Open in new window / Try shogun cloud
--- Log opened Mon Jun 02 00:00:14 2014
-!- PirosB3 [~pirosb3@host80-223-dynamic.246-95-r.retail.telecomitalia.it] has joined #shogun01:21
-!- PirosB3 [~pirosb3@host80-223-dynamic.246-95-r.retail.telecomitalia.it] has quit [Quit: PirosB3]02:38
-!- kislay [~Abhijeet@103.240.205.180] has joined #shogun02:40
shogun-buildbot_build #17 of nightly_fedora is complete: Failure [failed git]  Build details are at http://buildbot.shogun-toolbox.org/builders/nightly_fedora/builds/17  blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com>, khalednasr <k.nasr92@gmail.com>03:00
-!- kislay [~Abhijeet@103.240.205.180] has quit [Ping timeout: 252 seconds]03:59
-!- pickle27 [~pickle27@192-0-136-118.cpe.teksavvy.com] has quit [Remote host closed the connection]06:15
-!- lisitsyn1 [~qdrgsm@5-28-22-111.clients.tlt.100megabit.ru] has quit [Quit: Leaving.]07:27
-!- kislay [~Abhijeet@103.240.206.249] has joined #shogun08:05
-!- PirosB3 [~pirosb3@host80-223-dynamic.246-95-r.retail.telecomitalia.it] has joined #shogun08:52
-!- PirosB3 [~pirosb3@host80-223-dynamic.246-95-r.retail.telecomitalia.it] has quit [Client Quit]08:52
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun09:16
shogun-notifier-shogun: Parijat Mazumdar :develop * 66a5d02 / src/shogun/multiclass/tree/C45ClassifierTree.cpp: https://github.com/shogun-toolbox/shogun/commit/66a5d0206a644db671265bda697cc144e095f4e209:16
shogun-notifier-shogun: coverity scan defects fixed09:16
shogun-notifier-shogun: Fernando Iglesias :develop * 1436ba2 / src/shogun/multiclass/tree/C45ClassifierTree.cpp: https://github.com/shogun-toolbox/shogun/commit/1436ba2fb1c3e254d38e284d490af8b7f9a49cd609:16
shogun-notifier-shogun: Merge pull request #2274 from mazumdarparijat/C4509:16
shogun-notifier-shogun:09:16
shogun-notifier-shogun: coverity scan defects fixed09:16
-!- PirosB3 [~pirosb3@host80-223-dynamic.246-95-r.retail.telecomitalia.it] has joined #shogun09:35
thoralfHallo.09:42
-!- lambday [67157f4c@gateway/web/freenode/ip.103.21.127.76] has joined #shogun10:13
@besser82thoralf, Guten Morgen!10:16
thoralfHey Besser10:21
-!- PirosB3 [~pirosb3@host80-223-dynamic.246-95-r.retail.telecomitalia.it] has quit [Quit: PirosB3]10:50
-!- PirosB3 [~pirosb3@host80-223-dynamic.246-95-r.retail.telecomitalia.it] has joined #shogun11:27
-!- PirosB3 [~pirosb3@host80-223-dynamic.246-95-r.retail.telecomitalia.it] has quit [Quit: PirosB3]12:03
shogun-buildbot_build #326 of debian wheezy - memcheck is complete: Failure [failed memory check]  Build details are at http://buildbot.shogun-toolbox.org/builders/debian%20wheezy%20-%20memcheck/builds/326  blamelist: Fernando Iglesias <fernando.iglesiasg@gmail.com>, Parijat Mazumdar <mazumdarparijat@gmail.com>12:12
-!- kislay [~Abhijeet@103.240.206.249] has quit [Quit: Leaving]12:13
-!- lisitsyn [~lisitsin@mxs.kg.ru] has joined #shogun12:14
-!- PirosB3 [~pirosb3@host80-223-dynamic.246-95-r.retail.telecomitalia.it] has joined #shogun12:15
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout]12:16
-!- thoralf_ [~thoralf@91-66-177-150-dynip.superkabel.de] has joined #shogun12:52
-!- thoralf [~thoralf@91-66-177-150-dynip.superkabel.de] has quit [Read error: Connection reset by peer]12:52
-!- zxtx [~zv@c-98-223-196-32.hsd1.in.comcast.net] has quit [Ping timeout: 252 seconds]14:13
