| --- Log opened Thu Apr 25 00:00:29 2013 | ||
| @sonney2k | wiking, then only disable them for osx | 00:08 |
|---|---|---|
| @sonney2k | wiking, or anything !linux | 00:08 |
| @sonney2k | the integration tests we have cover much more than the few tiny tests we currently have so they are invaluable and turning them of completely is no option | 00:09 |
| -!- makokal [~bokal@pD9EF09DC.dip0.t-ipconnect.de] has quit [Quit: makokal] | 00:44 | |
| -!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving] | 01:23 | |
| -!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout] | 01:26 | |
| -!- abinash [75ef5e6e@gateway/web/freenode/ip.117.239.94.110] has joined #shogun | 01:34 | |
| -!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has joined #shogun | 05:34 | |
| deerishi | whois @lisitsyn | 05:47 |
| deerishi | @lisitsyn:der? | 05:47 |
| deerishi | ? | 05:47 |
| -!- abinash [75ef5e6e@gateway/web/freenode/ip.117.239.94.110] has quit [Ping timeout: 245 seconds] | 06:26 | |
| -!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 08:05 | |
| -!- mode/#shogun [+o iglesiasg] by ChanServ | 08:06 | |
| -!- sids_aquarius_ [~sidi_@li400-124.members.linode.com] has quit [Ping timeout: 264 seconds] | 08:47 | |
| -!- sids_aquarius_ [~sidi_@li400-124.members.linode.com] has joined #shogun | 08:49 | |
| -!- sids_aquarius [~sidi_@li400-124.members.linode.com] has joined #shogun | 09:02 | |
| -!- sids_aquarius_ [~sidi_@li400-124.members.linode.com] has quit [Ping timeout: 246 seconds] | 09:02 | |
| -!- lisitsyn [~blackburn@46.20.65.68] has quit [Quit: Leaving.] | 09:07 | |
| -!- makokal [~bokal@pD9EF09DC.dip0.t-ipconnect.de] has joined #shogun | 09:37 | |
| -!- makokal [~bokal@pD9EF09DC.dip0.t-ipconnect.de] has quit [Quit: makokal] | 09:49 | |
| @iglesiasg | wiking: around? | 09:50 |
| -!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has quit [Ping timeout: 245 seconds] | 10:04 | |
| -!- lisitsyn [~lisitsin@mxs.kg.ru] has joined #shogun | 10:17 | |
| -!- mode/#shogun [+o lisitsyn] by ChanServ | 10:17 | |
| @lisitsyn | wiking: http://www.wikicfp.com/cfp/servlet/event.showcfp?eventid=29897©ownerid=38935 :D | 10:17 |
| @lisitsyn | would you like to visit that city I live in? ;) | 10:18 |
| -!- hoijui [~hoijui@dslb-088-074-121-064.pools.arcor-ip.net] has joined #shogun | 10:19 | |
| -!- sonne|work [~sonnenbu@sams-office-nat.tomtomgroup.com] has joined #shogun | 10:24 | |
| @lisitsyn | sonne|work: welcome back hehe | 10:24 |
| sonne|work | lisitsyn: hey hey | 10:25 |
| sonne|work | man what a neural network overoverhype! | 10:25 |
| sonne|work | I felt like in the early days of kernel machines! | 10:25 |
| @lisitsyn | sonne|work: haha because they work | 10:25 |
| @lisitsyn | for images | 10:25 |
| sonne|work | they work under very specific conditions | 10:26 |
| sonne|work | with lots and lots of data | 10:26 |
| sonne|work | and lots and lots of computing power | 10:26 |
| sonne|work | and they are one of the few methods that drastically benefit from GPUs | 10:26 |
| sonne|work | so I guess that is why | 10:26 |
| @lisitsyn | sonne|work: yes they work nice on gpus | 10:27 |
| sonne|work | but as usual no free lunch :D | 10:27 |
| @lisitsyn | of course | 10:27 |
| @lisitsyn | but it appears to be better than older stuff | 10:27 |
| @lisitsyn | for some tasks | 10:27 |
| @lisitsyn | iglesiasg: may be you want to visit samara in september? :) | 10:28 |
| sonne|work | lisitsyn: no it is the older stuff | 10:32 |
| sonne|work | it is the same stuff from the 80's or so but done with more data & computing power | 10:33 |
| @lisitsyn | sonne|work: true | 10:33 |
| sonne|work | and they still have local minima and need clever data normalization | 10:33 |
| sonne|work | suitable number of hidden layers | 10:33 |
| sonne|work | structure etc | 10:33 |
| sonne|work | I would say it didn't get easier but hey I guess one can say that about any method | 10:34 |
| @lisitsyn | sonne|work: so no more svms? :) | 10:35 |
| sonne|work | I saw cascades of decision / regression trees | 10:35 |
| sonne|work | and the usual stuff like logistic regression etc | 10:36 |
| sonne|work | and NN's | 10:36 |
| @iglesiasg | lisitsyn: oh nice a conference in your city :) | 10:38 |
| -!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has joined #shogun | 10:38 | |
| @lisitsyn | iglesiasg: yeah - there is some time to come up with something | 10:40 |
| @iglesiasg | lisitsyn, sonne|work : btw could you guys please this PR? | 10:40 |
| @iglesiasg | https://github.com/shogun-toolbox/shogun/pull/1012 | 10:40 |
| @lisitsyn | alright | 10:40 |
| @iglesiasg | it doesn't break anything new according to travis | 10:40 |
| @iglesiasg | https://travis-ci.org/shogun-toolbox/shogun/builds/6621915 | 10:40 |
| -!- shogun-notifier- [~irker@7nn.de] has joined #shogun | 10:41 | |
| shogun-notifier- | shogun: Fernando Iglesias :develop * 5a265c2 / src/shogun/machine/ (4 files): https://github.com/shogun-toolbox/shogun/commit/5a265c23b0db994224f946ba10f1acef83ebbc73 | 10:41 |
| shogun-notifier- | shogun: Setters/getters in structured machines. | 10:41 |
| shogun-notifier- | shogun: They are required to implement SO-SVMs from interfaces. | 10:41 |
| shogun-notifier- | shogun: Sergey Lisitsyn :develop * 7b4c255 / src/shogun/machine/ (4 files): https://github.com/shogun-toolbox/shogun/commit/7b4c255e3c5c2314867c142998e6902310063a6b | 10:41 |
| shogun-notifier- | shogun: Merge pull request #1012 from iglesias/feature/improve_structure | 10:41 |
| shogun-notifier- | shogun: | 10:41 |
| shogun-notifier- | shogun: Setters/getters in structured machines. | 10:41 |
| @iglesiasg | my python bindings for structured learning uses it and I think Patrick wanted to test something | 10:41 |
| @iglesiasg | lisitsyn: thanks! | 10:41 |
