| --- Log opened Sun Apr 21 00:00:28 2013 | ||
| -!- saket [0e8b5206@gateway/web/freenode/ip.14.139.82.6] has quit [Quit: Page closed] | 00:51 | |
| -!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 00:58 | |
| -!- mode/#shogun [+o iglesiasg] by ChanServ | 00:58 | |
| -!- utsavjain [~tiripS@14.139.82.6] has quit [Quit: Leaving.] | 01:14 | |
| -!- k_matsuda [~k_matsuda@h101-111-222-195.catv02.itscom.jp] has joined #shogun | 01:28 | |
| -!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving] | 02:42 | |
| -!- ry_ [~smuxi@tu073053.ip.tsinghua.edu.cn] has quit [Ping timeout: 240 seconds] | 02:48 | |
| -!- foulwall [~foulwall@li379-21.members.linode.com] has joined #shogun | 02:56 | |
| shogun-buildbot | build #368 of nightly_default is complete: Failure [failed test] Build details are at http://www.shogun-toolbox.org/buildbot/builders/nightly_default/builds/368 | 04:05 |
|---|---|---|
| -!- foulwall [~foulwall@li379-21.members.linode.com] has quit [Remote host closed the connection] | 04:15 | |
| -!- rgirish28 [~rgirish28@61.17.226.91] has joined #shogun | 04:16 | |
| rgirish28 | hey, anyone free for a chat on GSoC projects? | 04:21 |
| -!- rgirish28 [~rgirish28@61.17.226.91] has quit [Ping timeout: 258 seconds] | 06:07 | |
| -!- abhi_iit_10691 [7ab0d1d0@gateway/web/freenode/ip.122.176.209.208] has joined #shogun | 06:39 | |
| -!- abhi_iit_10691 [7ab0d1d0@gateway/web/freenode/ip.122.176.209.208] has quit [Client Quit] | 06:39 | |
| -!- sidbee [~sidbee@202.78.172.162] has joined #shogun | 06:48 | |
| -!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has joined #shogun | 07:38 | |
| -!- foulwall [~foulwall@li379-21.members.linode.com] has joined #shogun | 07:43 | |
| -!- saket [0e8b5206@gateway/web/freenode/ip.14.139.82.6] has joined #shogun | 08:02 | |
| -!- rgirish28 [~rgirish28@111.92.71.249] has joined #shogun | 08:27 | |
| -!- romzee [ca4eaca2@gateway/web/freenode/ip.202.78.172.162] has joined #shogun | 08:29 | |
| rgirish28 | hey, anyone free for a chat on GSoC projects? | 08:32 |
| -!- sidbee [~sidbee@202.78.172.162] has quit [] | 08:33 | |
| -!- rgirish28 [~rgirish28@111.92.71.249] has quit [Quit: rgirish28] | 08:37 | |
| -!- rgirish28 [~rgirish28@111.92.71.249] has joined #shogun | 08:45 | |
| -!- saket [0e8b5206@gateway/web/freenode/ip.14.139.82.6] has quit [Ping timeout: 245 seconds] | 09:01 | |
| -!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 09:15 | |
| -!- mode/#shogun [+o iglesiasg] by ChanServ | 09:15 | |
| -!- rgirish28 [~rgirish28@111.92.71.249] has quit [Quit: rgirish28] | 09:25 | |
| -!- romzee [ca4eaca2@gateway/web/freenode/ip.202.78.172.162] has quit [Quit: Page closed] | 09:30 | |
| foulwall | Hi iglesiasg sonney2k | 09:35 |
| foulwall | There is something weird, the GaussianLikehood class can be imported from my local machine, but the same thing cannot be imported on my server | 09:36 |
| foulwall | I've used the same command making the shogun, and almost the same version of numpy and scipy. | 09:37 |
| foulwall | Is there any special configure options for the GaussianLikehood class? | 09:38 |
| foulwall | sorry GaussianLikelihood | 09:38 |
| -!- vikram360 [~vikram360@117.192.169.4] has joined #shogun | 09:39 | |
| -!- rgirish28 [~rgirish28@121.246.65.119] has joined #shogun | 09:44 | |
| @iglesiasg | foulwall: mmm not that I know honestly | 10:14 |
| @iglesiasg | let me check | 10:14 |
| @iglesiasg | foulwall: is Eigen3 in your server? | 10:16 |
| @iglesiasg | I see that GaussianLikelihood is guarded by ifdef HAVE_EIGEN3 | 10:16 |
| -!- vikram360 [~vikram360@117.192.169.4] has quit [Ping timeout: 264 seconds] | 10:20 | |
| foulwall | iglesiasg: thanks for the clue | 10:37 |
| -!- foulwall [~foulwall@li379-21.members.linode.com] has quit [Ping timeout: 264 seconds] | 10:42 | |
| -!- vikram360 [~vikram360@117.192.163.118] has joined #shogun | 10:53 | |
| -!- votjakovr [~chatzilla@host-46-241-3-209.bbcustomer.zsttk.net] has joined #shogun | 11:34 | |
| -!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has joined #shogun | 11:35 | |
| -!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has quit [Ping timeout: 245 seconds] | 11:49 | |
| vgorbati | lisitsyn: hello, are you here? | 11:52 |
| @lisitsyn | vgorbati: yes | 11:52 |
| vgorbati | lisitsyn: as I said, I will finish the test on Monday, but I was thinking ahead of what to do next, and the issue https://github.com/shogun-toolbox/shogun/issues/946 (Develop an adapter for t-SNE) looked suitable. Is it open/is there any blockers for this issue? | 11:54 |
| @lisitsyn | vgorbati: the only blocker is that I should update the code | 11:54 |
| @lisitsyn | I'll do it today | 11:54 |
