IRC logs of #shogun for Thursday, 2016-05-19

« Wednesday, 2016-05-18 Index Friday, 2016-05-20 »
--- Log opened Thu May 19 00:00:02 2016
-!- HeikoS [~heiko@host-92-0-162-192.as43234.net] has joined #shogun00:30
-!- mode/#shogun [+o HeikoS] by ChanServ00:30
-!- HeikoS [~heiko@host-92-0-162-192.as43234.net] has quit [Quit: Leaving.]00:40
-!- besser82_ [~besser82@fedora/besser82] has joined #shogun01:26
-!- mode/#shogun [+o besser82_] by ChanServ01:26
-!- besser82_ [~besser82@fedora/besser82] has quit [Ping timeout: 276 seconds]02:24
shogun-buildbotbuild #2 of clang - thread analysis is complete: Failure [failed compile]  Build details are at http://buildbot.shogun-toolbox.org/builders/clang%20-%20thread%20analysis/builds/2  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>, Viktor Gal <viktor.gal@maeth.com>03:31
shogun-buildbotbuild #1 of clang - undefined behaviour analysis is complete: Failure [failed compile]  Build details are at http://buildbot.shogun-toolbox.org/builders/clang%20-%20undefined%20behaviour%20analysis/builds/1  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>, Viktor Gal <viktor.gal@maeth.com>03:33
-!- sanuj [~sanuj@45.127.55.123] has joined #shogun03:33
-!- sanuj [~sanuj@45.127.55.123] has quit [Ping timeout: 260 seconds]05:40
shogun-buildbotbuild #2 of memleak - valgrind is complete: Failure [failed memory check]  Build details are at http://buildbot.shogun-toolbox.org/builders/memleak%20-%20valgrind/builds/2  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>, Viktor Gal <viktor.gal@maeth.com>06:16
shogun-buildbotbuild #1006 of nightly_all is complete: Success [build successful]  Build details are at http://buildbot.shogun-toolbox.org/builders/nightly_all/builds/100606:29
shogun-buildbotbuild #1005 of nightly_none is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/nightly_none/builds/1005  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>, Viktor Gal <viktor.gal@maeth.com>06:40
-!- sanuj [~sanuj@45.127.55.123] has joined #shogun07:10
Saurabh7hello07:51
Saurabh7wiking: ping07:55
shogun-buildbotbuild #1134 of nightly_default is complete: Failure [failed test notebooks]  Build details are at http://buildbot.shogun-toolbox.org/builders/nightly_default/builds/1134  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>, Viktor Gal <viktor.gal@maeth.com>08:02
shogun-buildbotbuild #832 of deb4 - python3 is complete: Failure [failed configure]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb4%20-%20python3/builds/832  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>08:08
shogun-buildbotbuild #1222 of precise - libshogun is complete: Failure [failed compile]  Build details are at http://buildbot.shogun-toolbox.org/builders/precise%20-%20libshogun/builds/1222  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>08:17
shogun-buildbotbuild #8 of xenial - libshogun is complete: Failure [failed compile]  Build details are at http://buildbot.shogun-toolbox.org/builders/xenial%20-%20libshogun/builds/8  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>08:18
shogun-buildbotbuild #2872 of bsd1 - libshogun is complete: Failure [failed configure]  Build details are at http://buildbot.shogun-toolbox.org/builders/bsd1%20-%20libshogun/builds/2872  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>, Viktor Gal <vigsterkr@gmail.com>08:18
shogun-buildbotbuild #9 of xenial - libshogun is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/xenial%20-%20libshogun/builds/9  blamelist: Viktor Gal <vigsterkr@gmail.com>08:25
shogun-buildbotbuild #1223 of precise - libshogun is complete: Success [build successful]  Build details are at http://buildbot.shogun-toolbox.org/builders/precise%20-%20libshogun/builds/122308:28