| vgorbati | lisitsyn: hi there | 10:48 |
| @lisitsyn | vgorbati: hey | 10:48 |
| -!- van51 [~van51@athedsl-320854.home.otenet.gr] has joined #shogun | 10:50 | |
| vgorbati | lisitsyn: it looks like you are in good mood for merging PR's - do you have time to review my PR too?:) https://github.com/shogun-toolbox/shogun/pull/1015 | 10:52 |
| @lisitsyn | vgorbati: could you please rebase it? | 10:53 |
| @lisitsyn | you have 4 merge commits here | 10:53 |
| vgorbati | lisitsyn: basically I need to call 'git rebase develop'? | 10:55 |
| @lisitsyn | no it won't help there | 10:56 |
| vgorbati | lisitsyn: and what will help? Previously I just called 'git pull upstream develop' from time-to-time to stay in touch with develop branch. | 10:59 |
| -!- travis-ci [~travis-ci@ec2-174-129-147-163.compute-1.amazonaws.com] has joined #shogun | 10:59 | |
| 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/6623316 | 10:59 |
| -!- travis-ci [~travis-ci@ec2-174-129-147-163.compute-1.amazonaws.com] has left #shogun [] | 10:59 | |
| -!- makokal [~bokal@eduroam-pool3-128.wlan.uni-bremen.de] has joined #shogun | 11:01 | |
| @iglesiasg | lisitsyn: have you seen the cool perceptron we have now? https://f.cloud.github.com/assets/1158006/414322/980ff0ae-ac14-11e2-9897-1dc19c22f9b8.png | 11:10 |
| @iglesiasg | :D | 11:10 |
| @lisitsyn | iglesiasg: is he drunk? | 11:10 |
| @lisitsyn | rosenblatt gets drunk | 11:11 |
| @iglesiasg | lisitsyn: haha not, just randomly initialized | 11:11 |
| @lisitsyn | I see | 11:11 |
| @iglesiasg | I used that figure for my report | 11:11 |
| @lisitsyn | vgorbati: always do 'git pull --rebase upststream develop' :) | 11:11 |
| @lisitsyn | okay lets not mess with that now | 11:11 |
| vgorbati | lisitsyn: so, just to make it clear for me, in future, when I want to get changes from upstream/develop I should run 'git pull --rebase upstream develop' instead of 'git pull upstream develop | 11:12 |
| vgorbati | ', right? | 11:13 |
| @lisitsyn | vgorbati: yes | 11:16 |
| @lisitsyn | it won't produce these merge commits | 11:16 |
| @lisitsyn | just rebases your commits on top of upstream branch | 11:16 |
| vgorbati | lisitsyn: as a result there wont be any 'merge ...' commits when I do PR? | 11:17 |
| vgorbati | lisitsyn: sorry, I should read before writing:) | 11:17 |
| @lisitsyn | antihistaminic stuff makes me sleeeep | 11:18 |
| -!- HeikoS [~heiko@nat-180-153.internal.eduroam.ucl.ac.uk] has joined #shogun | 11:20 | |
| -!- mode/#shogun [+o HeikoS] by ChanServ | 11:20 | |
| van51 | from what i read between you two, i am beginning to understand that I should have done something differently | 11:22 |
| van51 | because my PR is on the master branch | 11:23 |
| van51 | am I right? | 11:25 |
| vgorbati | lisitsyn: :) caught a cold? | 11:25 |
| vgorbati | van51: as far as I understand, all the PR's should be send to develop branch | 11:25 |
| @lisitsyn | vgorbati: allergy | 11:26 |
| sonne|work | van, vgorbati: exactly just do it against develop | 11:27 |
| vgorbati | lisitsyn: I guess the allergy is not to cats, because then your git-photo is strange:) | 11:28 |
| @lisitsyn | vgorbati: spring stuff | 11:29 |
| @lisitsyn | vgorbati: that's because I am a cat | 11:29 |
| vgorbati | lisitsyn: :) | 11:29 |
| vgorbati | lisitsyn: during spring you go crazy?) | 11:30 |
| vgorbati | lisitsyn: like a cat? | 11:30 |
| @lisitsyn | vgorbati: hah probably no | 11:31 |
| van51 | ok so I re-did a PR, can I delete the previous one? :S | 11:34 |
| @lisitsyn | van51: just clost it | 11:35 |
| @lisitsyn | close* | 11:35 |
| van51 | lisitsyn: ok, done :) | 11:35 |
| -!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has quit [Quit: Page closed] | 11:36 | |
| -!- lisitsyn [~lisitsin@mxs.kg.ru] has quit [Ping timeout: 240 seconds] | 12:00 | |
| -!- lisitsyn [~lisitsin@mxs.kg.ru] has joined #shogun | 12:05 | |
| -!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving] | 12:13 | |
| -!- van51 [~van51@athedsl-320854.home.otenet.gr] has quit [Ping timeout: 268 seconds] | 12:20 | |
| -!- HeikoS [~heiko@nat-180-153.internal.eduroam.ucl.ac.uk] has quit [Quit: Leaving.] | 12:43 | |
| @wiking | sonne|work: ping? | 13:13 |
| @wiking | sonney2k: pong :) | 13:14 |
| -!- hoijui_ [~hoijui@dslb-088-074-121-064.pools.arcor-ip.net] has joined #shogun | 13:22 | |
| sonne|work | wiking: pong | 13:24 |
| -!- sijin_ [~smuxi@144.214.222.109] has joined #shogun | 13:26 | |
| @wiking | sonne|work: sent u an email and to the other core developers | 13:29 |
| -!- Netsplit *.net <-> *.split quits: hoijui, sijin | 13:29 | |
| sonne|work | ok so I wait 30mins | 13:29 |
| @wiking | hehehe | 13:29 |
| @wiking | i dont think so | 13:29 |
| @wiking | it should be already there | 13:29 |
| @wiking | or? | 13:29 |
| @wiking | sent it to Soeren Sonnenburg <sonne@debian.org> | 13:30 |
| @wiking | sonne|work: it's all about the intergration tests... | 13:30 |
| @wiking | and that we should basically disable it temporarily | 13:30 |
| @wiking | or at least return always 0 | 13:30 |
| @wiking | so that it does not fail even if some of it fails :P | 13:30 |
| sonne|work | wiking: greylisting | 13:31 |
| @wiking | sonne|work: oh motherfucker | 13:31 |
| @wiking | :) | 13:31 |
| @wiking | sonne|work: evne from gmail? :) | 13:31 |
| @wiking | sonne|work: *eve | 13:31 |
| @wiking | *even | 13:31 |
| @wiking | sonne|work: u shoudl have it in /msg :P | 13:34 |
| @wiking | unless your irc client put me on flood :P | 13:34 |
| sonne|work | wiking: yeah but I already said - why disable it? | 13:34 |
| sonne|work | it is very valuable for linux | 13:34 |
| sonne|work | if you want to disable it then just for !linux systems | 13:34 |
| sonne|work | Admittedly I don't feel very comfortable of having our own rng in shogun | 13:35 |