| vgorbati | lisitsyn: ok, then I will take this issue next, if you do not mind. I guess, this task is pretty similar to "Change kPCA to use the Tapkee library", so the steps would be pretty similar? | 11:56 |
| @lisitsyn | vgorbati: yes | 11:56 |
| vgorbati | lisitsyn: ok, thanks for you answers) | 11:57 |
| @lisitsyn | votjakovr: Heiko will answer your mail rather soon so keep calm ;) | 12:18 |
| votjakovr | lisitsyn: Thanks:) | 12:19 |
| -!- vikram360 [~vikram360@117.192.163.118] has quit [Ping timeout: 256 seconds] | 12:21 | |
| -!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has quit [Quit: Page closed] | 12:37 | |
| -!- votjakovr [~chatzilla@host-46-241-3-209.bbcustomer.zsttk.net] has quit [Quit: ChatZilla 0.9.90 [Firefox 20.0.1/20130409194949]] | 12:51 | |
| -!- votjakovr [~chatzilla@host-46-241-3-209.bbcustomer.zsttk.net] has joined #shogun | 12:59 | |
| -!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has joined #shogun | 13:06 | |
| -!- saket [0e8b5206@gateway/web/freenode/ip.14.139.82.6] has joined #shogun | 13:09 | |
| -!- vikram360 [~vikram360@117.221.120.7] has joined #shogun | 13:23 | |
| -!- rgirish28 [~rgirish28@121.246.65.119] has quit [Ping timeout: 264 seconds] | 13:38 | |
| -!- saket [0e8b5206@gateway/web/freenode/ip.14.139.82.6] has quit [Ping timeout: 245 seconds] | 13:51 | |
| deerishi | @lisitsyn : could you please run this code and get back to me about the output (kernelpca)http://pastebin.com/aU0VFVB2 | 14:00 |
| deerishi | @iglesiasg : could you check it out? | 14:03 |
| deerishi | ? | 14:04 |
| deerishi | ? | 14:04 |
| deerishi | anyone der? | 14:10 |
| deerishi | ? | 14:10 |
| deerishi | ? | 14:10 |
| -!- vikram360 [~vikram360@117.221.120.7] has quit [Ping timeout: 256 seconds] | 14:20 | |
| @iglesiasg | deerishi: hey | 14:20 |
| @iglesiasg | I can try it later | 14:20 |
| -!- saket [0e8b5206@gateway/web/freenode/ip.14.139.82.6] has joined #shogun | 14:22 | |
| deerishi | @iglesiasg: k. | 14:23 |
| -!- HeikoS [~heiko@90.196.14.164] has joined #shogun | 14:38 | |
| deerishi | hello HeikoS, did you check the code i had sent you about kpca? | 14:40 |
| HeikoS | hi deerishi | 14:40 |
| -!- rgirish28 [~rgirish28@61.17.21.87] has joined #shogun | 14:41 | |
| HeikoS | I have been away from the computer over the last hours, but currently looking into everything ... will answer soon! | 14:41 |
| deerishi | thank you HeikoS.i will start working on another patch | 14:44 |
| HeikoS | lisitsyn: could you also have a look at https://github.com/shogun-toolbox/shogun/pull/991/files | 14:47 |
| HeikoS | since it modifies all the multiclass stuff | 14:47 |
| -!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving] | 14:49 | |
| -!- vgorbati [c12249c6@gateway/web/freenode/ip.193.34.73.198] has joined #shogun | 15:03 | |
| HeikoS | deerishi: | 15:13 |
| HeikoS | you are applying kernelPCA twice | 15:13 |
| HeikoS | preprocessor.apply_to_feature_matrix(features) | 15:13 |
| HeikoS | X=preprocessor.get_transformation_matrix() | 15:13 |
| HeikoS | X2=preprocessor.apply_to_feature_matrix(features) | 15:13 |
| -!- vgorbati [c12249c6@gateway/web/freenode/ip.193.34.73.198] has quit [Ping timeout: 245 seconds] | 15:16 | |
| -!- HeikoS [~heiko@90.196.14.164] has quit [Quit: Leaving.] | 15:20 | |
| -!- sijin [~sijin@144.214.222.109] has joined #shogun | 15:20 | |
| @lisitsyn | HeikoS: will take a look tonight | 16:01 |
| -!- vikram360 [~vikram360@117.192.167.69] has joined #shogun | 16:01 | |
| -!- sijin [~sijin@144.214.222.109] has quit [Quit: irc2go] | 16:05 | |
| saket | hi | 16:21 |
| saket | I am new to git flow ...and want to confirm on the following | 16:21 |
| -!- sijin [~sijin@144.214.222.109] has joined #shogun | 16:21 | |
| -!- abinash [75ef5e6e@gateway/web/freenode/ip.117.239.94.110] has joined #shogun | 16:22 | |
| saket | git flow feature start <name> ; git add filename ; git commit ; git flow feature finish; git flow feature publish <name> ....This set of commands is equivalent to a pull request right ? | 16:24 |
| sijin | are you there? | 16:25 |
| saket | me ? | 16:25 |
| -!- vikram360 [~vikram360@117.192.167.69] has quit [Ping timeout: 272 seconds] | 16:25 | |
| abinash | Hello, I just wanted to add a newly created example in the makefile | 16:26 |
| sijin | sorry. I forgot to use @ | 16:26 |
| abinash | Please help me out in knowing which file to changed | 16:26 |
| saket | lisitsyn: "git flow feature start <name> ; git add filename ; git commit ; git flow feature finish; git flow feature publish <name>" ....This set of commands is equivalent to a pull request right ? | 16:26 |
| saket | sonney2k: "git flow feature start <name> ; git add filename ; git commit ; git flow feature finish; git flow feature publish <name> ....This set of commands is equivalent to a pull request right ? | 16:31 |