shogun-buildbotbuild #226 of FC22 - libshogun is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/FC22%20-%20libshogun/builds/226  blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>08:30
shogun-buildbotbuild #227 of FC22 - libshogun is complete: Success [build successful]  Build details are at http://buildbot.shogun-toolbox.org/builders/FC22%20-%20libshogun/builds/22708:34
-!- leagoetz [~leagoetz@host-92-0-162-192.as43234.net] has joined #shogun08:49
shogun-buildbotbuild #833 of deb4 - python3 is complete: Success [build successful]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb4%20-%20python3/builds/83309:01
-!- leagoetz [~leagoetz@host-92-0-162-192.as43234.net] has quit [Remote host closed the connection]09:13
-!- shogun-buildbot [~shogun-bu@7nn.de] has quit [Quit: buildmaster reconfigured: bot disconnecting]09:35
-!- shogun-buildbot [~shogun-bu@7nn.de] has joined #shogun09:35
@wikingshogun-toolbox: force build --branch=develop 'trusty - libshogun - viennacl'09:36
@wikingshogun-buildbot: force build --branch=develop 'trusty - libshogun - viennacl'09:36
shogun-buildbotbuild forced [ETA 3m32s]09:36
shogun-buildbotI'll give a shout when the build finishes09:36
shogun-buildbotHey! build trusty - libshogun - viennacl #658 is complete: Success [build successful]09:38
shogun-buildbotBuild details are at http://buildbot.shogun-toolbox.org/builders/trusty%20-%20libshogun%20-%20viennacl/builds/65809:38
-!- sanuj [~sanuj@45.127.55.123] has quit [Ping timeout: 260 seconds]09:48
-!- sanuj [~sanuj@117.204.243.44] has joined #shogun10:00
-!- sanuj [~sanuj@117.204.243.44] has quit [Ping timeout: 276 seconds]10:07
-!- sanuj [~sanuj@117.204.243.44] has joined #shogun10:11
sanujlisitsyn, there?10:18
lisitsynyes10:19
sanujlisitsyn, https://gist.github.com/sanuj/56f03cd242473137fad851e68fa0f2c110:19
sanujwiking, ^10:19
sanujlist of abstract classes from shogun10:19
sanuji grepped for "()=0" and "() = 0"10:20
sanuji have made a markdown so we can decide which classes to keep in base shogun10:20
sanuji will convert into swig interface afterwards10:21
lisitsynbut what about virtual functions *with* arguments?10:21
sanujheiko also wanted to make some classifications for these classes10:22
sanujsome tree10:22
sanujoh10:22
sanuji'll add those10:22
sanuji missed them10:22
lisitsynits like (.*) = 010:22
lisitsynor something10:22
lisitsynand there also can be const10:22
sanujokay10:22
lisitsynvirtual f(int a) const = 0;10:22
sanuji see10:22
sanujwill try these10:22
sanujoh lots of classes need to be added then10:23
sanujlisitsyn, check it later then10:23
sanuji'll ping you when i've added all of them10:23
lisitsynyeah10:23
lisitsyncool10:23
sanujlisitsyn,  should i grep for anything else?10:23
@wikinghello hello10:24
lisitsynhey10:24
@wikingwhere is heiko?:)10:24
lisitsynsanuj: probably no10:24
sanujokay10:24
@wikinghahaha10:24
@wikinglisitsyn: do you have an answer10:24
@wikingfor that email on the mailing list10:24
lisitsynthere should be virtual and ()10:24
@wikingwhat's the diff between10:24
lisitsynwiking: basemulticlass?10:24
@wikingyes10:24
lisitsynlet me check10:24
sanujwiking, you are talking about the thing that i asked you earlier10:25
sanuj?10:25
lisitsynwiking: for some reason10:27
lisitsyne.g. TreeMachine is not MulticlassMachine10:27
lisitsynbut BaseMulticlassMachine10:27
lisitsynI guess it was too limitedt10:28
-!- sanuj [~sanuj@117.204.243.44] has quit [Ping timeout: 276 seconds]10:41
-!- Saurabh7 [Saurabh7@gateway/shell/panicbnc/x-txiyhlblsirgakfb] has quit [Ping timeout: 260 seconds]11:45
-!- Saurabh7 [Saurabh7@gateway/shell/panicbnc/x-nzmfsjhfxlaprvmf] has joined #shogun11:45
-!- besser82_ [~besser82@fedora/besser82] has joined #shogun11:47
-!- mode/#shogun [+o besser82_] by ChanServ11:47
-!- leagoetz [~leagoetz@nat-170-181.internal.eduroam.ucl.ac.uk] has joined #shogun12:04
