Open in new window / Try shogun cloud
--- Log opened Sun Feb 19 00:00:31 2017
-!- suhas2go [uid201652@gateway/web/] has quit [Quit: Connection closed for inactivity]00:26
@sukeyPull Request #3399 "LinalgRefactor - Remove old vector_sum method"  synchronized by OXPHOS -
@sukeyPull Request #3641 "add tests for KNN and fix an error in KDTree solver"  synchronized by MikeLing -
-!- suhas2go [uid201652@gateway/web/] has joined #shogun04:44
-!- Hieronymous [~sudarshan@] has joined #shogun05:17
-!- goksinen [~goksinen@2604:2000:c591:8400:8d3e:9aa9:1628:cfc6] has joined #shogun05:39
-!- goksinen [~goksinen@2604:2000:c591:8400:8d3e:9aa9:1628:cfc6] has quit [Ping timeout: 240 seconds]05:44
-!- Hieronymous [~sudarshan@] has quit [Ping timeout: 260 seconds]06:10
-!- goksinen [~goksinen@2604:2000:c591:8400:8d3e:9aa9:1628:cfc6] has joined #shogun06:33
-!- goksinen [~goksinen@2604:2000:c591:8400:8d3e:9aa9:1628:cfc6] has quit [Ping timeout: 240 seconds]06:38
-!- Hieronymous [~sudarshan@] has joined #shogun07:00
-!- Hieronymous [~sudarshan@] has quit [Ping timeout: 240 seconds]07:08
-!- IRCFrEAK [] has joined #shogun07:09
-!- IRCFrEAK [] has left #shogun []07:10
-!- sudarshan_ [~sudarshan@2405:204:6388:9da2:285f:f89f:8a1a:2855] has joined #shogun07:21
-!- goksinen [~goksinen@2604:2000:c591:8400:8d3e:9aa9:1628:cfc6] has joined #shogun07:28
-!- goksinen [~goksinen@2604:2000:c591:8400:8d3e:9aa9:1628:cfc6] has quit [Ping timeout: 240 seconds]07:32
-!- kesslerfrost [~textual@2405:204:8d:d03b:4c05:6bea:84fc:8c8a] has joined #shogun07:39
kesslerfrosthi, I want to join the mailing list of shogun but I couldn't find any link on their github repository for gsoc07:40
kesslerfrostCan anyone send a link for the same. Thanks.07:41
-!- sudarshan_ [~sudarshan@2405:204:6388:9da2:285f:f89f:8a1a:2855] has quit [Quit: Leaving]08:02
-!- Hieronymous [~sudarshan@2405:204:6388:9da2:285f:f89f:8a1a:2855] has joined #shogun08:03
-!- kesslerfrost [~textual@2405:204:8d:d03b:4c05:6bea:84fc:8c8a] has quit [Ping timeout: 240 seconds]08:04
-!- goksinen [~goksinen@2604:2000:c591:8400:8d3e:9aa9:1628:cfc6] has joined #shogun08:22
-!- goksinen [~goksinen@2604:2000:c591:8400:8d3e:9aa9:1628:cfc6] has quit [Ping timeout: 240 seconds]08:26
-!- goksinen [~goksinen@2604:2000:c591:8400:8d3e:9aa9:1628:cfc6] has joined #shogun09:16
-!- goksinen [~goksinen@2604:2000:c591:8400:8d3e:9aa9:1628:cfc6] has quit [Ping timeout: 240 seconds]09:21
-!- kesslerfrost [~textual@2405:204:8d:d03b:299f:b5f8:a900:e616] has joined #shogun09:30
-!- kesslerfrost [~textual@2405:204:8d:d03b:299f:b5f8:a900:e616] has quit [Ping timeout: 240 seconds]10:10
-!- goksinen [~goksinen@2604:2000:c591:8400:8d3e:9aa9:1628:cfc6] has joined #shogun10:10
-!- goksinen [~goksinen@2604:2000:c591:8400:8d3e:9aa9:1628:cfc6] has quit [Ping timeout: 240 seconds]10:15
-!- goksinen [~goksinen@2604:2000:c591:8400:8d3e:9aa9:1628:cfc6] has joined #shogun11:03
-!- kesslerfrost [~textual@2405:204:8d:d03b:2059:9af5:f2c6:64a] has joined #shogun11:05
-!- goksinen [~goksinen@2604:2000:c591:8400:8d3e:9aa9:1628:cfc6] has quit [Ping timeout: 240 seconds]11:08
@wikingkesslerfrost, hi12:06
-!- Hieronymous [~sudarshan@2405:204:6388:9da2:285f:f89f:8a1a:2855] has quit [Quit: Leaving]12:07
@wikingkesslerfrost, for mailing list to join12:07
kesslerfrost@wiking thanks for the github link :) but I actually want the link for mailing list to join :)12:08
@wikingk cool12:08
kesslerfrostwiking thanks again12:08
kesslerfrostand one more thing. I have to discuss about the project of "Applying Shogun to the Real World" should I discuss it here or on the mailing list? (if I'm not wrong you are one of the mentors for GSoC 2017 right?)12:10