| @sonney2k | saket, I haven't done this yet from a student perspective it might be that you don't have to finish the feature but just ask for a PR | 16:33 |
| @sonney2k | for that branch | 16:33 |
| saket | I did a PR (https://github.com/shogun-toolbox/shogun/pull/994 ) but i @iglesias asked to do it via git flow | 16:35 |
| saket | sonney2k: I request you to please guide me here .. @iglesias said that you will be able to help better in the second last comment on the thread in the above PR | 16:36 |
| @sonney2k | saket, just don't do the PR against master but develp! | 16:40 |
| @sonney2k | develop | 16:40 |
| saket | ohh | 16:40 |
| saket | thanks | 16:41 |
| saket | will send one right now :) | 16:41 |
| -!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 16:42 | |
| -!- mode/#shogun [+o iglesiasg] by ChanServ | 16:42 | |
| -!- mdragan [8d55007a@gateway/web/freenode/ip.141.85.0.122] has joined #shogun | 16:43 | |
| -!- sijin_ [~smuxi@144.214.222.109] has joined #shogun | 16:48 | |
| -!- sijin [~sijin@144.214.222.109] has quit [Quit: irc2go] | 16:48 | |
| sijin_ | @lisitsyn , are you there? I have a question about Kpca warper #968, can I add some code in tapkee_shogun.cpp ? | 16:52 |
| -!- vikram360 [~vikram360@117.192.175.188] has joined #shogun | 17:00 | |
| @lisitsyn | sijin_: I am back | 17:20 |
| sijin_ | hello | 17:20 |
| sijin_ | I just wonder whether i should change tapkee_shougun.cpp file since I don't find the support for KPCA in TAPKEE_METHODS_FOR_SHOGUN | 17:22 |
| @lisitsyn | sijin_: yes but I'd have to update tapkee | 17:23 |
| @lisitsyn | sijin_: is that you asked this quetion in the issue? | 17:23 |
| sijin_ | yes, it's me | 17:23 |
| @lisitsyn | alright | 17:23 |
| @lisitsyn | sijin_: I have to update the library and then we can continue | 17:24 |
| @lisitsyn | I didn't foresee we need an out-of-sample projection | 17:24 |
| sijin_ | ok , i will check out, thx @lisitsyn | 17:25 |
| @lisitsyn | sijin_: if there is anything else you would like to contibute with you may do that meanwhile | 17:26 |
| abinash | anyone help me out in adding a newly created example in the makefile in examples directory | 17:26 |
| sijin_ | okay. | 17:27 |
| @lisitsyn | abinash: libshogun examples? | 17:28 |
| abinash | lisitsyn: yes libshogun examples | 17:28 |
| @lisitsyn | abinash: just put that example to the list in makefile | 17:28 |
| abinash | which make file, the make file in examples directory or the one present in libshogun directory | 17:29 |
| @lisitsyn | abinash: shogun/examples/undocumented/libshogun/Makefile | 17:29 |
| abinash | I have added the one in documented example | 17:30 |
| @lisitsyn | *un*documented, documented examples are generated | 17:30 |
| abinash | so shouldn't it be shogun/examples/documented/libshogun/Makefile | 17:30 |
| abinash | ok thank you | 17:32 |
| -!- abhi_iit_10691 [7ab0b46b@gateway/web/freenode/ip.122.176.180.107] has joined #shogun | 17:34 | |
| -!- dvalcarce [5b757556@gateway/web/freenode/ip.91.117.117.86] has joined #shogun | 17:34 | |
| -!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has joined #shogun | 17:39 | |
| dvalcarce | hi all | 17:40 |
| dvalcarce | sonney2k: i've just made a pull request | 17:41 |
| abhi_iit_10691 | Hi I am new to shogun and interested in few of the ideas mentioned on the gsoc ideas page. Can anyone of you direct me to the instructions for getting shogun setup on my machine ? | 17:48 |
| rgirish28 | hi, anyone free to discuss GSoC ideas? | 17:53 |
| dvalcarce | abhi_iit_10691: you can find an installation guide at http://shogun-toolbox.org/doc/en/current/installation.html | 17:54 |
| abhi_iit_10691 | dvalcarce : Thank You | 17:56 |
| -!- rgirish28 [~rgirish28@61.17.21.87] has quit [Ping timeout: 252 seconds] | 18:03 | |
| -!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has quit [Ping timeout: 245 seconds] | 18:06 | |
| -!- abinash [75ef5e6e@gateway/web/freenode/ip.117.239.94.110] has quit [Ping timeout: 245 seconds] | 18:07 | |
| -!- utsavjain [~tiripS@14.139.82.6] has joined #shogun | 18:33 | |
| -!- rgirish28 [~rgirish28@61.17.21.190] has joined #shogun | 18:34 | |
| -!- shogun-notifier- [~irker@7nn.de] has joined #shogun | 18:46 | |
| shogun-notifier- | shogun: Abinash Panda :develop * 583fbff / examples/documented/libshogun/probability_libsvm.cpp: https://github.com/shogun-toolbox/shogun/commit/583fbff60f9cbd0268077eabccfd317e892144d2 | 18:46 |
| shogun-notifier- | shogun: example probability for SVM | 18:46 |
| shogun-notifier- | shogun: Abinash Panda :develop * 7f003b2 / examples/documented/libshogun/ (2 files): https://github.com/shogun-toolbox/shogun/commit/7f003b2b99a97493df356b3c53ab4abc542a9c1c | 18:46 |