| sonne|work | it is a cruzial thing - does eigen ship one btw? | 13:35 |
| @wiking | ok but they are destined to fail of !linux | 13:36 |
| @wiking | *on !linux | 13:36 |
| @wiking | either we same the models for other OS as well in data repo | 13:36 |
| @wiking | or what u suggest is that we don't run them on non-linux systems | 13:37 |
| @wiking | but then in my opinion is a badly designed test :( | 13:37 |
| @wiking | it's kind of like kill it's purpose :P | 13:37 |
| @wiking | if the test can only be run on a given system.... | 13:37 |
| -!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout] | 13:41 | |
| sonne|work | wiking: I didn't say we cannot fix it | 13:41 |
| sonne|work | but I am against disabling it | 13:41 |
| sonne|work | these integration tests are much more helpful then our few unit tests | 13:41 |
| sonne|work | and it takes long ages to get the whole suite back to work if we ignore them for 6 months (==this gsoc) | 13:42 |
| @wiking | ok but imo we should do something about it, becuase now one really has to check the travis builds in detail to see whether a PR or a commit does indeed break the shogun build, instead of really just checking a small icon which is either red or green...;P | 13:45 |
| sonne|work | wiking: wait why? | 13:47 |
| sonne|work | everything should work now right?! | 13:47 |
| @wiking | sonne|work: no now it fails | 13:47 |
| @wiking | sonne|work: everything | 13:47 |
| sonne|work | wiking: why? | 13:47 |
| @wiking | sonne|work: check out this for example | 13:48 |
| @wiking | sonne|work: https://github.com/shogun-toolbox/shogun/pull/1018 | 13:48 |
| @wiking | according to travis this PR is not good... it fails the build | 13:48 |
| @wiking | but it's not because of that given PR | 13:48 |
| sonne|work | serialization_complex_example.py setting 1/2 | 13:49 |
| @wiking | it's because the python_modular integration tests fails | 13:49 |
| sonne|work | only that fails | 13:49 |
| sonne|work | nothing more | 13:49 |
| @wiking | sonne|work: indeed | 13:49 |
| @wiking | sonne|work: but still you don't know this | 13:49 |
| sonne|work | so we just have to fix that | 13:49 |
| @wiking | unless you check yourself | 13:49 |
| @wiking | the travis builds | 13:49 |
| sonne|work | err? | 13:49 |
| @wiking | well on github you only know that the travis build failed | 13:49 |
| sonne|work | when something breaks you have to check the stuff and fix it | 13:49 |
| sonne|work | yes sure | 13:49 |
| @wiking | yes but now it's always broken | 13:49 |
| sonne|work | but how else would you do it? | 13:49 |
| @wiking | no matter what u do | 13:49 |
| sonne|work | wiking: well then just fix the example / the bug | 13:50 |
| sonne|work | "just" | 13:50 |
| @wiking | sonne|work: ok... let me put it another way | 13:50 |
| @wiking | i myself have no possible means to find out why that one little integration test fails on travis | 13:50 |
| @wiking | and this is like this for the last 1-2 weeks | 13:51 |
| sonne|work | wiking: yeah me neither | 13:51 |
| @wiking | hence nobody has any time/motivation etc | 13:51 |
| sonne|work | because it works here | 13:51 |
| @wiking | to fix this | 13:51 |
| sonne|work | so I just don't know how to fix it | 13:51 |
| @wiking | and i suggested a way to overcome this problem | 13:51 |
| @wiking | 'just' | 13:51 |
| sonne|work | how? | 13:51 |
| sonne|work | we could just blacklist this particular test | 13:51 |
| sonne|work | (like we do with lots of the others already) | 13:52 |
| @wiking | sonne|work: either remove the integration test, or just don't return 1 if one of the the integration tests fails | 13:52 |
| sonne|work | we have to fail | 13:52 |
| sonne|work | otherwise no one will check test | 13:52 |
| sonne|work | s | 13:52 |
| sonne|work | then better blacklist it | 13:52 |
| @wiking | ok then let's blacklist that one serialization test | 13:52 |
| sonne|work | wiking: just open up generator.py | 13:53 |
| sonne|work | and add the test to the blacklist | 13:53 |
| @wiking | sonne|work: btw it's really funny enough that *that* given test does not fail on osx... | 13:53 |
| sonne|work | it will kill both cases though | 13:53 |
| @wiking | sonne|work: but a lot of others does.... | 13:53 |
| sonne|work | I don't understand how it can fail | 13:53 |
| sonne|work | I mean it works for one and not for the other | 13:54 |
| sonne|work | maybe we should not blacklist but print the differences | 13:54 |
| @wiking | well that's an option as well | 13:54 |
| @wiking | sonne|work: http://pastebin.com/pquRSArj | 13:57 |
| @wiking | or you have a more sophisticated one? | 13:58 |
| sonne|work | wiking: maybe better add a failed_string | 14:01 |
| sonne|work | then add a tuple to failed | 14:01 |
| sonne|work | failed.append((setting_str, failed_str)) | 14:01 |
| sonne|work | so that we see it as summary at the end | 14:01 |
| sonne|work | otherwise it gets lots | 14:01 |
| sonne|work | but otherwise yes - that is whay I would do | 14:02 |
| @wiking | and failed_str is the 2 tuples? | 14:02 |
| sonne|work | and also mention if the length doesn't fit etc | 14:02 |
| sonne|work | wiking: there is a failed=[] array at the top | 14:02 |
| @wiking | yes | 14:02 |
| sonne|work | failed_str should be some string in the end | 14:02 |
| @wiking | and i see that setting_str is appeneded | 14:02 |
| sonne|work | it probably should also list if len(a) != len(b) etc | 14:03 |
| sonne|work | but hey it is a start | 14:03 |
| -!- pranjaldaga1 [~pranjalda@115.248.50.23] has joined #shogun | 14:07 | |
| @wiking | sonne|work: | 14:07 |
| @wiking | ok from this one why does it fail | 14:07 |
| @wiking | i dont see it | 14:07 |