-!- lambday [67157f4c@gateway/web/freenode/ip.103.21.127.76] has quit [Ping timeout: 240 seconds]14:17
-!- khalednasr [~k.nasr92@41.69.213.164] has joined #shogun14:30
-!- zxtx [~zv@129-79-241-111.dhcp-bl.indiana.edu] has joined #shogun14:52
-!- thoralf_ is now known as thoralf15:04
-!- kislay [~Abhijeet@103.240.205.197] has joined #shogun16:12
-!- pickle27f [~pickle27@216.191.231.218] has joined #shogun16:12
-!- pickle27f [~pickle27@216.191.231.218] has quit [Quit: This computer has gone to sleep]17:25
-!- HeikoS [~heiko@nat-164-229.internal.eduroam.ucl.ac.uk] has joined #shogun17:26
-!- mode/#shogun [+o HeikoS] by ChanServ17:26
-!- HeikoS [~heiko@nat-164-229.internal.eduroam.ucl.ac.uk] has quit [Ping timeout: 260 seconds]17:32
-!- pickle27f [~pickle27@216.191.231.218] has joined #shogun17:33
-!- lambday [67157f4c@gateway/web/freenode/ip.103.21.127.76] has joined #shogun17:37
-!- pickle27f [~pickle27@216.191.231.218] has quit [Quit: This computer has gone to sleep]17:50
-!- khalednasr [~k.nasr92@41.69.213.164] has quit [Ping timeout: 240 seconds]18:06
-!- lisitsyn [~lisitsin@mxs.kg.ru] has quit [Quit: Leaving.]18:11
-!- HeikoS [~heiko@host-92-24-110-177.ppp.as43234.net] has joined #shogun18:18
-!- mode/#shogun [+o HeikoS] by ChanServ18:18
-!- PirosB3 [~pirosb3@host80-223-dynamic.246-95-r.retail.telecomitalia.it] has quit [Quit: PirosB3]18:22
-!- pickle27f [~pickle27@216.191.231.218] has joined #shogun18:55
lambdayHeikoS: there?19:46
-!- thor|home [~thoralf@91-65-141-142-dynip.superkabel.de] has joined #shogun20:09
-!- PirosB3 [~pirosb3@host80-223-dynamic.246-95-r.retail.telecomitalia.it] has joined #shogun20:15
thor|homeHeyho.20:20
-!- HeikoS [~heiko@host-92-24-110-177.ppp.as43234.net] has quit [Ping timeout: 264 seconds]20:23
-!- HeikoS [~heiko@dab-hlw1-h-29-10.dab.02.net] has joined #shogun20:58
-!- mode/#shogun [+o HeikoS] by ChanServ20:58
-!- PirosB3 [~pirosb3@host80-223-dynamic.246-95-r.retail.telecomitalia.it] has quit [Quit: PirosB3]20:59
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun21:16
shogun-notifier-shogun: abinashpanda :develop * 1414741 / / (8 files): https://github.com/shogun-toolbox/shogun/commit/14147416a19e93f2dc1a493f6d0fa1434c08768921:16
shogun-notifier-shogun: CMultilabelAccuracy class created along with modifications of CStructuredAccuracy and CMultilabelSOLabels for multi-label classification accuracy21:16
shogun-notifier-shogun: abinashpanda :develop * 6eb5c8d / examples/undocumented/libshogun/ (2 files): https://github.com/shogun-toolbox/shogun/commit/6eb5c8d323e4a0b05887422a05b22bc74001b8dc21:16
shogun-notifier-shogun: so_multilabel.cpp example added for multilabel classification21:16
shogun-notifier-shogun: tklein23 :develop * a4e9550 / / (10 files): https://github.com/shogun-toolbox/shogun/commit/a4e95507f767d9622b73cf572e458e50f36cd84021:16
shogun-notifier-shogun: Merge pull request #2273 from abinashpanda/develop21:16
shogun-notifier-shogun:21:16
shogun-notifier-shogun: CStructuredAccuracy class modified to compute accuracy for multilabel classification21:16
shogun-buildbot_build #2988 of deb1 - libshogun is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb1%20-%20libshogun/builds/2988  blamelist: abinashpanda <abinash.panda.ece10@itbhu.ac.in>, tklein23 <tklein23@users.noreply.github.com>21:24
-!- HeikoS [~heiko@dab-hlw1-h-29-10.dab.02.net] has quit [Quit: Leaving.]21:35