| shogun-notifier- | shogun: example probabilities for SVM | 18:46 |
| shogun-notifier- | shogun: Abinash Panda :develop * 1e7569e / examples/documented/libshogun/probability_libsvm.cpp: https://github.com/shogun-toolbox/shogun/commit/1e7569e1d6e4bf065b7c26c3a9b8238067115a46 | 18:46 |
| shogun-notifier- | shogun: file removed | 18:46 |
| shogun-notifier- | shogun: Abinash Panda :develop * 857cb8c / examples/undocumented/libshogun/classifier_libsvm_probabilities.cpp: https://github.com/shogun-toolbox/shogun/commit/857cb8c2b3dc76e437022ed2c07da70c9e58d6af | 18:46 |
| shogun-notifier- | shogun: examples probabilities for SVM | 18:46 |
| shogun-notifier- | shogun: Abinash Panda :develop * 15f26c3 / examples/undocumented/libshogun/ (2 files): https://github.com/shogun-toolbox/shogun/commit/15f26c353f10d0aba407edf1adfed53f9a92c917 | 18:46 |
| shogun-notifier- | shogun: example probabilities for SVM | 18:46 |
| shogun-notifier- | shogun: Abinash Panda :develop * ff251a3 / examples/documented/libshogun/ (2 files): https://github.com/shogun-toolbox/shogun/commit/ff251a3e6a2109e35b6ea4356333081265089cf4 | 18:46 |
| shogun-notifier- | shogun: files removed | 18:46 |
| shogun-notifier- | shogun: Roman Votyakov :develop * f42672a / tests/unit/regression/gp/LaplacianInferenceMethod_unittest.cc: https://github.com/shogun-toolbox/shogun/commit/f42672a40c8392615069ea51eeb9fea13537c2ef | 18:47 |
| shogun-notifier- | shogun: Added unit-test for Laplacian inference method | 18:47 |
| shogun-notifier- | shogun: Heiko Strathmann :develop * 35f0277 / tests/unit/regression/gp/LaplacianInferenceMethod_unittest.cc: https://github.com/shogun-toolbox/shogun/commit/35f027793014674bb9cdbeac720ca7307ea84d41 | 18:47 |
| shogun-notifier- | shogun: Merge pull request #993 from votjakovr/develop | 18:47 |
| shogun-notifier- | shogun: | 18:47 |
| shogun-notifier- | shogun: Added unit-test for Laplacian inference method | 18:47 |
| shogun-notifier- | shogun: lambday :develop * d15a268 / / (3 files): https://github.com/shogun-toolbox/shogun/commit/d15a2686a8901d3e921c1e7ac9d9332ebd7077ed | 18:56 |
| shogun-notifier- | shogun: sample_from_gaussian for sparse cov matrix added | 18:56 |
| shogun-notifier- | shogun: lambday :develop * 8eeac14 / / (2 files): https://github.com/shogun-toolbox/shogun/commit/8eeac14229d7d4c1029647f85dc8f0b8d107d8a3 | 18:56 |
| shogun-notifier- | shogun: fixed sample_from_gaussian for sparse cov | 18:56 |
| shogun-notifier- | shogun: lambday :develop * 388dbdb / / (4 files): https://github.com/shogun-toolbox/shogun/commit/388dbdbcc0bc8c83f2dec30af53a52a0d0660253 | 18:56 |
| shogun-notifier- | shogun: added SimplicialLLT class in eigen3.h for older eigen | 18:56 |
| shogun-notifier- | shogun: lambday :develop * 306053c / src/shogun/mathematics/Statistics.cpp: https://github.com/shogun-toolbox/shogun/commit/306053c204e19bc5809cd8b7609708f53c1efd87 | 18:56 |
| shogun-notifier- | shogun: minor changes | 18:56 |
| shogun-notifier- | shogun: Heiko Strathmann :develop * f50778e / / (5 files): https://github.com/shogun-toolbox/shogun/commit/f50778e0402f6eb8f08b55d222ef2cb1f89a0d39 | 18:56 |
| shogun-notifier- | shogun: Merge pull request #989 from lambday/develop | 18:56 |
| shogun-notifier- | shogun: | 18:56 |
| shogun-notifier- | shogun: sample_from_gaussian for sparse cov matrix added | 18:56 |
| -!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has joined #shogun | 18:57 | |
| -!- abhi_iit_10691 [7ab0b46b@gateway/web/freenode/ip.122.176.180.107] has quit [Ping timeout: 245 seconds] | 18:59 | |
| -!- abhi_iit_10691 [7ab0c13e@gateway/web/freenode/ip.122.176.193.62] has joined #shogun | 19:02 | |
| saket | sonney2k: I sent a PR ( https://github.com/shogun-toolbox/shogun/pull/995 ) to the developer branch | 19:12 |
| saket | *develop branch | 19:12 |
| saket | as per "[20:06] <saket> sonney2k: I request you to please guide me here .. @iglesias said that you will be able to help better in the second last comment on the thread in the above PR [20:09] <@sonney2k> saket, just don't do the PR against master but develp! [20:09] <@sonney2k> develop" | 19:13 |
| -!- foulwall_ [~foulwall@li379-21.members.linode.com] has joined #shogun | 19:17 | |
| foulwall_ | Hi sonney2k? | 19:17 |
| -!- travis-ci [~travis-ci@ec2-54-234-189-28.compute-1.amazonaws.com] has joined #shogun | 19:18 | |
| travis-ci | [travis-ci] it's Heiko Strathmann'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/6516492 | 19:18 |
| -!- travis-ci [~travis-ci@ec2-54-234-189-28.compute-1.amazonaws.com] has left #shogun [] | 19:18 | |
| -!- gsomix [~gsomix@85.26.233.11] has joined #shogun | 19:28 | |
| gsomix | wow, how many people | 19:29 |