| @wiking | ('statistics_mmd_kernel_selection_combined.py setting 1/2', [(CombinedKernel, [True, True, True, True, True], [True, True, True, True, True]), (CombinedKernel, [True, True, True, True, True], [True, True, True, True, True])]) | 14:07 |
| sonne|work | wiking: argh | 14:08 |
| -!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 14:08 | |
| -!- mode/#shogun [+o iglesiasg] by ChanServ | 14:08 | |
| @wiking | but it fails on my machine :S | 14:08 |
| sonne|work | so it is the guts of combined kernel | 14:09 |
| sonne|work | wiking: for SGObjects we have some function to print parameters IIRC | 14:09 |
| sonne|work | but I don't know whether we can print the actual current values | 14:10 |
| @wiking | ok anyhow | 14:10 |
| @wiking | i did this | 14:10 |
| @wiking | - failed.append(setting_str) | 14:10 |
| @wiking | + failed_string = [] | 14:10 |
| @wiking | want this in or not? | 14:10 |
| @wiking | + failed_string.append(a) | 14:10 |
| @wiking | + failed_string.append(b) | 14:10 |
| @wiking | just to see what travis says | 14:10 |
| @wiking | + failed.append((setting_str, failed_string)) print "%-60s ERROR" % setting_str | 14:10 |
| sonne|work | good enough | 14:12 |
| sonne|work | just push it | 14:12 |
| @wiking | ok | 14:12 |
| sonne|work | wiking: let me check if sgobject can print values | 14:12 |
| @wiking | ok | 14:12 |
| @wiking | pushed | 14:12 |
| @wiking | let's see what travis does with it | 14:13 |
| -!- shogun-notifier- [~irker@7nn.de] has joined #shogun | 14:13 | |
| shogun-notifier- | shogun: Viktor Gal :develop * f05af6d / tests/integration/python_modular/tester.py: https://github.com/shogun-toolbox/shogun/commit/f05af6d20abd96c2558c5f221358d529445252eb | 14:13 |
| shogun-notifier- | shogun: More debug information for python_modular integration tests | 14:13 |
| sonne|work | wiking: the only alternative idea I have is to serialize this sgobject into an asciifile | 14:13 |
| sonne|work | and then show the actual diff -u | 14:13 |
| @wiking | well | 14:15 |
| @wiking | anything is good just commit or tell me what to do | 14:15 |
| @wiking | i just want to fix this shit with travis | 14:16 |
| @wiking | as it's really a lot of help for all of us | 14:16 |
| @wiking | sonne|work: btw: have u figured out what the heck is with buildbot? | 14:16 |
| -!- sijin_ [~smuxi@144.214.222.109] has left #shogun [] | 14:17 | |
| -!- sijin_ [~smuxi@144.214.222.109] has joined #shogun | 14:18 | |
| -!- sijin_ [~smuxi@144.214.222.109] has quit [] | 14:18 | |
| -!- pranjaldaga1 [~pranjalda@115.248.50.23] has quit [Ping timeout: 260 seconds] | 14:19 | |
| -!- Netsplit over, joins: sijin | 14:20 | |
| sonne|work | wiking: no - I didn't get it to work | 14:20 |
| sonne|work | wiking: in principle version 0.8.7 supports multiple branches | 14:21 |
| sonne|work | but gitpoller always says - no changes | 14:21 |
| sonne|work | I simply don't understand why | 14:21 |
| sijin | Sorry to interupt, I just found Ckernel got segmentation fault when using it twice. | 14:23 |
| sonne|work | sijin: what are you exactly doing? | 14:25 |
| sijin | I just use it like this: Ckernel*p = new CGaussianKernel(); p->init(feature1,feature2);p->cleanup(); p->init(feature1, feature2); p->cleanup() | 14:26 |
| sijin | then i got segmentation fault | 14:27 |
| shogun-notifier- | shogun: iglesias :develop * 34056f3 / src/shogun/classifier/Perceptron.cpp,src/shogun/classifier/Perceptron.h: https://github.com/shogun-toolbox/shogun/commit/34056f39e1f06110cf86cb667354f1f5f169630a | 14:28 |
| shogun-notifier- | shogun: Add boolean member to Perceptron to allow model initialization. | 14:28 |
| shogun-notifier- | shogun: This is useful to show that the perceptron converges to different | 14:28 |
| shogun-notifier- | shogun: solutions depending on initial values of w and bias. | 14:28 |
| shogun-notifier- | shogun: iglesias :develop * 6fd3a5a / examples/undocumented/python_modular/graphical/classifier_perceptron_graphical.py: https://github.com/shogun-toolbox/shogun/commit/6fd3a5af40c8d9c3e095ba1ef21c2d7893ed9504 | 14:28 |
| shogun-notifier- | shogun: Perceptron graphical example. | 14:28 |
| shogun-notifier- | shogun: For the same data, the solution is different depending on initial values | 14:28 |
| shogun-notifier- | shogun: of w and bias | 14:28 |
| shogun-notifier- | shogun: Soeren Sonnenburg :develop * e10f78b / / (3 files): https://github.com/shogun-toolbox/shogun/commit/e10f78b0d4b97357d5831fe15aebcfdc41e4df10 | 14:28 |
| shogun-notifier- | shogun: Merge pull request #1007 from iglesias/feature/custom_init_perceptron | 14:28 |
| shogun-notifier- | shogun: | 14:28 |
| shogun-notifier- | shogun: Add small functionality to perceptron and graphical example | 14:28 |
| sijin | feature1, and feature1 are pointers of generated data. | 14:29 |
| sonne|work | sijin: please post the exact code (some small example) for us to reproduce | 14:30 |
| sonne|work | it can be anything currently (also a bug on your side) | 14:30 |
| -!- foulwall [~foulwall@li379-21.members.linode.com] has joined #shogun | 14:30 | |
| sonne|work | foulwall: hey there | 14:31 |
| foulwall | Hello everybody, sorry for the absence these days. | 14:31 |
| foulwall | sonne|work: I prepared for the mid term exams. | 14:32 |
| sonne|work | I guessed so | 14:32 |
| sonne|work | foulwall: just don't forget to submit you application in time | 14:33 |
| sonne|work | foulwall: btw I showed you demos to some professors | 14:33 |
| sijin | http://pastebin.com/c9aFkXbj here is the code | 14:33 |
| sonne|work | they liked it | 14:33 |
| sonne|work | foulwall: and they asked whether they can use this in teaching | 14:34 |
| sonne|work | foulwall: so if we do it really well lots of students will see your demos :D | 14:34 |