@wikingkesslerfrost, mailing list is the best idea12:40
@wikingbecause then other people know about it12:40
@wikingnot only the mentors12:40
@wikingwe alway encourage people to discuss gsoc related stuff on the mailing list12:40
-!- kesslerfrost [~textual@2405:204:8d:d03b:2059:9af5:f2c6:64a] has quit [Ping timeout: 240 seconds]13:13
-!- shogitter [] has joined #shogun13:26
-!- HeikoS [] has joined #shogun14:01
-!- mode/#shogun [+o HeikoS] by ChanServ14:01
@wikingHeikoS, y014:03
@wikingHeikoS, got the invite14:03
@wikingthnx heaps for pushing this14:03
@wikingHeikoS, ok so since i'm sooo near to finish up with the cloud14:04
@wikinghow about we do a release14:04
@wikingafter the cloud is online?14:04
CaBais there any way to tell sg to put all available parallelism into the machine backend and not to use any for kernel matrix computation for example?14:04
@wikingCaBa, heheh14:05
CaBahola :)14:05
@wikingCaBa, sorry i couldn't answer your answer before14:05
@wikingso no not atm14:05
@wikingif you want you need to define openmp groups14:05
@wikingand then you can do this trick14:05
@wikingi reckon it's not that complicated14:05
@HeikoSwiking: yes up for it14:05
@wikingCaBa, ok i used the wrong term14:06
@wikingCaBa, openmp team14:06
@sukeyPull Request #3642 "Add predictive variance to gp examples"  synchronized by karlnapf -
CaBawiking: ok, i'll look at that. how does shogun currently distribute parallelism?14:06
@wiking#pragma omp14:07
@wikingno teams14:07
@wikingthere's 1 big team14:07
@wikingand that shares all the threads14:07
@HeikoSwiking: i like it that scala is now part of build14:08
@HeikoSand the integration tests all pass :)14:08
@HeikoSand that it was very little work to add it14:08
@wikingdid that pr get rebased?14:08
@wikingbut seems yes14:09
@wikinghave u merged it? :)14:09
@wikingbtw we can do the same for all jvm language14:09
CaBathat parallel x-val that Saurabh7_ (?) implemented - it's still not in develop?14:09
@wikingCaBa, ah it was14:09
@wikingbut then we reverted14:09
@wikingbut i was thinking to re-revert14:09
CaBayes, that's the last thing i heard14:09
@wikinghaven't touched it since sorry14:09
@wikingHeikoS, groovy, kotlin, clojure14:10
@wikingthese could be all supportes14:10
CaBawiking: what's the PR / branch?14:10
@wikingas they are jvm14:10
@wikingCaBa, it was already in the pr14:10
@wikingi mean develop14:10
@HeikoSwiking: yep sure, I mean support is a strong term14:10
@HeikoSits part of our build thats all14:10
@wikingCaBa, 0666b6bc7df79028844c2ef6acd4c9b4f5d1947014:11
@wikingcheck that revert14:12
@wikingthat was actually the parallel xval14:12
@wikingi mean that commit is the removal of it :)14:12
CaBawiking: what was the problem with that imp?14:13
@sukeyPull Request #3633 "clean up language error and fluffy comments"  merged by karlnapf -
@sukeyNew Commit "Merge pull request #3633 from MikeLing/clean_up_KNN14:14
@sukeyclean up language error and fluffy comments" to shogun-toolbox/shogun by karlnapf:
@wikingCaBa, i thnk on the end there was no problem... :) we just did not see the forest from the tree14:14
@wikingso we killed the very first thing we saw :)14:14
@wikingbut hey14:15
@wikingsend in patches plz14:15
CaBawell if there is nothing to patch ^^14:15
@wikingas we are only the handful of us14:15
@wikingand we are dying here  :)14:15
@wikingahhaha there's too much to fix in this shit14:16
CaBai know. i do what i can. tricky thing is i have to finish my phd until yesterday ;)14:16
@wikingi know the feeling14:16
@wikinghave fun14:16
@wikingi think HeikoS is not far from finishing :OP14:16
@wikingto took me forever14:17
@wikingto finish14:17