| gsomix | hi | 19:29 |
| -!- travis-ci [~travis-ci@ec2-54-234-189-28.compute-1.amazonaws.com] has joined #shogun | 19:34 | |
| travis-ci | [travis-ci] it's Heiko Strathmann'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/6516525 | 19:34 |
| -!- travis-ci [~travis-ci@ec2-54-234-189-28.compute-1.amazonaws.com] has left #shogun [] | 19:34 | |
| foulwall_ | hi gsomix | 19:38 |
| @sonney2k | foulwall_, haha very cool anim | 19:43 |
| @sonney2k | for the clustering! | 19:43 |
| foulwall_ | sonney2k: thanks | 19:43 |
| @sonney2k | foulwall_, btw one thing - I tried the clustering demo today from firefox | 19:44 |
| @sonney2k | it doesn't work | 19:44 |
| @sonney2k | I cannot click with the left mouse button | 19:44 |
| @sonney2k | nothing appears | 19:44 |
| foulwall_ | I'll do some browser check tomorrow. | 19:44 |
| @sonney2k | foulwall_, thanks | 19:45 |
| foulwall_ | sonney2k: have you tried the svr demo? | 19:45 |
| @sonney2k | foulwall_, ohh yes | 19:47 |
| @sonney2k | it works in firefox | 19:47 |
| @sonney2k | and it morpsh | 19:48 |
| @sonney2k | morpsh | 19:48 |
| @sonney2k | very cool | 19:48 |
| foulwall_ | sonney2k: thanks | 19:48 |
| @sonney2k | foulwall_, dvalcarce I guess you both know that you are both trying to get in for the same idea | 19:48 |
| -!- travis-ci [~travis-ci@ec2-54-234-189-28.compute-1.amazonaws.com] has joined #shogun | 19:49 | |
| travis-ci | [travis-ci] it's Heiko Strathmann'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/6516686 | 19:49 |
| -!- travis-ci [~travis-ci@ec2-54-234-189-28.compute-1.amazonaws.com] has left #shogun [] | 19:49 | |
| foulwall_ | I'm doing some debug job for the gp regression. | 19:49 |
| gsomix | foulwall_, sonney2k, good evening. where can I check demos for shogun now? | 19:49 |
| @sonney2k | I am currently thinking about creating 2 projects | 19:49 |
| @sonney2k | out of this | 19:50 |
| gsomix | I mean online demos. | 19:50 |
| foulwall_ | sonney2k: I knew that. His work's excellent too. | 19:50 |
| @sonney2k | gsomix, sec | 19:50 |
| dvalcarce | sonney2k, foulwall_: hi | 19:50 |
| @sonney2k | gsomix, http://shogun-toolbox.org/demos/svr/entrance | 19:51 |
| @sonney2k | gsomix, http://shogun-toolbox.org/demos/clustering/entrance | 19:51 |
| gsomix | sonney2k, thanks | 19:51 |
| @sonney2k | gsomix, that is foulwall_ 's work | 19:51 |
| @sonney2k | and then dvalcarce did http://shogun-toolbox.org/demos/classification/index | 19:51 |
| foulwall_ | Maybe I can do some mldata integration, whatever I have a whole summer's time | 19:51 |
| dvalcarce | foulwall and me thought about spliting the project if it is possible | 19:51 |
| foulwall_ | Hi dvalcarce | 19:52 |
| @sonney2k | dvalcarce, I have a couple of ideas - not sure who of you would prefer what | 19:52 |
| @sonney2k | I don't really know what your background is foulwall_ / dvalcarce | 19:52 |
| @sonney2k | so one is more or less the original plan | 19:53 |
| @sonney2k | do graphical demos for everythign that is in shogun | 19:53 |
| @sonney2k | lets mark it: | 19:53 |
| @sonney2k | 1) do graphical demos for everythign that is in shogun | 19:53 |
| @sonney2k | 2) do shogun applications | 19:53 |
| @sonney2k | that would be the ocr demo we have on the website (currently written in pyqt & shogun) | 19:54 |
| @sonney2k | and a couple of tools from bioinformatics | 19:54 |
| @sonney2k | and maybe some more we can think of | 19:54 |
| @sonney2k | 3) do a machine learning web framework | 19:54 |
| @sonney2k | that would mean - people could upload their data | 19:55 |
| @sonney2k | define the task underneath | 19:55 |
| @sonney2k | and shogun would come up with some predictions | 19:55 |
| @sonney2k | difficulty ranges from 1) easy 2) medium 3) tough | 19:55 |
| @sonney2k | foulwall_, dvalcarce any thoughts / preferences? | 19:56 |
| dvalcarce | all ideas seem interesting to me | 19:56 |
| foulwall_ | I prefer the 2) | 19:57 |
| dvalcarce | idea 1 would be (more or less) the task we are doing right now? | 19:57 |
| @sonney2k | lisitsyn, http://www.aibirds.org/ :) | 19:57 |
| rgirish28 | hi sonney2k, hope I can also join the discussion | 19:57 |
| @sonney2k | foulwall_, 1) yes | 19:57 |
| dvalcarce | shogun applications means shogun "desktop" applications? | 19:57 |
| @sonney2k | rgirish28, hey there | 19:58 |
| rgirish28 | The machine learning web framework seems like an awesome idea. | 19:58 |
| dvalcarce | btw, i like idea 3 | 19:58 |
| @sonney2k | dvalcarce, no via the web but of course the difference is rather small - there is some gui frontend done on the web site | 19:58 |
| foulwall_ | sonney2k: But I dont clear what's the framework running on. | 19:58 |
| @sonney2k | and some backend done in python | 19:59 |