| -!- travis-ci [~travis-ci@ec2-174-129-147-163.compute-1.amazonaws.com] has joined #shogun | 14:35 | |
| travis-ci | [travis-ci] it's Viktor Gal'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/6628019 | 14:35 |
| -!- travis-ci [~travis-ci@ec2-174-129-147-163.compute-1.amazonaws.com] has left #shogun [] | 14:35 | |
| foulwall | sonne|work: I'll fix the GP regression demo tonight(add the region as you said) | 14:36 |
| foulwall | sonne|work: of course they can. | 14:36 |
| sonne|work | foulwall: yeah but it would be great wouldn't it? I mean people really learning / seeing what you've done | 14:37 |
| foulwall | sonne|work: I'll try to make the demo show the immediate step | 14:37 |
| sonne|work | foulwall: that also needs some changes inside shogun | 14:37 |
| foulwall | sonne|work: that'll be great | 14:37 |
| sonne|work | lets first do the demos | 14:38 |
| sonne|work | then add more | 14:38 |
| foulwall | sonne|work: and for the 3 choice, I choice the easiest one. | 14:38 |
| foulwall | the first one. I'll make the demo more beautiful | 14:38 |
| foulwall | sonne|work: for the 3 choices, I choose the first one. | 14:40 |
| foulwall | sonne|work: I'll do some work later today. I'll go for a class. | 14:41 |
| foulwall | sonne|work: cu. | 14:41 |
| -!- foulwall [~foulwall@li379-21.members.linode.com] has left #shogun [] | 14:41 | |
| sijin | sonne|work: did you see the code? any problems? | 14:42 |
| sonne|work | not clear yet | 14:46 |
| -!- travis-ci [~travis-ci@ec2-174-129-147-163.compute-1.amazonaws.com] has joined #shogun | 14:51 | |
| travis-ci | [travis-ci] it's Soeren Sonnenburg's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: http://travis-ci.org/shogun-toolbox/shogun/builds/6628392 | 14:51 |
| -!- travis-ci [~travis-ci@ec2-174-129-147-163.compute-1.amazonaws.com] has left #shogun [] | 14:51 | |
| -!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has joined #shogun | 14:54 | |
| sijin | sonne|work: can you try that code on your computer? | 14:58 |
| sonne|work | sijin: no time yet | 15:02 |
| sijin | ok, thx all the same | 15:02 |
| -!- abinash [75ef5e6e@gateway/web/freenode/ip.117.239.94.110] has joined #shogun | 15:13 | |
| -!- gsomix [~Miranda@188.168.13.211] has joined #shogun | 15:18 | |
| gsomix | hi | 15:18 |
| gsomix | sonne|work: are you back? | 15:20 |
| sonne|work | gsomix: yes berlin | 15:21 |
| -!- van51 [~van51@athedsl-397303.home.otenet.gr] has joined #shogun | 15:21 | |
| gsomix | sonne|work: great! | 15:22 |
| -!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: Lost terminal] | 15:37 | |
| @wiking | sonne|work: ok so the error is atm meaningless: 'serialization_complex_example.py setting 1/2', [(GMNPSVM, GMNPSVM), (GMNPSVM, GMNPSVM)]) | 15:40 |
| @wiking | sonne|work: any ideas what we should do next? | 15:40 |
| sonne|work | wiking: yeah we need a method to print the parameters and *values* of shogun objects | 16:02 |
| sonne|work | I think we already have some print_parameters or sth | 16:02 |
| sonne|work | so it should be very doable to also print the actual value in there | 16:02 |
| sonne|work | wiking: on the other hand this is more or less the same as serializing a shogun object into an ASCII file | 16:03 |
| sonne|work | and maybe the latter is better | 16:03 |
| sonne|work | and then just do a diff on top of it | 16:03 |
| sonne|work | wiking: that woudl work like this:fstream = SerializableAsciiFile("foo.asc", "w") | 16:04 |
| sonne|work | obj.save_serializable(fstream) | 16:05 |
| sonne|work | and then just outputting diff -u | 16:05 |
| @wiking | lisitsyn: ping? | 16:16 |
| @wiking | sonne|work: would we consider a library like this for integration: http://vision.mas.ecp.fr/Personnel/teboul/software.php | 16:17 |
| @wiking | or it's too strict? | 16:17 |
| -!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has quit [Ping timeout: 245 seconds] | 16:17 | |
| @wiking | ok no worries i've found this one which is totally gpl 2: http://code.google.com/p/librf/source/browse/#svn%2Ftrunk%2Flibrf | 16:19 |
| -!- abinash [75ef5e6e@gateway/web/freenode/ip.117.239.94.110] has quit [Ping timeout: 245 seconds] | 16:21 | |
| lisitsyn | wiking: yes? | 16:23 |
| @wiking | lisitsyn: can we just put this into shogun/lib: //code.google.com/p/librf/source/browse/#svn%2Ftrunk%2Flibrf | 16:23 |
| @wiking | http://code.google.com/p/librf/source/browse/#svn%2Ftrunk%2Flibrf | 16:23 |
| @wiking | and add a wrapper around it | 16:23 |
| @wiking | for shogun? | 16:23 |
| lisitsyn | looks like it is not free | 16:24 |
| @wiking | where? | 16:24 |
| @wiking | even weka is using this | 16:24 |
| lisitsyn | I mean it is just non-commercial | 16:24 |
| @wiking | http://code.google.com/p/librf/source/browse/trunk/COPYING | 16:24 |
| lisitsyn | ah alright | 16:25 |
| lisitsyn | if it s gpl then ye | 16:25 |
| @wiking | lisitsyn: ah u r checking the other | 16:25 |
| @wiking | lisitsyn: ok question lib/external/librf | 16:25 |
| @wiking | or lib/librf? | 16:25 |
| lisitsyn | wiking: external probably | 16:25 |
| @wiking | ok let's add a new branch finally to git | 16:29 |
| @wiking | feature/RandomForest | 16:29 |
| @wiking | :) | 16:29 |
| @wiking | huh this import will be rather funny | 16:35 |
| @wiking | ./lib/external/librf/tree.cpp:15:25: error: librf/utils.h: No such file or directory | 16:35 |
| -!- hoijui_ [~hoijui@dslb-088-074-121-064.pools.arcor-ip.net] has quit [Quit: Leaving] | 16:38 | |
| van51 | can someone help me on how to make two commits into one? :) | 16:52 |
| sonne|work | van51: git reset <oldgitid> | 17:04 |
| sonne|work | git commit --amend | 17:04 |
| -!- makokal [~bokal@eduroam-pool3-128.wlan.uni-bremen.de] has quit [Quit: makokal] | 17:05 | |
| -!- gsomix [~Miranda@188.168.13.211] has quit [Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org] | 17:09 | |