CaBathesis writing offer whole new opportunities of procrastination O.O... spent two days perfecting a simple intro figure...14:17
@wikingbeen there14:19
@wikingdone that14:19
@HeikoSI have funding until at least september14:21
@HeikoSand then maybe 5 more months14:21
@HeikoSso all good here :)14:21
@wikingHeikoS, :>14:21
@HeikoShappy hacking14:21
@wikingcan you just put together your published articles14:21
@wikingor you have to write the whole thing?14:21
@wiking'put together' = staple :)14:21
@wikingsome places allow that14:22
@wikinglucky bastards14:22
CaBayeah, cumulative dissertation...14:22
@wikingafaik the worst place to make a phd is finland14:22
@wikingand then france :)14:22
@HeikoSwiking: yes stapling things14:22
@wikingbecause public defense is a real defense in a war :D14:22
@HeikoSonce it comes to that14:23
@HeikoSI mean I have to polish it14:23
@HeikoSand maybe extend here and there14:23
@HeikoSbut generally, I just take my papers and put them together14:23
@wikingah that is cool14:23
@wikingso any of you invented something brand new to this world... some ground shacking stuff? :)14:24
@HeikoSyou mean like the kernel adaptive metropolis hastings algorithm?14:24
@HeikoSof the kernel hamiltonian monte carlo?14:24
@HeikoSof kernel steinway?14:24
@HeikoSor nystrom approximate infinite dimensional exponential family?14:25
@wikingi meant like ROBOTS14:25
@HeikoSall pretty unmissable14:25
@HeikoSnope sorry ;)14:25
@HeikoSand also just did not get back to an offer to do a robotics internship for google brain in NY this summer ;)14:25
@wikingyou almost got it14:26
@wikingbut then you had to be smart about it :D14:26
@HeikoSas for google, for now I prefer gsoc to be my only place of being involved14:26
@wikingtoday i've managed to eliminate my spinnaker14:26
@wikingthat was all my contribution to this world14:26
@HeikoSwhat is spinnaker?14:27
@sukeyIssue #3643 "unify FITC and varDTC in Shogun" karlnapf added label: "entrance" -
@wikingthat big sail14:27
@sukeyIssue #3624 "Unable to build shogun" closed by karlnapf -
@wikingthose are the best14:28
@wikingwhen people come and say14:28
@wikingi cannot do it14:28
@HeikoSwiking: indeed :D14:29
@HeikoSwhat about your sail then?14:29
@wikingnow it's in 2 pieces14:29
@wikingbut yeah there are 2 other sails on the boat14:30
@wikingso there's still something to go with14:30
lisitsynare you like in sea or what? :)14:30
@wikinglisitsyn, now? no14:30
@wikingwas the hole weekend14:30
lisitsynI see14:31
@sukeyIssue #3615 "Unable to make due to svm_sgd error." closed by karlnapf -
lisitsynwiking: or agent trumpovsky as we call him14:33
lisitsynwiking: I think it was bill maher who said 'agent orange' which is even more ridiculous14:33
@wikingthis guy took too mcuh pills14:34
@wikingi think :)14:34
lisitsynorange ones?14:34
CaBauh, not everything is openmp in shogun?14:41
CaBaCKernel::get_kernel_matrix() uses pthreads14:42
@sukeyPull Request #3642 "Add predictive variance to gp examples"  synchronized by karlnapf -
@HeikoSCaBa: it does14:44
@HeikoSwill be part of gsoc to clean this up14:44
@HeikoSI am having second thoughts here14:57
@HeikoStheir goals are not ours14:57
@HeikoSso its more like a secretary for us14:57
@HeikoSnot sure that makes a lot of sense14:57
@HeikoSwiking: on the other hand:
@HeikoSthis aligns so well14:58
@wikingHeikoS, man we've been over this14:58
@wikingsoooo many times14:58
@wikingsooooooooo fuckng many times14:58
@wikingand yes14:58
@HeikoSwiking: I know and never resolved this right?14:59
@wikingfsf is a secretary14:59
@wikingyes we did14:59
@wikingwe did14:59
@wikingliek 2 times already14:59
@wikingonce in person14:59