-!- Tonatiuh_ [4d08d476@gateway/web/freenode/ip.77.8.212.118] has joined #shogun21:39
-!- Tonatiuh_ [4d08d476@gateway/web/freenode/ip.77.8.212.118] has left #shogun []21:42
-!- zxtx [~zv@129-79-241-111.dhcp-bl.indiana.edu] has quit [Ping timeout: 260 seconds]22:02
-!- lisitsyn [~qdrgsm@80.252.20.67] has joined #shogun22:16
-!- zxtx [~zv@129-79-241-150.dhcp-bl.indiana.edu] has joined #shogun22:22
-!- pickle27f [~pickle27@216.191.231.218] has quit [Quit: This computer has gone to sleep]22:25
-!- pickle27f [~pickle27@216.191.231.218] has joined #shogun22:27
-!- iglesiasg [~iglesiasg@524AE0A7.cm-4-3d.dynamic.ziggo.nl] has joined #shogun22:35
-!- mode/#shogun [+o iglesiasg] by ChanServ22:35
kislayhey iglesiasg :)23:07
@iglesiasgsup kislay23:07
kislayHi there23:07
kislayI have a problem. http://pastebin.com/9CKwYSei .23:08
kislayiglesiasg, check it . I couldn't find anyways to get a workaround this.23:09
@iglesiasgkislay, get_name?23:10
@iglesiasgI just read the first part so far23:10
@iglesiasgok23:10
kislayiglesiasg, get_name does exactly this!!23:11
@iglesiasgso you mean finding out that it is a float32_t, not a CDenseFeatures, right?23:11
kislayiglesiasg, yes23:11
@iglesiasgkislay, ah ok, maybe not23:11
@iglesiasgkislay, well, then get_name doesn't do that, I believe :)23:11
@iglesiasgget_name would give you a string DenseFeatures or CDenseFeatures IIRC23:11
kislayyeah, exactly23:11
kislayI asked pickle27f about this too. He's been searching along with me :)23:12
@iglesiasgkislay, mmm the way I come up with to do that is to specialise the method23:13
@iglesiasglike one for CDenseFeatures<float32_t>, another for CDenseFeatures<float64_t> etc23:13
@iglesiasgbut that is a pain23:13
kislayyup. we want to escape that.23:13
@iglesiasgmmm but wait23:13
pickle27fyeah I thought of that too23:13
@iglesiasgwhat if you just make the method templated?23:13
kislaywhich method?23:14
pickle27figlesiasg, I think it might already need to be templates based on the output we desire23:14
pickle27ffor example I think this object should be able to take a DenseFeature Float64 and return a cv::Mat float32 if we want it to23:14
@iglesiasgI see23:16
@iglesiasglisitsyn, what do you think? ^23:16
lisitsynoh maan :D23:16
pickle27fI thought of overloading the method calls too, if we have to we can define a ton of this methods which then call out to a single method with the actually implementation23:16
lisitsyngotta dive into the discussion23:16
pickle27fand we pass a flag indicating the type23:16
@iglesiasglisitsyn, it is a short one :)23:16
lisitsynok you have23:17
pickle27fdoes SGMatrix not know its own type internally?23:17
lisitsynCDenseFeatures<T>23:17
lisitsynand wanna know T23:17
pickle27fmaybe can just expose a getter for the typer23:17
kislayyeah!23:17
kislaylisitsyn, yup23:17
lisitsynit does but in C++ a class is not a first-class citizen23:18
pickle27fright23:18
pickle27fhow does opencv return their internal types?23:18
lisitsynwhy do we need to know this type?23:18
lisitsynpickle27f: enum23:18
pickle27fwe basically need the same thing23:18
lisitsynyes23:18
lisitsynenums or template specialization23:18
pickle27fbecause opencv to shogun is done and super smooth, its going back that isn't as awesome right now23:18
lisitsynokay sorry I am mentally disabled lately23:19
lisitsyn:D23:19