| -!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has joined #shogun | 17:11 | |
| van51 | sonne|work: thanks, but what if I don't have the code locally at the moment? | 17:11 |
| @wiking | damn | 17:20 |
| @wiking | how do i publish a git flow feature on shogun-toolbox/shogun.git? :S | 17:20 |
| -!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout] | 17:28 | |
| -!- gsomix [~Miranda@188.168.13.211] has joined #shogun | 17:42 | |
| -!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has quit [Ping timeout: 245 seconds] | 17:46 | |
| -!- lisitsyn [~lisitsin@mxs.kg.ru] has quit [Quit: Leaving.] | 18:01 | |
| @wiking | ok | 18:07 |
| @wiking | i have something, but not the right place | 18:07 |
| @wiking | https://github.com/vigsterkr/shogun/tree/feature/RandomForest | 18:07 |
| -!- sijin [~smuxi@144.214.222.109] has quit [] | 18:09 | |
| -!- theo [802818b4@gateway/web/freenode/ip.128.40.24.180] has joined #shogun | 18:10 | |
| -!- theo [802818b4@gateway/web/freenode/ip.128.40.24.180] has quit [Client Quit] | 18:14 | |
| @wiking | sonne|work: ping | 18:22 |
| -!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 18:26 | |
| -!- mode/#shogun [+o iglesiasg] by ChanServ | 18:26 | |
| -!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has joined #shogun | 18:31 | |
| @wiking | lol | 18:38 |
| @wiking | wtf is this | 18:38 |
| @wiking | modshogun_wrap.cxx:5476:3: error: "SWIG wrapped code invalid in 32 bit architecture, regenarete code using -DSWIGWORDSIZE32" | 18:38 |
| -!- pranjaldaga [~pranjalda@115.248.50.23] has joined #shogun | 18:40 | |
| @iglesiasg | wiking: the reason why travis fails in python_modular? | 18:40 |
| -!- HeikoS [~heiko@nat-160-69.internal.eduroam.ucl.ac.uk] has joined #shogun | 18:42 | |
| -!- mode/#shogun [+o HeikoS] by ChanServ | 18:42 | |
| -!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has quit [Ping timeout: 245 seconds] | 18:47 | |
| @wiking | iglesiasg: no it's the reason why python modular fails on freebsd | 18:51 |
| @wiking | ;P | 18:51 |
| -!- lisitsyn [~blackburn@188.168.13.211] has joined #shogun | 18:54 | |
| -!- mode/#shogun [+o lisitsyn] by ChanServ | 18:54 | |
| -!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: Reconnecting] | 18:58 | |
| -!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 18:58 | |
| -!- mode/#shogun [+o iglesiasg] by ChanServ | 18:59 | |
| gsomix | sonne|work: sonney2k: lisitsyn: iglesiasg: HeikoS: wiking: good evening. can you take a look at it? http://pastebin.com/XMW7NyrP :) | 18:59 |
| gsomix | I'm opened for any suggestions | 18:59 |
| @iglesiasg | it doesn't seem to load here :S | 19:01 |
| @HeikoS | same here :) | 19:02 |
| gsomix | oh gosh :( | 19:02 |
| gsomix | https://dl.dropboxusercontent.com/u/19029407/ideas%20for%20gsoc%202013.txt | 19:03 |
| @HeikoS | gsomix: like all the ideas! | 19:04 |
| @iglesiasg | gsomix: yeah, nice ones | 19:04 |
| @HeikoS | especially R and matlab bindings :) but they might be challenging | 19:05 |
| @iglesiasg | gsomix: my favourite are the 4th and the 6th | 19:05 |
| gsomix | iglesiasg: thanks | 19:07 |
| gsomix | HeikoS: yep, most challenging task is matlab interface, because swig doesn't have officially support (only for octave) | 19:08 |
| @iglesiasg | gsomix: and one more suggesting that I would love seeing is cmake support | 19:08 |
| @iglesiasg | suggesting -> suggestion | 19:08 |
| @HeikoS | gsomix: did you have a look at that project by the guy who wanted to improve swig for matlab? | 19:08 |
| @HeikoS | gsomix: but this might be quite hard in fact | 19:08 |
| @iglesiasg | HeikoS: gsomix: IMHO matlab interface doesn't pay off the effort | 19:09 |
| @iglesiasg | HeikoS: I know you are a matlab supporter though :) | 19:09 |
| @HeikoS | iglesiasg: I think I recently changed the sides :) | 19:11 |
| @HeikoS | but still, we need to have it | 19:11 |
| @HeikoS | so many people use it | 19:11 |
| gsomix | HeikoS: a little, just looked at code. not have time for test right now. I plan to test it after proposal submiting. | 19:11 |
| @HeikoS | ok | 19:12 |
| gsomix | HeikoS: so, we can lay aside this idea for near future. | 19:17 |
| @HeikoS | I would be very keen in having automatic interfaces for matlab | 19:17 |
| @HeikoS | almost everyone I know uses it | 19:17 |
| @HeikoS | I tell them to use python, but it doesnt help :) | 19:18 |
| gsomix | iglesiasg: cmake support. added to list | 19:18 |
| @HeikoS | same for R btw | 19:18 |
| @HeikoS | both bio and stats people mostly use R | 19:18 |
| @iglesiasg | gsomix: it would be pretty awesome, but it can be tought. I am not sure | 19:18 |
| @HeikoS | (and matlab) | 19:18 |
| gsomix | HeikoS: yep, I use matlab for my research, and my supervisor, and all in my lab. it is very popular. | 19:19 |
| @HeikoS | exactly | 19:20 |
| * gsomix afk | 19:21 | |
| -!- Vikalp [~AndChat49@223.185.83.159] has joined #shogun | 19:49 | |
| Vikalp | Hi everybody | 20:06 |
| Vikalp | M getting linking error while using shogun sgmatrix | 20:07 |
| Vikalp | Can anybody please help me out? | 20:07 |
| @lisitsyn | Vikalp: hey | 20:15 |
| @lisitsyn | are you using -lshogun key? | 20:15 |
| Vikalp | Yes | 20:16 |
| @lisitsyn | what is the error? | 20:16 |
| @wiking | ok | 20:16 |
| @wiking | why is this: | 20:16 |
| @wiking | -LINKFLAGS_PYTHON="-ldl -lutil" | 20:16 |
| @wiking | -POSTLINKFLAGS_PYTHON="-ldl -lutil" | 20:16 |
| @wiking | in ./configure | 20:16 |
| @lisitsyn | wiking: they should not be here? | 20:17 |
| @wiking | lisitsyn: well it's fine as long as it's only intended for linux ;) | 20:18 |
| @lisitsyn | wiking: ahh | 20:18 |
| @wiking | there's no libdl on *bsd | 20:18 |
| @lisitsyn | wiking: I should buy a mac | 20:18 |
| @lisitsyn | :D | 20:18 |