| @sonney2k | that is using shogun for the computation heavy stuff | 19:59 |
| foulwall_ | sonney2k: PaaS? | 19:59 |
| @sonney2k | foulwall_, yeah kind of | 20:00 |
| @sonney2k | or SaaS - we ship the code with it so one could do that locally | 20:00 |
| foulwall_ | sonney2k: So that is kind of challenging | 20:00 |
| @sonney2k | yes it is | 20:00 |
| @sonney2k | one could start with very simple steps first though | 20:00 |
| @sonney2k | like a) just read in .csv files | 20:01 |
| @sonney2k | assuming everything is numeric | 20:01 |
| dvalcarce | it can be useful in laboratory environments | 20:01 |
| @sonney2k | then consider what the target value is | 20:01 |
| dvalcarce | a powerful computer can run the server side | 20:01 |
| @sonney2k | if it is a regression or classification task | 20:01 |
| dvalcarce | and scientists uploads their data via web | 20:01 |
| @sonney2k | and then run some algorithm on top of that end store the model | 20:01 |
| @sonney2k | dvalcarce, yeah or streaming via url | 20:02 |
| @sonney2k | then one could do the same with e.g. textural features and also allow more options | 20:03 |
| @sonney2k | like what you do now in the demos | 20:03 |
| @sonney2k | some values to tweak | 20:03 |
| @sonney2k | with some nice doc | 20:03 |
| foulwall_ | dvalcarce: what do you think | 20:03 |
| rgirish28 | right, so basically porting the whole toolbox to the web | 20:03 |
| rgirish28 | with demos, maybe a gui, the works. i suppose? | 20:05 |
| @sonney2k | rgirish28, not porting - but creating an interface to it. That would allow people who don't know much about data science to benefit from that | 20:05 |
| dvalcarce | foulwall_: i think i prefer idea #3, but i also like ideas 1 and 2 | 20:05 |
| dvalcarce | and you? | 20:05 |
| foulwall_ | sonney2k: just the same to you. | 20:05 |
| foulwall_ | sonney2k: sorry | 20:05 |
| foulwall_ | dvalcarce: just the same to you | 20:05 |
| rgirish28 | right. | 20:06 |
| dvalcarce | haha | 20:06 |
| rgirish28 | that would be quite ambitious. but extremely interesting. | 20:06 |
| @sonney2k | yes it is ambitious - but there is time until end of september | 20:07 |
| @lisitsyn | sonney2k: ah so you are getting to my idea! | 20:07 |
| @lisitsyn | :D | 20:07 |
| foulwall_ | sonney2k: framework means what? easy interface | 20:07 |
| rgirish28 | so would that be in python for backend | 20:07 |
| @sonney2k | and I think some basic .csv upload / training some model would barely take a few days | 20:07 |
| rgirish28 | ? | 20:08 |
| @sonney2k | of course there are all kinds of issues like potentially user management etc but having written mldata.org / mloss.org I know it is not difficult w/ django | 20:08 |
| @sonney2k | rgirish28, yes | 20:08 |
| rgirish28 | right | 20:08 |
| rgirish28 | sonney2k, I'm also interested in the fast food implementation. very interesting optimisation. | 20:09 |
| dvalcarce | yeah, django is amazing for web development | 20:09 |
| @sonney2k | lisitsyn, no you wanted to do some distributed learning - I want a ML learning platform | 20:10 |
| foulwall_ | sonney2k: I'm interested in that. | 20:10 |
| @sonney2k | dvalcarce, I have some comments/bug reports about your demo | 20:10 |
| dvalcarce | tell me | 20:10 |
| @sonney2k | dvalcarce, File "/home/shogun/django/shogun/demos/classification.py", line 51, in run_multiclass | 20:10 |
| @sonney2k | width = json.loads(request.GET["width"]) | 20:10 |
| -!- pararth [744988b1@gateway/web/freenode/ip.116.73.136.177] has joined #shogun | 20:10 | |
| @sonney2k | MultiValueDictKeyError: 'Key \'width\' not found | 20:11 |
| @sonney2k | dvalcarce, an also something is wrong with your coordinate system | 20:11 |
| @lisitsyn | sonney2k: isn't that the same? | 20:11 |
| @sonney2k | IIRC it is between 0 and 1 | 20:11 |
| @sonney2k | but underneath it seems to be much much different | 20:11 |
| @sonney2k | dvalcarce, points you give to shogun should also have the 0...1 scale | 20:11 |
| @sonney2k | dvalcarce, then the value for sigma would be much more sane in the ~ 1 range | 20:12 |
| dvalcarce | sonney2k: ok, let me check | 20:12 |
| @sonney2k | lisitsyn, no it is not like prepare jobs - execute | 20:12 |
| @sonney2k | lisitsyn, that would be for experts again | 20:13 |
| @lisitsyn | sonney2k: how do you see that then | 20:13 |
| @sonney2k | lisitsyn, if you want it is just a GUI | 20:13 |
| gsomix | foulwall_, nice demos. but I have error in svr. | 20:13 |
| gsomix | with linear kernel | 20:13 |
| @lisitsyn | sonney2k: well machine learning as a service is not new I guess :) | 20:13 |
| foulwall_ | gsomix: what's that | 20:14 |
| gsomix | Chrome Version 26.0.1410.63, "Error generated!", POST http://shogun-toolbox.org/demos/svr/point 404 (NOT FOUND) | 20:14 |