-!- besser82_ [~besser82@fedora/besser82] has quit [Remote host closed the connection]12:19
-!- besser82 [~besser82@fedora/besser82] has joined #shogun12:23
-!- mode/#shogun [+o besser82] by ChanServ12:24
-!- leagoetz_ [~leagoetz@nat-170-181.internal.eduroam.ucl.ac.uk] has joined #shogun12:26
-!- HeikoS [~heiko@80.169.91.26] has joined #shogun12:30
-!- mode/#shogun [+o HeikoS] by ChanServ12:30
-!- leagoetz [~leagoetz@nat-170-181.internal.eduroam.ucl.ac.uk] has quit [Ping timeout: 260 seconds]12:30
-!- HeikoS [~heiko@80.169.91.26] has quit [Quit: Leaving.]12:42
-!- HeikoS [~heiko@80.169.91.26] has joined #shogun12:42
-!- mode/#shogun [+o HeikoS] by ChanServ12:42
-!- sanuj [~sanuj@117.204.242.93] has joined #shogun12:42
sanujthere are too many classes in shogun12:57
sanuj:)12:58
sanujlisitsyn, do i also include classes from lib/* for shogun-base?13:08
lisitsynsanuj: like?13:08
sanujtapkee13:08
sanujlib/external/*13:08
sanujlisitsyn, ^13:09
lisitsynsanuj: ah no13:10
sanujokay13:10
lisitsynsanuj: they are used by converters/13:10
-!- sanuj [~sanuj@117.204.242.93] has quit [Ping timeout: 276 seconds]13:40
-!- sanuj [~sanuj@117.204.242.93] has joined #shogun13:52
-!- HeikoS [~heiko@80.169.91.26] has quit [Quit: Leaving.]14:10
sanujlisitsyn, ping14:47
lisitsynsanuj: sup14:47
sanujhttps://gist.github.com/sanuj/56f03cd242473137fad851e68fa0f2c114:47
sanujupdated14:47
sanujlisitsyn, if you think it's okay then we can tell others and refactor this list14:48
lisitsynsanuj: well its huge14:49
sanujyeah14:50
-!- leagoetz_ [~leagoetz@nat-170-181.internal.eduroam.ucl.ac.uk] has quit []15:26
-!- mizari [~mizari@95-174-213-100.nts.su] has joined #shogun15:35
lisitsynsanuj: ok lets divide and conquer15:36
lisitsynsanuj: could you put things we are sure15:37
lisitsyninto another list there15:37
sanujlisitsyn, i'll make another gist for that then15:37
sanujwhat are we sure about15:38
sanujthe things you mentioned in your comment?15:38
lisitsynsanuj: you can start with my first comment15:39
sanujlisitsyn, cool15:41
sanujlisitsyn, i hope this list is not a waste then :)15:41
lisitsynsanuj: I think it is better to put them there15:41
lisitsynjust moving them from one list to another15:41
sanujokay15:41
lisitsynto keep all of them at the same place15:41
sanujlisitsyn, so all kernels and distances from my list15:42
sanujand so on15:42
-!- OXPHOS [9d8b131c@gateway/web/freenode/ip.157.139.19.28] has joined #shogun15:47
-!- besser82 [~besser82@fedora/besser82] has quit [Ping timeout: 250 seconds]15:49
-!- Saurabh7 [Saurabh7@gateway/shell/panicbnc/x-nzmfsjhfxlaprvmf] has quit [Ping timeout: 260 seconds]15:53
-!- Saurabh7 [Saurabh7@gateway/shell/panicbnc/x-zklydguxtatbrtlp] has joined #shogun15:53
@wikinguuuuuuuuueeeeeeeefiiii16:11
sanujwiking, ssup16:13
-!- HeikoS [~heiko@nat-228-166.internal.eduroam.ucl.ac.uk] has joined #shogun16:23
-!- mode/#shogun [+o HeikoS] by ChanServ16:23
sanujlisitsyn, there?16:30
sanujHeikoS, hi16:33
shogun-buildbotbuild #235 of deb1 - libshogun - PR is complete: Failure [failed cookbook]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb1%20-%20libshogun%20-%20PR/builds/235  blamelist: OXPHOS16:47
@wikingHeikoS: see my comment & react plz16:54
-!- sanuj [~sanuj@117.204.242.93] has quit [Ping timeout: 260 seconds]16:55
-!- sanuj [~sanuj@117.204.242.93] has joined #shogun17:24
-!- leagoetz [~leagoetz@nat-170-181.internal.eduroam.ucl.ac.uk] has joined #shogun18:02
sanujlisitsyn, do you want CRandomKitchenSinksDotFeatures also in the other list?18:20
lisitsynno18:20
sanujokay18:20
lisitsynthat's just dot features18:20
sanujdensefeatures and sparsefeatures are not abstract classes18:20
sanujbut derived from dotfeatures18:24
sanujlisitsyn, ^18:24
lisitsynsanuj: yes18:24
lisitsynthat's an exception18:24
sanujokay18:25
