Strathmann :master * b847c03 / examples/undocumented/python_modular/library_time.py: https://github.com/shogun-toolbox/shogun/commit/b847c03973f7855bb5a3f873a6ce081aebe98e8e18:57
shogun-notifier-shogun: added more test cases and make use of parameter list18:57
shogun-notifier-shogun: Heiko Strathmann :master * 7c25b13 / src/shogun/regression/ (2 files): https://github.com/shogun-toolbox/shogun/commit/7c25b13b5d880ea87aa89f28a3c3e4285030085518:57
shogun-notifier-shogun: fixes the bug that GP regression returns nonsense18:57
shogun-notifier-shogun: Heiko Strathmann :master * f37c2d8 / src/shogun/regression/gp/InferenceMethod.h: https://github.com/shogun-toolbox/shogun/commit/f37c2d87321ab9dc71c77a8d615343d4510c6ae618:57
shogun-notifier-shogun: added a note18:57
shogun-notifier-shogun: Heiko Strathmann :master * 8bd57ff / src/shogun/regression/gp/ExactInferenceMethod.cpp: https://github.com/shogun-toolbox/shogun/commit/8bd57ff2b969f63f66f29f8ddc7b9c039a33c26218:57
shogun-notifier-shogun: replaced lapack by eigen3 calls and make code much easier to read18:57
shogun-notifier-shogun: Heiko Strathmann :master * 58366b5 / examples/undocumented/libshogun/ (2 files): https://github.com/shogun-toolbox/shogun/commit/58366b5899ed5b890fd27ca379780023bd8a296c18:57
shogun-notifier-shogun: added a simply GP regression example18:57
shogun-notifier-shogun: Heiko Strathmann :master * 42a7929 / tests/unit/regression/ (2 files): https://github.com/shogun-toolbox/shogun/commit/42a79292ec8311cc2c8f4a4bb94e02189a86e17618:57
shogun-notifier-shogun: new unit tests for gp regression with exact inference18:57
shogun-notifier-shogun: Merge pull request #919 from karlnapf/master18:57
shogun-notifier-shogun:18:57
shogun-notifier-shogun: Made classic GP regression work again18:57
blackburnheiko: big thin18:57
blackburng18:57
heikoblackburn: when you fix ocas, also update the integration test since it is broken18:58
heikoblackburn: yes, that was a nasty one18:58
blackburnalright18:58
heikoeigen3 makes things much easier btw18:58
heiko30 lines -> 4 lines18:58
heikoand no bugs18:58
blackburnheiko: you get it18:58
heikoblackburn: converter_multidimensionalscaling_modular.py also has a broken integration test, could you add unit test to make sure it works and then replace integration test?18:59
blackburnheiko: yeah I changed it and will replace integration test18:59
-!- lambday [b613104d@gateway/web/freenode/ip.182.19.16.77] has quit [Quit: Page closed]19:01
blackburnwhite smoke!19:10
-!- AnkitAgrawal [3ba21749@gateway/web/freenode/ip.59.162.23.73] has joined #shogun19:13
-!- AnkitAgrawal [3ba21749@gateway/web/freenode/ip.59.162.23.73] has quit [Ping timeout: 245 seconds]19:31
blackburnheiko: I like the system I built in tapkee19:38
blackburnterminate called after throwing an instance of 'tapkee::wrong_parameter_type_error'19:38
blackburn  what():  Wrong type for parameter target_dimension. Should be IndexType19:38
blackburnheiko: quite easy to fix anything :)19:39
heikoblackburn19:42
heikomulticlass labels cannot be serialized19:42
heikodue to SGVector<SGVector ...19:42
heikocould you replace that by a SGMatrix?19:42
blackburnheiko: oops19:42
heikonumber is known19:42
blackburnyeah19:42
blackburnyes sure19:42
heikocool, let me know when its done, I have more integration tests working then :)19:43
blackburnheiko: how many are failing?19:43
heikonot too many19:43
heikogetting there19:43
heikolet me check19:43
blackburnheiko: we won't get it to zero though19:43
heikoblackburn:  I know19:43
heikobut we fix bugs19:44
heikothat is important19:44
heikomore importantly: yet unknown bugs19:44
heikothe known ones are in the giuthub issue list19:44
heikoso everyone can check19:44
heikoblackburn: btw I really like this matrix view of eigen19:45
blackburnheiko: which one?19:45
blackburnselfadjoint?19:45
heikono the map in general19:45
blackburnah19:45
heikothats cool stuff19:46
heikoblackburn: I will go home now, need some rest, worked all day on this stuff19:46
blackburnheiko: people have understood how to C++ finally19:46
blackburn:D19:46
heikoblackburn:  yeah :)19:46
blackburnyeah have a rest, I am on the duty now19:46
heikoafter killing themselves a few times in the future :D19:46
heikoblackburn: I have some time tomorrow so I can do final things19:47
blackburnexactly19:47
heikowhen to release tomorrow?19:47
heikoevening?19:47
blackburnheiko: I don't know19:47
blackburnthe master should say19:47
heikook then19:47
heikoso see you later or tomorrow more probably19:47
heikolet me know when the multiclass labels are changed19:48
blackburnheiko: I am writing tests for tapkee19:48
heikoI will change the test the19:48
heikon19:48
blackburnI am not sure if I should include it too19:48
blackburnin shogun19:48
heikowhat?19:48
heikotests?19:48
blackburnyeah19:48
heikodont know19:48
blackburnheiko: yeah me too19:48
heikois it that distinct?19:48
blackburnthe code of tapkee is just the same19:49
blackburnin standalone and in shogun19:49
blackburnso in principle they are not really required19:49
heikoI would say yes since then we can see them from shogun19:49
blackburnbut tests for interface of tapkee in shogun (CClasses) should be here for sure19:49
heikoyou know, I dont have to get involved into yet another framework then :D19:49