lisitsyninput: CDenseFeatures<T>23:19
lisitsynwe need what?23:19
@iglesiasgsomething like the enum in opencv for our templated features sounds good to me23:19
lisitsynSGMatrix<T>?23:19
lisitsynkislay: do you need function definition or some internals?23:21
lisitsynwhat's input/output?23:21
kislayinput is the CDenseFeature object.23:21
lisitsynalright23:22
lisitsynwe need opencv mat?23:22
kislayand we are trying to get SGMatrix...then opencv mat23:22
kislayyeah23:22
lisitsyngot it23:22
pickle27fwe just need to know the type of the Shogun obj so we can do the conversion properly23:22
lisitsynI am pretty sure we have it already23:22
lisitsyngive me a second23:22
lisitsynget_feature_type23:23
lisitsynreturns EFeatureType23:23
lisitsyndispatch it23:23
lisitsyndefined in shogun/features/FeatureTypes.h23:23
lisitsynF_INT, F_WORD blabla23:23
lisitsynkislay: pickle27f: does this solve the problem?23:24
kislaylisitsyn, 2 mins23:24
lisitsynsee shogun/features/DenseFeatures.cpp:640 for type/enum value correspondence23:26
pickle27fthat sounds like it will work23:26
kislayyeah23:26
pickle27fwe don't have anything similar for SGMatrix though right>23:26
pickle27fwe should add something23:26
lisitsynyeah sure23:27
lisitsynmakes sense to add23:27
shogun-notifier-shogun: Parijat Mazumdar :develop * 084fc09 / src/shogun/multiclass/tree/ (3 files): https://github.com/shogun-toolbox/shogun/commit/084fc09c3856e68df78112729201197219f1080823:27
shogun-notifier-shogun: cross validation pruning added + changes in clone_tree method in TreeMachine23:27
shogun-notifier-shogun: Parijat Mazumdar :develop * f8245b5 / / (4 files): https://github.com/shogun-toolbox/shogun/commit/f8245b56df31472e6a97b03e98dbdebccab3699723:27
shogun-notifier-shogun: CART cv_prune test added, TreeMachine clone_tree refactored23:27
shogun-notifier-shogun: Fernando Iglesias :develop * 5040da2 / / (4 files): https://github.com/shogun-toolbox/shogun/commit/5040da259ff6c9268b7f5bec65b7bd1ce12753ec23:27
shogun-notifier-shogun: Merge pull request #2261 from mazumdarparijat/CART23:27
shogun-notifier-shogun:23:27
shogun-notifier-shogun: cross validation pruning added + changes in clone_tree method in TreeMachine23:27
lisitsyn(if it is missed)23:27
lisitsyniglesiasg: out of curiousity23:27
lisitsyniglesiasg: is parijat planning to work on some gradient boosted regression trees?23:28
@iglesiasglisitsyn, yeah?23:28
pickle27falso guys23:28
@iglesiasglisitsyn, I recall that from his proposal23:28
pickle27fis sonney2k around these days?23:28
lisitsyniglesiasg: I am gonna be an expert with GBRT23:28
lisitsyn:D23:28
lisitsynpickle27f: barely online :)23:28
pickle27fI asked him for something for the website a few days ago and haven't heard anything23:28
pickle27falso23:28
pickle27fIm working on a total new website, Im almost ready to demo it23:28
lisitsynpickle27f: really?23:29
lisitsynnew look or?23:29
pickle27fnew look and new backend23:29
lisitsynnot django?23:29
pickle27fI was tired of all the really sketchiness in the current site23:29
pickle27fyeah new site is flask23:29
pickle27fno db23:29
lisitsynflassk23:29
lisitsyn:)23:29
lisitsynthat's cool23:29
@iglesiasglisitsyn, that's great! You can maybe give some suggestions when Parijat works on it23:29
pickle27fthere is no reason for us to have a db23:29
pickle27falso you can actually run this site locally23:29