lisitsynas they are really important18:25
sanujyeah18:25
sanujlisitsyn, do you want all abstract class distances?18:26
sanujrealdistance18:26
sanujstringdistance18:26
sanujsparsedistance18:26
lisitsynsanuj: looks like18:26
lisitsynnot sure though18:26
lisitsynsanuj: lets keep just base distance18:26
lisitsynit looks like string/sparse/real is just a notion of features they are defined18:27
sanuji.e. CDistance?18:27
lisitsynon18:27
lisitsynyes18:27
sanujokay18:27
sanujlisitsyn,  do we need to confirm this with others?18:27
lisitsynsanuj: yeah it could be good to discuss that18:28
sanujHeikoS, wiking ping18:28
lisitsynit's really a major change18:28
@HeikoSsanuj: jojo18:28
@HeikoSwiking: checking18:28
sanujHeikoS, all abstract classes --> https://gist.github.com/sanuj/56f03cd242473137fad851e68fa0f2c118:28
@HeikoS?18:28
@HeikoScool checking18:28
@HeikoSlisitsyn: ^?18:29
sanujsaw18:29
lisitsynHeikoS: yes I know18:29
@HeikoSlisitsyn: what is a major change?18:30
lisitsynHeikoS: well defining base classes18:30
lisitsyn:D18:30
lisitsynit determines everything18:30
@HeikoSyes18:30
@HeikoSwe need a class diagram18:30
@HeikoSa list is not very intuitive :)18:30
@HeikoSyou know dia?18:31
sanujno18:31
lisitsyngraphviz is ok as well18:31
@HeikoSlambday used that for his crazy hierarchy in gsoc a few years agp18:31
sanuji'll have a look at them18:31
@HeikoSsanuj, lisitsyn I think this should be approached differently18:32
@HeikoStoo many details right now18:32
@HeikoSwhat are the base classes and their interfaces?18:32
@HeikoSthis is most important18:32
@HeikoSfor that, we need to write down algorithm classes18:32
@HeikoSthen, and only then, we can try to fit existing things in there18:32
@HeikoSand iterate18:32
lisitsynHeikoS: my idea is to make a few iterations to select classes we are sure about18:32
@HeikoSI think this leads to bad design18:33
@HeikoSas we are biased by the current class structure which is a disaster18:33
@HeikoSthe old CMachine interface for supervised learning is good,18:33
@HeikoSbut thats all thats good18:33
@HeikoSso rather think from algorithm classes18:33
@HeikoSnot concrete algorithms18:33
@HeikoSI am missing:18:33
@HeikoSCSupervisedLearning18:33
@HeikoSCUnsupervisedLearning18:33
lisitsynSemisupervised? :)18:34
@HeikoSwhere the first has fit(X,y), the second only has fit(X)18:34
sanujHeikoS, that's what your comment is about on the gist18:34
lisitsynHeikoS: would you go18:34
lisitsynfor fit/predict18:34
@HeikoSlisitsyn: dont care18:34
lisitsynthe sklearn gold standard?18:34
lisitsyn;)18:34
@HeikoSfit, apply18:35
@HeikoStrain, apply18:35
lisitsynfit predict18:35
lisitsyn:)18:35
@HeikoSfit, predict18:35
@HeikoSI dont catre18:35
sanujtrain predict sounds good :P18:35
lisitsynit goes against default names18:35
@HeikoSsanuj: but do you get what I mean when I say start general18:35
lisitsynnot good18:36
@HeikoSnot with classes that exist18:36
sanujHeikoS, yes i do18:36
@HeikoSI mean things like variational likelihood18:36
@HeikoSno point in even thinking about this thing18:36
@HeikoSway to specific18:36
@HeikoSmore important to think about the interface of a GP18:36
sanujthis will be a part of a plugin18:36
@HeikoSdifferent ways it can be used in18:36
@HeikoSOr even CLikelihoodModel18:36
@HeikoSthe core things to get right are:18:37
@HeikoSsupervised/unsupervised and some subclasses (not concrete algorithms yet)18:37
@HeikoSand the features18:37
sanujreinforcement?18:37
@HeikoSwe dont have that18:37
@HeikoSthen pick a single representative algo, and try to put it under the framework18:38
@HeikoSdraw class diagram examples18:38
@HeikoSwith the method signatures (not too detailed)18:38
@HeikoSCareful interface design18:38
@HeikoSthinking of what users might want to do, keep it clean18:38
@HeikoSetc etc18:38
sanujokay18:39