blackburnheiko: heh I use gtest too19:50
blackburnso I can just put them directly..19:50
blackburnheiko: almost no alternative to gtest for C++ I think19:51
blackburnboost.test may be19:51
heikodont know19:51
heikoblackburn: have a good evening, see you around19:51
blackburnsee you19:51
@sonney2kheiko, afternoon19:51
@sonney2ktomorrow would be good19:51
blackburnsonney2k: tomorrow evening?19:52
@sonney2kafternoon19:52
blackburnalright19:53
heikosonney2k: ok then19:57
heikosonney2k: why not evening?19:57
heikosonney2k: I have some more time in evening not so much in the afternoon (even earlier here)19:57
blackburnlibshogun.a 280 mb19:58
blackburnhah19:58
@sonney2kheiko, the release process takes hours... and I want to get some sleep....20:03
heikosonney2k: okay, thats a good point20:04
heiko :)20:04
-!- heiko [~heiko@pat-191-250.internal.eduroam.ucl.ac.uk] has left #shogun []20:06
shogun-notifier-shogun: Sergey Lisitsyn :master * d3d518c / src/shogun/lib/tapkee/ (25 files): https://github.com/shogun-toolbox/shogun/commit/d3d518cacac72cf33d7078be3063250c35293c2d20:35
shogun-notifier-shogun: A big update for Tapkee20:35
shogun-notifier-shogun:20:35
shogun-notifier-shogun: - Fixed a few VS 2012 compilations20:35
shogun-notifier-shogun: - Parallel MDS20:35
shogun-notifier-shogun: - Parallel20:35
shogun-buildbotbuild #959 of deb1 - libshogun is complete: Failure [failed compile]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/959  blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com>20:38
blackburndamn20:38
-!- travis-ci [~travis-ci@ec2-50-19-78-12.compute-1.amazonaws.com] has joined #shogun20:39
travis-ci[travis-ci] it's Sergey Lisitsyn's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: http://travis-ci.org/shogun-toolbox/shogun/builds/548071720:39
-!- travis-ci [~travis-ci@ec2-50-19-78-12.compute-1.amazonaws.com] has left #shogun []20:39
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun20:47
shogun-notifier-shogun: Sergey Lisitsyn :master * 6e01346 / src/shogun/lib/tapkee/tapkee_shogun.cpp: https://github.com/shogun-toolbox/shogun/commit/6e013467d73a06c61fa89c808eaa06927af546e320:59
shogun-notifier-shogun: Fixed compilation20:59
-!- travis-ci [~travis-ci@ec2-50-19-78-12.compute-1.amazonaws.com] has joined #shogun21:06
travis-ci[travis-ci] it's Sergey Lisitsyn's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: http://travis-ci.org/shogun-toolbox/shogun/builds/548123821:06
-!- travis-ci [~travis-ci@ec2-50-19-78-12.compute-1.amazonaws.com] has left #shogun []21:06
shogun-buildbotbuild #960 of deb1 - libshogun is complete: Success [build successful]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/96021:10
shogun-buildbotbuild #772 of bsd1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/772  blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com>21:16
shogun-notifier-shogun: Sergey Lisitsyn :master * 48366db / / (4 files): https://github.com/shogun-toolbox/shogun/commit/48366db1a2f832edcae22d05382a91fb4e5bef0d22:03
shogun-notifier-shogun: Some improvement for multiclass confidences22:03
shogun-notifier-shogun:22:03
shogun-notifier-shogun: - Changed the way confidences are stored (array of sgvectors -> sgmatrix)22:03
shogun-notifier-shogun:22:03
shogun-notifier-shogun: - Added a unit test for confidences22:03
blackburnoops22:06
-!- travis-ci [~travis-ci@ec2-50-19-78-12.compute-1.amazonaws.com] has joined #shogun22:11
travis-ci[travis-ci] it's Sergey Lisitsyn's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: http://travis-ci.org/shogun-toolbox/shogun/builds/548271922:11
-!- travis-ci [~travis-ci@ec2-50-19-78-12.compute-1.amazonaws.com] has left #shogun []22:11
shogun-buildbotbuild #787 of deb2 - static_interfaces is complete: Failure [failed test cmdline_static]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/787  blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com>22:18
blackburnbad attempt haha22:19
shogun-buildbotbuild #773 of bsd1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/773  blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com>22:21
shogun-notifier-shogun: Sergey Lisitsyn :master * c80ac5e / src/shogun/machine/MulticlassMachine.cpp: https://github.com/shogun-toolbox/shogun/commit/c80ac5e045d35c0d37d857e7b3dcec5a926cd7e222:21
shogun-notifier-shogun: Fixed MC machine confidences allocation22:21
shogun-buildbotbuild #774 of bsd1 - libshogun is complete: Failure [failed test]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/774  blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com>22:29
-!- travis-ci [~travis-ci@ec2-50-19-78-12.compute-1.amazonaws.com] has joined #shogun22:30
travis-ci[travis-ci] it's Sergey Lisitsyn's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: http://travis-ci.org/shogun-toolbox/shogun/builds/548321422:30
-!- travis-ci [~travis-ci@ec2-50-19-78-12.compute-1.amazonaws.com] has left #shogun []22:30
shogun-buildbotbuild #912 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/912  blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com>22:54
shogun-buildbotbuild #788 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/78823:06
shogun-buildbotbuild #913 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/91323:11
--- Log closed Thu Mar 14 00:00:12 2013