@wikingonce via email14:59
@wikingfsf is not only a secretary story14:59
@wikingthey do help you14:59
@wikinga) with funding14:59
@wikingb) getting complicated legal stuff straigh14:59
@wiking(getting funding from country a but spending it in country b)15:00
@wikingwhat about that list plesae15:00
@wikingi dont fucking care15:00
@wikingthat those are some datascience projects15:00
@wikingand why do we have to be with them15:01
@wikingwe've been through this sooo many times15:01
@wikingthat if we join numfocus15:01
@HeikoSread the fsf list15:01
@wikingthen we loose control15:01
@wikingvery good library15:01
@wikingvery good library15:01
@HeikoSthe one i sent15:01
@HeikoSwe dont fall under "high priority"15:01
@HeikoSin their work15:01
@wikingi mean man15:02
@wikingwe are having now NOTHING15:02
@HeikoSthe other list is much much closer to what we do15:02
@wikingthey are more restricting15:02
@wikingwe have really been through this15:02
@HeikoSyes, but does that really hurt us?15:02
@wikingit does15:02
@wikingwe have been through this as well15:03
@HeikoSI dont care15:03
@wikingeven Soeren15:03
@HeikoSthat was wrong15:03
@wikingso i dont know what you dont care about15:03
@HeikoSgsoc stays as it is15:03
@wikingwhat was wrong?15:03
@HeikoSsee numpy15:03
@wikingsee scikit15:03
@wikingand before you fucking start about gpl15:03
@wikingboost is almsot list bsd15:03
@wikingit's really annoying15:04
@HeikoSI dont really care about license, I am just thinking pragmatic15:04
@wikingto go through this so many times15:04
@wikingand you are always coming back15:04
@wikingbecause for some reason15:04
@wikingyou dont agree15:04
@wikingand you really really want numfocus15:04
@wikingfor whatsoever reason15:04
@HeikoSits not like that15:04
@wikingyes it is15:04
@wikingthere hs been15:04
@wiking2 cases15:04
@wikingwhen you agreed15:04
-!- kesslerfrost [~textual@2405:204:100:83e6:5cea:8ffc:8825:a16a] has joined #shogun15:04
@HeikoSjust wonder whether the restrictive bit would be still ok for the alignment15:04
@wikingthat numfocus should not be15:04
@wikingin the picture15:04
@wikingnow after 2 months15:04
@wikingagain we are here where we started15:05
@wikingfor me not at all15:05
@wikingi dont like this15:05
@wikingthat they are restrictive15:05
@wikingbecause we are restricted15:05
@HeikoSyes I dont like it either15:05
@HeikoSbut concretely15:05
@wikingbut we didn't get anything15:05
@HeikoSwhat happens?15:05
@wikingyou joing15:05
@HeikoSsince gsoc is as before15:05
@wikingand then if ever want something else15:05
@wikingyou have to 'somehow' magically15:05
@wikingand NOBODY15:05
@wikingknows how's that working15:05
@wikingbecause that never happened before15:06
@wikingbut man15:06
@HeikoSits just this, that causes me not being happy with fsf15:06
@wikingyou have15:06
@wikingalways come up15:06
@wikingwith some dislikes with FSF15:06
@wikingfirst the gpl15:06
@HeikoSwiking: the leaving thing can be discussed if it is not clear right?15:06
@wikingthen this15:06
@wikingno its not fuck15:06
@HeikoSyeah gpl thing I dont like either15:06
@wikinghave you heard the fsf peopl15:06
@wikinghow fucking complicated thing is it15:06
@wikingto leave15:06
@wikingbut they been through that15:06
@wikingnot once15:06
@wikingwhen somebody wanted to start their own shit15:07
@wikingbecause they grew so big15:07
@wikingas otherwise15:07
@wikingyou can join whoever15:07
@wikingyou want as well15:07
@wiking(with fsf)15:07
@wikingboost is not gpl at all15:07
@wiking"The FSF maintains a list of "high priority projects" to which the Foundation claims that "there is a vital need to draw the free software community's attention".["15:07