| @lisitsyn | http://about.wise.io/mlaas.html | 20:14 |
| @lisitsyn | https://bigml.com/ | 20:14 |
| foulwall_ | gsomix: I'll check that | 20:14 |
| @sonney2k | lisitsyn, surely not - but open source? | 20:14 |
| @sonney2k | lisitsyn, all closed | 20:14 |
| @sonney2k | and you forgot google predict | 20:14 |
| @sonney2k | microsoft and all the other big guys doing sth similar | 20:15 |
| @lisitsyn | sonney2k: we will never reach that being open source :D | 20:15 |
| @lisitsyn | well not really open source issue | 20:15 |
| @lisitsyn | it is about funding and a way to get money from that | 20:15 |
| @sonney2k | lisitsyn, I don't know what you mean | 20:16 |
| @sonney2k | lisitsyn, last time I heard a talk from the bigml guy | 20:16 |
| @sonney2k | it was just decision trees | 20:16 |
| rgirish28 | sonney2k, who can I talk to regarding the fast food implementation | 20:16 |
| @lisitsyn | sonney2k: it is not possible to develop such a thing during our spare time | 20:16 |
| @lisitsyn | what's what I mean | 20:16 |
| @sonney2k | so surely we can do sth similar | 20:16 |
| @lisitsyn | that* | 20:16 |
| @sonney2k | lisitsyn, why not? | 20:16 |
| -!- votjakovr [~chatzilla@host-46-241-3-209.bbcustomer.zsttk.net] has quit [Quit: ChatZilla 0.9.90 [Firefox 20.0.1/20130409194949]] | 20:16 | |
| @sonney2k | it is easy | 20:17 |
| @lisitsyn | sonney2k: depends - UI is not easy | 20:17 |
| @sonney2k | lisitsyn, look mldata/mloss.org were not that much work | 20:17 |
| @sonney2k | maybe 1-2 months | 20:17 |
| @lisitsyn | sonney2k: but it was developed when you were 'free' right? | 20:17 |
| @sonney2k | yes? | 20:18 |
| @sonney2k | but if say dvalcarce would do it he would be too | 20:18 |
| @sonney2k | and he would get support from us | 20:18 |
| @lisitsyn | yes syre | 20:18 |
| @lisitsyn | sure | 20:18 |
| @sonney2k | and you have seen how much he did already *without* asking anything | 20:18 |
| @sonney2k | IMHO it just needs some plan on how to solve the most simple tasks first | 20:19 |
| @lisitsyn | I am sorry I am a little out of it | 20:19 |
| foulwall_ | sonney2k: How much work time would you like me to spend a day? | 20:19 |
| @lisitsyn | 20-22 hours per day of course ;) | 20:20 |
| foulwall_ | sonney2k: 10 hours a day in the summer ? | 20:20 |
| @sonney2k | foulwall_, normal working hours | 20:20 |
| @lisitsyn | bah | 20:20 |
| @lisitsyn | 10? | 20:20 |
| @sonney2k | foulwall_, lisitsyn is joking of course | 20:20 |
| @sonney2k | it is supposed to be fun too | 20:20 |
| foulwall_ | bah, 10 hours is just the normal time. | 20:21 |
| @lisitsyn | sonney2k: would you like to work 10 hours at tomtom? ;) | 20:21 |
| @lisitsyn | :D | 20:21 |
| foulwall_ | sonney2k: we Chinese people got crazy in summer. | 20:21 |
| dvalcarce | haha | 20:22 |
| @lisitsyn | I usually get a hour or more late to job and it is kinda 6 hours per day | 20:22 |
| @lisitsyn | and I do get tired | 20:22 |
| @sonney2k | foulwall_, 8 hours/day is normal in germany so I don't expect more but hey it all depends | 20:22 |
| @lisitsyn | foulwall_: well if it doesn't burn you out | 20:22 |
| @sonney2k | when stuff is /really/ fun (like when you do research) then it can happen that you forget the world around you | 20:22 |
| @sonney2k | rgirish28, me | 20:23 |
| @lisitsyn | sonney2k: have you heard anything about 'alfa' hotel in berlin? | 20:23 |
| @sonney2k | lisitsyn, no - maybe you better take some hostel no? | 20:24 |
| @lisitsyn | sonney2k: an agency found that and '?njoy hotel berlin city messe' | 20:24 |
| foulwall_ | sonney2k: I can guaranty that all summer. | 20:24 |
| @lisitsyn | costs are pretty reasonable | 20:24 |
| @sonney2k | lisitsyn, didn't know you got rich :D | 20:25 |
| @lisitsyn | sonney2k: it costs less than *one* ticket to berlin | 20:25 |
| @lisitsyn | for two of us | 20:25 |
| @sonney2k | dvalcarce, btw where are you from? | 20:26 |
| -!- deerishi [73f88294@gateway/web/freenode/ip.115.248.130.148] has quit [Ping timeout: 245 seconds] | 20:26 | |
| rgirish28 | fantastic. so, what exactly are you looking for in the implementation? i mean deliverables. | 20:26 |
| @sonney2k | foulwall_, I take it you are from china right :) | 20:26 |
| foulwall_ | yes, and it's late in the night. | 20:26 |
| dvalcarce | sonney2k, from spain | 20:26 |
| @lisitsyn | sonney2k: the cheapest airflight from here to berlin is ~450$ | 20:26 |
| @lisitsyn | :D | 20:27 |
| @iglesiasg | Spanish majority around here! ^_^ | 20:27 |
| @lisitsyn | sonney2k: if things are going well I am in berlin on the night of 9th, July | 20:27 |
| @sonney2k | foulwall_, yeah well so shogun gets around the clock development :) | 20:27 |