| @wiking | hehehe | 20:18 |
| @wiking | lisitsyn: any idea how i can move the feature/RandomForest branch to shogun-toolbox/shogun repo? | 20:19 |
| @lisitsyn | wiking: I actually had a chance but bought that shitty samsung | 20:19 |
| @wiking | https://github.com/vigsterkr/shogun/tree/feature/RandomForest | 20:19 |
| @wiking | the gitflow feature publish put it to my own repo | 20:19 |
| @wiking | :S | 20:19 |
| @lisitsyn | wiking: but why do you want to have it in main repo | 20:19 |
| @wiking | lisitsyn: so that you can contribute as well? | 20:20 |
| @lisitsyn | wiking: yeah makes sense | 20:20 |
| @lisitsyn | wiking: okay gitflow just names branches | 20:20 |
| @wiking | i mean in any feature if you publish it | 20:20 |
| @lisitsyn | so git pull it | 20:20 |
| @wiking | it would be great to go to the upstream repo | 20:20 |
| @wiking | since u want to publish it | 20:21 |
| @lisitsyn | wiking: just pull and push it | 20:21 |
| @wiking | in order that others can contribute as well | 20:21 |
| @wiking | you mean git push upstream feature/RandomForest | 20:22 |
| @wiking | ? | 20:22 |
| @wiking | there :) | 20:22 |
| -!- shogun-notifier- [~irker@7nn.de] has joined #shogun | 20:22 | |
| shogun-notifier- | shogun: Viktor Gal :feature/RandomForest * 8b17964 / src/shogun/lib/external/librf/ (17 files): https://github.com/shogun-toolbox/shogun/commit/8b17964144f2755f0e4df8209548168dbfd028ae | 20:22 |
| shogun-notifier- | shogun: Initial import of librf | 20:22 |
| shogun-notifier- | shogun: Random Forest implementation | 20:22 |
| shogun-notifier- | shogun: Viktor Gal :feature/RandomForest * e9928f8 / src/shogun/lib/external/librf/ (12 files): https://github.com/shogun-toolbox/shogun/commit/e9928f8f5c644d9bcd1ab849e51fddf84b53b7a4 | 20:22 |
| shogun-notifier- | shogun: Fix header include for librf | 20:22 |
| shogun-notifier- | shogun: Viktor Gal :feature/RandomForest * 68b8c99 / src/shogun/lib/external/librf/test.cpp,src/shogun/lib/external/librf/weights.cpp: https://github.com/shogun-toolbox/shogun/commit/68b8c9966c700ef6f6ee4d78a6ddd7729d01411d | 20:22 |
| shogun-notifier- | shogun: librf: remove unnecessary files | 20:22 |
| @lisitsyn | wiking: yeah if you can | 20:22 |
| @lisitsyn | bahm | 20:23 |
| @wiking | https://github.com/shogun-toolbox/shogun/tree/feature/RandomForest | 20:23 |
| @wiking | yey | 20:23 |
| @wiking | ok this one compiles fine | 20:23 |
| @wiking | we need wrappers and unit tests | 20:23 |
| @wiking | the original code does have unit tests | 20:23 |
| @wiking | so i'll integrate them | 20:23 |
| @wiking | hah | 20:23 |
| @wiking | instance_set_unittest.cc | 20:23 |
| @wiking | ;) | 20:23 |
| @wiking | yey | 20:23 |
| @wiking | only thing that it's not written using gtest :P | 20:24 |
| @wiking | but of course i don't know if it's a good idea to add unit tests that are testing none-shogun code... but i guess it wouldn't hurt | 20:24 |
| @lisitsyn | wiking: we should have some strict policy | 20:25 |
| @lisitsyn | how to handle external code | 20:25 |
| @wiking | true | 20:29 |
| @wiking | anyways this needs now a shogun wrapper | 20:29 |
| @wiking | btw: is this a linearmachine? :D | 20:30 |
| @lisitsyn | wiking: no! | 20:30 |
| -!- Vikalp [~AndChat49@223.185.83.159] has quit [Quit: Bye] | 20:31 | |
| @wiking | then? | 20:31 |
| -!- saket [0e8b5206@gateway/web/freenode/ip.14.139.82.6] has joined #shogun | 20:31 | |
| @lisitsyn | wiking: just a machine | 20:31 |
| @wiking | aaah | 20:31 |
| @lisitsyn | wiking: what is linear in random forest? ;) | 20:31 |
| @wiking | nada | 20:32 |
| @wiking | i'm just asking what should be the head | 20:32 |
| @wiking | btw <shogun/classifier/RandomForest.h> ? | 20:32 |
| @lisitsyn | wiking: good question | 20:32 |
| @lisitsyn | wiking: it handles multiclass right? | 20:33 |
| @wiking | lisitsyn: afaik it should | 20:33 |
| @lisitsyn | then makes sense to put it to multiclass | 20:33 |
| @wiking | ah ok | 20:33 |
| @wiking | knn is there as well | 20:33 |
| -!- gsomix [~Miranda@188.168.13.211] has quit [Read error: Connection reset by peer] | 20:46 | |
| -!- travis-ci [~travis-ci@ec2-23-22-96-232.compute-1.amazonaws.com] has joined #shogun | 20:54 | |
| travis-ci | [travis-ci] it's Viktor Gal'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/6638517 | 20:54 |
| -!- travis-ci [~travis-ci@ec2-23-22-96-232.compute-1.amazonaws.com] has left #shogun [] | 20:54 | |
| @wiking | woah cool | 20:55 |
| @wiking | actually travis is useful \o/ | 20:55 |
| -!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has joined #shogun | 21:01 | |
| -!- gsomix [~Miranda@188.168.13.211] has joined #shogun | 21:04 | |
| -!- pranjaldaga [~pranjalda@115.248.50.23] has quit [Ping timeout: 245 seconds] | 21:08 | |
| -!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has joined #shogun | 21:12 | |
| shogun-notifier- | shogun: vladislav.horbatiuk@gmail.com :develop * 5e03c62 / .travis.yml,src/Makefile,src/shogun/labels/DenseLabels.cpp: https://github.com/shogun-toolbox/shogun/commit/5e03c622e25c37670fbe9c31ffcd296a7f8a20e9 | 21:14 |
| shogun-notifier- | shogun: Merge branch 'develop' of https://github.com/shogun-toolbox/shogun into vgorbati-develop | 21:14 |
| shogun-notifier- | shogun: vladislav.horbatiuk@gmail.com :develop * f2874ed / : https://github.com/shogun-toolbox/shogun/commit/f2874edf599c83941ab6d2e5238ed5e9488c2fe6 | 21:14 |
| shogun-notifier- | shogun: Merge branch 'develop' of https://github.com/shogun-toolbox/shogun into vgorbati-develop | 21:14 |
| shogun-notifier- | shogun: vladislav.horbatiuk@gmail.com :develop * 2d84d67 / / (60 files): https://github.com/shogun-toolbox/shogun/commit/2d84d6742f7b6bbdde39ef540e527791b010808d | 21:14 |