@HeikoSits something else that concerns me15:07
@wikingit's an opinionated list15:07
@HeikoSI wonder how much we get from this15:07
@wikingstill that does not matter15:07
@wikingthat everything goes into the trash15:08
@HeikoSthats what I am comparing15:08
@wikingwhy would you even think this?15:08
@HeikoSlike the things that matter for us: funding for things15:08
@wikingwell until for example15:08
@wikingwe dont try FSF15:08
@wikingwhich is FOR FREEE15:08
@HeikoSand publicity to a community that cares15:08
@wikingthen we will not know exactly15:08
@HeikoSabout what we do15:08
@wikingif we go with somebody15:08
@wikingwho says BEFORE ANYTHING15:08
@wiking'sign here to lock yourself up15:08
@wikingyeah sorry i have ridden this horse before15:08
@wikingi do not fucking sign things like that15:09
@wikingit's bs15:09
@wikingand you know it15:09
@wikingif after year15:09
@wikingi have to sign15:09
@HeikoSnot so sure when I look at the software list15:09
@wikingit's fair15:09
@wikingbut right when you joiin15:09
@wikingi do not thing at all15:09
@wikingthat is fair15:09
@HeikoSfeel we would get more from being in the middle of that15:09
@wikingbecause you do not know what you are getting into15:09
@wikingso again15:09
@HeikoSspecially these days where a lot of attention is on the data science stuff and such lists15:10
@wikingi'm fine with the very loose thing with numfocus15:10
@wikingif you personally dislike fsf for whatever reason15:10
@wikinglet's look for another one15:10
@HeikoSI dont dislike fsf15:10
@wikingman it is obviously15:10
@wikingthere's always15:10
@wikingin the last 5 months15:10
@wikinga reasy15:10
@wiking*reason why not to join them15:10
@wikingor even to send out an email15:10
@wikingto ask them, hey what the fuck is up15:10
@wikingand for some reason15:10
@HeikoSthe reason i started talking15:10
@wikingwe are always back15:11
@wikingto square one15:11
@HeikoSis because I wrote the email, just waiting for a correction15:11
@HeikoSso all good, I will send15:11
@wikingthis is at least the 3rd time now15:11
CaBaHeikoS: with the pthread / openmp hybrid situation, is it possible that shogun spawns more threads than requested?15:11
@HeikoSjust discussing15:11
@wikingCaBa, could be15:11
@HeikoSwiking: well I dont agree with you I have to say15:11
@HeikoSlike the last times15:11
@wikingi know15:11
@wikingwe do not agree on this15:11
@wikingwhich is fine15:11
@wikingbut again exclusivity/restriction15:11
@wikingis a big thing15:11
@wikingand you have to agree on that15:11
@wikingi'm fine not to look at fsf15:11
@wikinglet's look for somebody else15:12
@wikingbut who is not restrictive15:12
@wikinglike numfocus15:12
@HeikoSfsf is fine15:12
@wikingthis is a free fucking thing15:12
@HeikoSits more the numfocus question15:12
@wikingi dont want ANYBODY15:12
@wikingto tell what we can do or not15:12
@wikingeven if it's a pair of socks15:12
@HeikoSwiking: btw what did you mean about sklearn earlier?15:12
@wikingyeah they are not having any representation15:13
@HeikoSthey are not even part of numfocus, they are just affiliated15:13
@HeikoSso thats bs15:13
@wikingyeah because15:13
@wikingthey are with another15:13
@wikingbullshit project15:13
@wikingcalled python15:13
@wikingwhich has restrictions as well15:13
@HeikoSso numpy is in two it seems15:13
@HeikoSpsf and numfocus15:14
@HeikoSand they applied via psd15:14
@wikingyeah ok so15:14
@wikinghow can they be in both15:14
@wikingand again there's this less restrictive thing with numfocus15:15
@wikingand i told you many times we can have that anytime15:15
@HeikoSno numpy is fully in there15:15
@HeikoSbut still applied via psf15:15