@HeikoSThe main one is CMachine for supervised learning18:39
@HeikoS(should be renamed)18:39
@HeikoSthe list I wrote in the comment is a pretty good start btw18:39
@HeikoSwhat would be the interfaces for these classes18:39
sanujcool, i'll start from there18:39
@HeikoSsanuj: great!18:40
sanujand then put the existing shogun algorithms in the correct category18:40
@HeikoSsanuj: can we discuss this over a diagram?18:42
@HeikoScan you make one in some online editor?18:42
@HeikoSsanuj: yes then that18:42
@HeikoSsanuj: and I would aim for rather fewer algorithms at the start, but a really good interface18:42
@HeikoSand then add things later18:42
sanujHeikoS, alright18:42
sanuji18:43
sanuji'll make a diagram and share that with you soon18:43
sanujby tomorrow18:43
@HeikoSsanuj: great!18:43
@HeikoSlooking forward to see it18:43
@HeikoSthis is a very fun part of the project18:43
sanujhaha18:44
sanujcool18:44
-!- OXPHOS [9d8b131c@gateway/web/freenode/ip.157.139.19.28] has quit [Ping timeout: 250 seconds]18:47
-!- Saurabh7 [Saurabh7@gateway/shell/panicbnc/x-zklydguxtatbrtlp] has quit [Ping timeout: 260 seconds]19:01
-!- Saurabh7 [Saurabh7@gateway/shell/panicbnc/x-hsyhoqqrfnnamdms] has joined #shogun19:03
-!- sanuj [~sanuj@117.204.242.93] has quit [Quit: Leaving]19:29
-!- besser82 [~besser82@fedora/besser82] has joined #shogun19:45
-!- mode/#shogun [+o besser82] by ChanServ19:45
@wikingHeikoS: oooooo19:50
@HeikoSwiking: jojo19:51
@HeikoSgotta leave soon19:51
@HeikoSanything urgent?19:51
@wikingvery19:54
@wikingyes19:54
@wiking:D19:54
@wikingso19:54
@wikingHeikoS: as you can see i've crunched a lot of bugs19:54
@wikingdunno how much i will be still able to do19:54
@wikingbut wont be able to make the relase tomorrow19:55
@wikingi'll make it tenitatively around sunday19:55
@wikinggood?19:55
-!- sonne|osx [~sonne@89.204.137.120] has joined #shogun19:55
-!- leagoetz [~leagoetz@nat-170-181.internal.eduroam.ucl.ac.uk] has quit [Remote host closed the connection]19:59
-!- sonne|osx [~sonne@89.204.137.120] has quit [Ping timeout: 276 seconds]20:00
-!- leagoetz [~leagoetz@nat-170-181.internal.eduroam.ucl.ac.uk] has joined #shogun20:01
-!- leagoetz [~leagoetz@nat-170-181.internal.eduroam.ucl.ac.uk] has quit [Remote host closed the connection]20:02
-!- leagoetz [~leagoetz@nat-170-181.internal.eduroam.ucl.ac.uk] has joined #shogun20:03
-!- leagoetz [~leagoetz@nat-170-181.internal.eduroam.ucl.ac.uk] has quit [Ping timeout: 260 seconds]20:07
-!- mizari [~mizari@95-174-213-100.nts.su] has quit [Quit: Leaving]20:17
@wikingargh20:19
@HeikoSwiking: good!20:33
@HeikoSvery good20:33
@HeikoSany news from gpl friend?20:33
-!- HeikoS [~heiko@nat-228-166.internal.eduroam.ucl.ac.uk] has quit [Quit: Leaving.]20:48
-!- yorkerlin [63e6385e@gateway/web/freenode/ip.99.230.56.94] has joined #shogun21:02
-!- yorkerlin [63e6385e@gateway/web/freenode/ip.99.230.56.94] has quit [Client Quit]21:03
-!- travis-ci [~travis-ci@ec2-54-87-10-142.compute-1.amazonaws.com] has joined #shogun21:07
travis-ciit's lambday's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/13150016621:07
-!- travis-ci [~travis-ci@ec2-54-87-10-142.compute-1.amazonaws.com] has left #shogun []21:07
-!- Saurabh7 [Saurabh7@gateway/shell/panicbnc/x-hsyhoqqrfnnamdms] has quit [Ping timeout: 260 seconds]22:03
-!- Saurabh7 [Saurabh7@gateway/shell/panicbnc/x-weoykeanaeysigkz] has joined #shogun22:03
-!- travis-ci [~travis-ci@ec2-54-163-125-129.compute-1.amazonaws.com] has joined #shogun22:21
travis-ciit's lambday's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/13150564522:21
-!- travis-ci [~travis-ci@ec2-54-163-125-129.compute-1.amazonaws.com] has left #shogun []22:21
--- Log closed Fri May 20 00:00:03 2016
« Wednesday, 2016-05-18 Index Friday, 2016-05-20 »

Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!