| shogun-notifier- | shogun: Merge branch 'develop' of https://github.com/shogun-toolbox/shogun into vgorbati-develop | 21:14 |
| shogun-notifier- | shogun: vladislav.horbatiuk@gmail.com :develop * b15aec8 / src/NEWS,src/shogun/lib/tapkee/tapkee_methods.hpp: https://github.com/shogun-toolbox/shogun/commit/b15aec84f2af486af734eb974a45c26cad691bda | 21:14 |
| shogun-notifier- | shogun: Merge branch 'develop' of https://github.com/shogun-toolbox/shogun into vgorbati-develop | 21:14 |
| shogun-notifier- | shogun: vladislav.horbatiuk@gmail.com :develop * 9d40286 / / (5 files): https://github.com/shogun-toolbox/shogun/commit/9d40286898134258b902fc11fa1991a9d5cb3ec8 | 21:14 |
| shogun-notifier- | shogun: Added a wrapper (CTDistributedStochasticNeighborEmbedding class) for t-SNE dimensionality reduction method, implemented in tapkee (according to https://github.com/shogun-toolbox/shogun/issues/946). | 21:14 |
| shogun-notifier- | shogun: Also added a basic unit test for it, that embeds the test data from 3D to 2D using implemented wrapper, and checks that resulting data is actually in 2D, and that number of features stay the same. | 21:14 |
| shogun-notifier- | shogun: Sergey Lisitsyn :develop * c88afbf / / (5 files): https://github.com/shogun-toolbox/shogun/commit/c88afbf666345a7908d3fb56ab67fcdbe6727ef4 | 21:14 |
| shogun-notifier- | shogun: Merge pull request #1015 from vladislav-horbatiuk/vgorbati-develop | 21:14 |
| shogun-notifier- | shogun: | 21:14 |
| shogun-notifier- | shogun: Added a wrapper for t-SNE method implemented in tapkee library and a basic unit test for it. | 21:14 |
| @lisitsyn | wiking: so we have t-sne in shogun now too | 21:14 |
| @HeikoS | hey guys, currently running a distributed program on 160 computers :D | 21:15 |
| @lisitsyn | HeikoS: superman! | 21:15 |
| -!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has quit [Ping timeout: 245 seconds] | 21:17 | |
| @HeikoS | lisitsyn: this is very cool stuff | 21:27 |
| @HeikoS | oh and btw | 21:28 |
| @HeikoS | a stats postdoc here will implement a simple mala sampler in shogun | 21:28 |
| @HeikoS | and with that hopefully create the foundation for some sampling action | 21:28 |
| @lisitsyn | HeikoS: what is it? | 21:28 |
| @HeikoS | Metropolis adjusted Langevin algorithm | 21:29 |
| @HeikoS | https://www.google.co.uk/url?sa=t&rct=j&q=&esrc=s&source=web&cd=4&cad=rja&ved=0CEgQFjAD&url=http%3A%2F%2Fwww.ece.duke.edu%2F~lcarin%2FGirolami2011.pdf&ei=_oN5UaPJJueG0AWps4CYBA&usg=AFQjCNGJbqNdUQI9I7lO9fmhlASnVkzQmw&sig2=iBwA1_GfIGupW4ICewOiGQ&bvm=bv.45645796,d.d2k | 21:29 |
| @lisitsyn | no idea! | 21:29 |
| @lisitsyn | :D | 21:29 |
| @HeikoS | a method to sample probability distributions that are complicated | 21:29 |
| @HeikoS | you can use that to solve integrals that are too hard otherwise | 21:30 |
| @HeikoS | think of it as brute force breaking things when you cannot do anything elegant ;) | 21:32 |
| @HeikoS | but in an elegant way :D | 21:32 |
| @lisitsyn | HeikoS: sounds like monte-carlo | 21:32 |
| @lisitsyn | HeikoS: I am still in 60s :D | 21:33 |
| @HeikoS | hah :) | 21:33 |
| -!- gsomix [~Miranda@188.168.13.211] has quit [Read error: Connection reset by peer] | 21:33 | |
| -!- makokal [~bokal@pD9EF09DC.dip0.t-ipconnect.de] has joined #shogun | 21:33 | |
| @HeikoS | but very nice that the guy here does it | 21:33 |
| @HeikoS | they use it for a project and I convinced them to do it in shogun | 21:33 |
| @lisitsyn | HeikoS: yeah sure | 21:34 |
| @HeikoS | this is nice since if we have many of those, we might attract a few more stats people | 21:34 |
| @lisitsyn | HeikoS: we should think how can we speed up development | 21:34 |
| @HeikoS | next year's gsoc will have a sampling project for sure! :) | 21:35 |
| @lisitsyn | HeikoS: for example we could add a script | 21:35 |
| @lisitsyn | that creates a class | 21:35 |
| @lisitsyn | puts it to modular interface | 21:35 |
| @lisitsyn | HeikoS: see what I mean? | 21:35 |
| @HeikoS | ? | 21:41 |
| @lisitsyn | HeikoS: I mean for your guy it is not obvious how to put a class to modular interface | 21:42 |
| @lisitsyn | HeikoS: and it would be nice to have a tool that does that | 21:42 |
| @lisitsyn | say I write 'new_class statistics DaCoolSampling' | 21:43 |
| @lisitsyn | and here we go | 21:43 |
| -!- saket [0e8b5206@gateway/web/freenode/ip.14.139.82.6] has quit [Quit: Page closed] | 21:45 | |
| -!- gsomix [~Miranda@188.168.13.211] has joined #shogun | 21:46 | |
| @HeikoS | lisitsyn: sorry gotta go.... bye :) | 21:46 |
| @lisitsyn | HeikoS: alrighto | 21:46 |
| -!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has joined #shogun | 21:56 | |
| @sonney2k | wiking, any progress on printing out stuff? | 22:24 |
| @wiking | sonney2k: not really | 22:24 |
| @wiking | sonney2k: i bascially gave up | 22:25 |
| @wiking | somebody should take over | 22:25 |
| @wiking | i'm doing now random forest rather | 22:25 |
| -!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has quit [Ping timeout: 245 seconds] | 22:30 | |
| gsomix | sonney2k: hey :) | 22:37 |
| -!- sids_aquarius [~sidi_@li400-124.members.linode.com] has quit [Ping timeout: 246 seconds] | 22:52 | |
| -!- sids_aquarius [~sidi_@li400-124.members.linode.com] has joined #shogun | 22:53 | |
| -!- travis-ci [~travis-ci@ec2-23-22-195-190.compute-1.amazonaws.com] has joined #shogun | 23:09 | |
| 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/6639944 | 23:09 |
| -!- travis-ci [~travis-ci@ec2-23-22-195-190.compute-1.amazonaws.com] has left #shogun [] | 23:09 | |
| @lisitsyn | my turn to suffer | 23:09 |
| -!- naywhaya1e [~ryan@spoon.lugatgt.org] has joined #shogun | 23:10 | |
| -!- Netsplit *.net <-> *.split quits: naywhayare | 23:15 | |
| --- Log closed Fri Apr 26 00:00:29 2013 | ||
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!