@HeikoSsince it is in there too15:15
@HeikoScheck the list15:15
@wikingok so how is that possible?15:15
@wikingwhen it shouldnt be15:15
@wikingor they just got a pass on this15:15
@wiking(which i highly suspect)15:15
@wikingsince in the agreement we've seen15:15
@wikingit's very very clear15:15
@wikingabout the exclusiveness15:16
@HeikoSI have put that in the email15:16
@HeikoSso askiung15:16
@HeikoSwiking: lea will proof read and clarity check the one in the googledoc and then ill send it off15:16
@HeikoSfsf will get the same email more or less15:17
@HeikoSThen we can go from there15:17
@HeikoSwiking: you ever set up a donation thing on a website?15:17
@HeikoSwiking: once we are in some foundation, then we should have that15:18
@wikingsince we still dont have foundation15:18
@wikingi have not gone down on that path15:18
@wikingshould be easy15:18
@wikingbut that means you'll have to pay somebody15:18
@wikingwho does the paper works for this15:18
@wikingwhich in case of shogun15:19
@wikingi rather think that it's gonna be a deficit story15:19
@HeikoSyep I know15:19
@HeikoSwiking: im thinking this:15:19
@wikingi less donation than the actual money we have to pay for the paperworks15:19
@wikinghence that's why we started to discuss this story15:19
@wikingwith the foundations15:19
@HeikoSwiking: when i asked the numfocus guys for funding for the hackathon 2 years ago, they said they would rather sponsor affiliated projects15:20
@HeikoSwiking: so now I wonder how can sponsor hackathons in the longer run15:20
@HeikoSI mean next one maybe gunnar can do it15:20
@HeikoSbut generally?15:20
@HeikoSwhere to get the 10k from?15:21
@HeikoSanyways, that all comes later ... I was mostly thinking loud15:21
@wikingi reckon15:21
@wikingif we could get somebody15:21
@wikingstart using this seriously15:21
@wikingwe could get some $15:21
@wikingas like 'sponsor'15:21
@wikingtill then it's all this15:21
@HeikoSwiking: yeah good point15:21
@wikingpersonal/gsoc funding15:22
@wikingwhich reminds me; sonney2k when you have the chance could you wire me the foss2017 tix from gsoc15:23
@HeikoSwiking: ok gotta go, see you later15:24
@HeikoSwiking: let me know about cloud and also if anything needs help15:25
@HeikoSotherwise ill continue with linalg stuff15:25
@wikingcheck that15:25
@HeikoSand experimenting with tags integration15:25
@wikingit's a bit demoralising15:25
@HeikoSits a bit more manpower ;)15:25
@wikingso bu tyeah i'll ping u15:25
@wikingwhen you could test the cloud15:25
@HeikoSwiking: this xla is the way to do it15:25
@HeikoSso we can use it as a backend when they are done15:25
@HeikoSthere is no way we could have built this even we had the idea for years15:26
@HeikoSbecause we are only 2.5 ppl ;)15:26
-!- HeikoS [] has quit [Ping timeout: 260 seconds]15:31
-!- HeikoS [] has joined #shogun15:31
-!- mode/#shogun [+o HeikoS] by ChanServ15:31
@HeikoSwiking: btw re cloud15:33
@HeikoSshall put the link already in the email? :D15:34
@wikingnot ready yet15:34
@wikingi mean yeah i wasted a lot of personal time on it15:35
@wikingbut cannot make things faster15:35
@HeikoSwiking: this wasnt a push ;)15:35
-!- Hieronymous [~sudarshan@2405:204:6388:9da2:285f:f89f:8a1a:2855] has joined #shogun15:35
-!- sudarshan_ [~sudarshan@2405:204:6388:9da2:285f:f89f:8a1a:2855] has joined #shogun15:35
-!- sudarshan__ [~sudarshan@2405:204:6388:9da2:285f:f89f:8a1a:2855] has joined #shogun15:37
-!- sudarshan_ [~sudarshan@2405:204:6388:9da2:285f:f89f:8a1a:2855] has quit [Client Quit]15:37
-!- travis-ci [] has joined #shogun15:42
travis-ciit's Heiko Strathmann's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun:
-!- travis-ci [] has left #shogun []15:42