pickle27fbut we have a lot of reliance on specific paths on the server that runs our site23:30
lisitsyniglesiasg: I think I will study it during autumn :D23:30
pickle27fbut at least now I've pulled them into constants so we can emulate them locally23:30
lisitsynpickle27f: I had some bad experience with flask though :D23:30
lisitsynnever managed to control sessions properly23:30
pickle27fahhh23:30
pickle27fshouldn't matter for our site though23:31
lisitsynthought it was like my university assignment so doesn't matter23:31
pickle27fwe won't have sessions or anything23:31
lisitsynpickle27f: what's the look based on?23:31
pickle27fawww pfft what the heck might as well just show you guys where I'm at23:31
lisitsynhah23:31
pickle27fhttp://shogun-web.herokuapp.com/23:32
lisitsynbootstrap :)23:32
lisitsynIguess?23:32
pickle27fits just a free bootstrap theme23:32
pickle27fyeah23:32
pickle27fbut its better than the 7 stylesheets or whatever our current site has23:32
lisitsynyeah this is much better23:32
pickle27fplus you can hack and slash and use stack overflow for bootstrap issues23:32
pickle27fcause I don't think any of us really know css :p23:33
pickle27falso this site just uses twitter as a backend for "news"23:33
lisitsynpickle27f: yeah I am only ok with editing it :D23:33
pickle27fso we can save duplicate efforts on broadcasting things23:33
lisitsynpickle27f: that's awesome you are moving towards better look23:33
lisitsynI am like frustrated with the current way it looks23:34
pickle27fI really want to showcase the activeness of shogun23:34
pickle27flike we have commits everyday pretty much23:34
pickle27fthat and the examples23:34
pickle27fstill need to write a nice way to display github docs on our site for things like the install guide23:35
@iglesiasgpickle27f, fancy!23:36
pickle27fthanks!23:36
pickle27fits still missing some stuff but I am happy with the direction its going23:36
pickle27fthen we'll all be able to make easy changes just with PRs on github23:36
@iglesiasgvery convenient!23:38
pickle27fplus the site will actually be able to be ran locally so it will be easier to work with23:39
pickle27fnot really a big fan of django23:39
pickle27ftheir database integration stuff just isn't that nice sometimes23:39
@iglesiasgtime for me to abandon23:39
@iglesiasggood night guys, and goo talk!23:40
pickle27flater dude!23:40
-!- iglesiasg [~iglesiasg@524AE0A7.cm-4-3d.dynamic.ziggo.nl] has quit [Quit: Leaving]23:40
lisitsynpickle27f: we need vectorized thing for these chinese things23:42
pickle27fyeah this is all I have at the moment but you are right23:43
lisitsynpickle27f: I can do that23:43
pickle27fthat would be sweet - just email it to me whenever23:43
lisitsynyeah I expect once I get sober after my defence I'll get more time23:44
lisitsyn:D23:44
pickle27fgood luck with your defence man!23:48
pickle27fwhen is it?23:48
lisitsynpickle27f: the footer rocks23:48
pickle27fthanks!23:48
lisitsynpickle27f: 6th23:48
pickle27fso soon!23:48
lisitsynit is determined to be ok though23:48
pickle27fgood luck man!23:48
lisitsynI mean everything is done :D23:48
pickle27fanyways talk to you later, I got to run!23:48
lisitsynpickle27f: thanks23:48
lisitsynsee you23:49
-!- pickle27f [~pickle27@216.191.231.218] has quit [Quit: This computer has gone to sleep]23:59
--- Log closed Tue Jun 03 00:00:16 2014