-!- HeikoS [] has quit [Ping timeout: 260 seconds]15:43
-!- sudarshan__ [~sudarshan@2405:204:6388:9da2:285f:f89f:8a1a:2855] has quit [Ping timeout: 240 seconds]16:06
-!- Hieronymous [~sudarshan@2405:204:6388:9da2:285f:f89f:8a1a:2855] has quit [Ping timeout: 240 seconds]16:06
-!- Hieronymous [~sudarshan@] has joined #shogun16:56
-!- sudarshan_ [~sudarshan@2405:204:6388:9da2:285f:f89f:8a1a:2855] has joined #shogun16:57
-!- kesslerfrost [~textual@2405:204:100:83e6:5cea:8ffc:8825:a16a] has quit [Ping timeout: 240 seconds]17:11
-!- sudarshan_ [~sudarshan@2405:204:6388:9da2:285f:f89f:8a1a:2855] has quit [Quit: Leaving]17:22
-!- Hieronymous [~sudarshan@] has quit [Ping timeout: 240 seconds]17:42
-!- Hieronymous [~sudarshan@2405:204:6388:9da2:285f:f89f:8a1a:2855] has joined #shogun17:54
CaBawiking: you still around?18:22
-!- HeikoS [] has joined #shogun18:32
-!- mode/#shogun [+o HeikoS] by ChanServ18:32
@sukeyPull Request #3642 "Add predictive variance to gp examples"  merged by karlnapf -
@sukeyNew Commit "Merge pull request #3642 from karlnapf/develop18:35
@sukeyAdd predictive variance to gp examples" to shogun-toolbox/shogun by karlnapf:
-!- lambday [6a3314c2@gateway/web/freenode/ip.] has joined #shogun18:50
-!- mode/#shogun [+o lambday] by ChanServ18:50
-!- kesslerfrost [~textual@] has joined #shogun18:55
-!- Hieronymous [~sudarshan@2405:204:6388:9da2:285f:f89f:8a1a:2855] has quit [Ping timeout: 240 seconds]19:03
@HeikoSwiking: you around?19:23
-!- suhas2go [uid201652@gateway/web/] has quit [Quit: Connection closed for inactivity]19:26
-!- travis-ci [] has joined #shogun19:28
travis-ciit's Heiko Strathmann's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun:
-!- travis-ci [] has left #shogun []19:28
CaBaHeikoS: is libsvm parallel internally or is all the libsvm related parallelism in shogun?19:38
@HeikoSCaBa: I think only the kernel bits are parallel or?19:44
@HeikoSI dont know, shall I check?19:44
CaBaHeikoS: compute_Q is parallel shogun-side, yes19:44
CaBaHeikoS: that's what you mean?19:45
@HeikoSCaBa: whats the question? :)19:47
CaBa:D sorry19:47
CaBaHeikoS: if that's all, libsvm parallel wise... or if there is more upstream in libsv19:48
@HeikoSI dont know19:55
@HeikoSnot sure19:55
@HeikoSgrep for "thread"19:55
@HeikoSand "openmp"19:55
@HeikoS :)19:55
@sukeyNew Commit "added skip doxygen macro for internal classes" to shogun-toolbox/shogun by lambday:
@sukeyNew Commit "Merge branch 'feature/bigtest' into develop" to shogun-toolbox/shogun by karlnapf:
shogun-buildbotbuild #137 of trusty - libshogun - viennacl is complete: Failure [failed test]  Build details are at  blamelist: Heiko Strathmann <>20:08
shogun-buildbotbuild #119 of deb1 - libshogun is complete: Failure [failed test]  Build details are at  blamelist: Heiko Strathmann <>20:15
@sukeyNew Commit "Update NEWS" to shogun-toolbox/shogun by lambday:
-!- HeikoS [] has quit [Ping timeout: 260 seconds]20:28
-!- kesslerfrost [~textual@] has quit [Quit: kesslerfrost]20:42
-!- travis-ci [] has joined #shogun20:56
travis-ciit's Heiko Strathmann's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun:
-!- travis-ci [] has left #shogun []20:56
-!- lambday [6a3314c2@gateway/web/freenode/ip.] has quit [Ping timeout: 260 seconds]20:58
-!- travis-ci [] has joined #shogun21:38
travis-ciit's Soumyajit De's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun:
-!- travis-ci [] has left #shogun []21:38
CaBafor an object to be clone() able, it must contain only SGObject based members?23:40
CaBaor is there a way can have a pointer to a non-primitive, non-sgobject object as a member?23:44
--- Log closed Mon Feb 20 00:00:32 2017