| @sonney2k | iglesiasg, you know each other? | 20:28 |
| dvalcarce | sonney2k, i don't think so | 20:28 |
| @iglesiasg | sonney2k: not before he appeared around here, I didn't have the pleasure | 20:28 |
| @iglesiasg | dvalcarce: hola hola BTW :) | 20:28 |
| @lisitsyn | sonney2k: do you mean all spains know each other? ;) | 20:28 |
| foulwall_ | sonney2k: haha | 20:28 |
| dvalcarce | igleasiasg, hola! :) | 20:28 |
| @iglesiasg | that's racist man :P | 20:29 |
| @iglesiasg | jk | 20:29 |
| dvalcarce | lol | 20:29 |
| @lisitsyn | iglesiasg: if you was a mentor you would get quite a lot candidates from spain I am sure | 20:30 |
| @sonney2k | iglesiasg, :P | 20:30 |
| @iglesiasg | lisitsyn: you think so? why? | 20:30 |
| @sonney2k | iglesiasg, dvalcarce I thought there are only 2 left of you | 20:30 |
| @lisitsyn | iglesiasg: just see how many russian speakers I get | 20:30 |
| @iglesiasg | interesting idd | 20:30 |
| gsomix | lisitsyn, vodka, medvedi, stalin | 20:31 |
| @lisitsyn | you see? | 20:31 |
| @lisitsyn | :D | 20:31 |
| @iglesiasg | dvalcarce: where do you come from in Spain? | 20:31 |
| dvalcarce | igleasiasg, im from coru?a, and you? | 20:31 |
| @iglesiasg | dvalcarce: Madrid | 20:32 |
| @lisitsyn | sonney2k: is wilmersdorf kind of far away from you? | 20:32 |
| @lisitsyn | if I remember your location it seems deadly far away | 20:32 |
| foulwall_ | sonney2k gsomix: I've pulled a request for the svr linear bug. | 20:33 |
| foulwall_ | and it's too late, I'll go for sleep | 20:33 |
| foulwall_ | cu all guys. | 20:33 |
| gsomix | foulwall_, nite :) | 20:34 |
| @lisitsyn | see you | 20:34 |
| @iglesiasg | foulwall_: bye bye | 20:34 |
| dvalcarce | foulwall_, good night :) | 20:34 |
| @lisitsyn | iglesiasg: where are you going to stay in berlin? | 20:34 |
| @iglesiasg | lisitsyn: I haven't checked yet | 20:34 |
| @iglesiasg | I should definitely do it soon | 20:34 |
| foulwall_ | dvalcarce: for the project we can discuss tomorrow. | 20:34 |
| @iglesiasg | lisitsyn: have you already decided? | 20:34 |
| @lisitsyn | iglesiasg: I am considering enjoyHotel Berlin City Messe | 20:34 |
| @lisitsyn | iglesiasg: but not decided, no | 20:35 |
| -!- foulwall_ [~foulwall@li379-21.members.linode.com] has left #shogun [] | 20:35 | |
| @lisitsyn | iglesiasg: and the most important - when you are in berlin? | 20:35 |
| dvalcarce | foulwall_: ok! | 20:35 |
| @iglesiasg | lisitsyn: so how many days are we going to have around there finally? | 20:35 |
| @lisitsyn | iglesiasg: I am planning to be in berlin on 9th and get back on 15th | 20:35 |
| @sonney2k | gsomix, does it work for you now? | 20:35 |
| @sonney2k | dvalcarce, coudl you make sense of the issue? | 20:36 |
| @lisitsyn | iglesiasg: well friday 12th is the date and may be two days at TU | 20:36 |
| @sonney2k | lisitsyn, yes | 20:36 |
| @iglesiasg | it is one day workshop and 2 internal days? | 20:36 |
| @iglesiasg | ok | 20:36 |
| @sonney2k | very far away | 20:36 |
| dvalcarce | sonney2k, i'm not able to reproduce your issue | 20:36 |
| @lisitsyn | sonney2k: how far from the workshop place? | 20:36 |
| dvalcarce | however, there isn't the last version at http://shogun-toolbox.org/demos/classification/multiclass | 20:36 |
| @iglesiasg | sonney2k: these two days after the workshop at TU are more or less confirmed? | 20:36 |
| gsomix | sonney2k, yep | 20:36 |
| @iglesiasg | sonney2k: I remember you said you were not going to be able to be around though :S | 20:37 |
| @iglesiasg | lisitsyn: BTW has anyone from KTH CVAP registered? I am curious | 20:37 |
| dvalcarce | sonney2k, http://shogun-toolbox.org/static/js/classification.js file is old | 20:38 |
| @lisitsyn | iglesiasg: don't you have an access? | 20:38 |
| @iglesiasg | lisitsyn: no AFAIK | 20:38 |
| @sonney2k | dvalcarce, I've forwarded you the backtrace | 20:38 |
| @lisitsyn | iglesiasg: no one from KTH | 20:38 |
| @iglesiasg | lisitsyn: how many have applied so far? | 20:38 |
| @lisitsyn | iglesiasg: 9 | 20:39 |
| @sonney2k | dvalcarce, could you also please check if the range of the points is sane | 20:39 |
| @sonney2k | lisitsyn, 10 :) | 20:39 |
| @lisitsyn | sonney2k: -1 | 20:39 |
| @sonney2k | but none of us | 20:39 |
| @sonney2k | I know that cheng will come | 20:39 |
| @lisitsyn | sonney2k: the first row | 20:39 |
| @lisitsyn | sonney2k: really? | 20:39 |
| @sonney2k | yes | 20:39 |
| @lisitsyn | would be nice to see him | 20:39 |
| @sonney2k | and gunnar obviously and well I hope n4nd0 and you and heiko too | 20:40 |
| @lisitsyn | sonney2k: so will I be able to get to hotel after our party? :D | 20:40 |
| @sonney2k | lisitsyn, too far away | 20:40 |
| @sonney2k | take sth more central | 20:40 |
| @lisitsyn | sonney2k: Ufnaustrasse 1 | 20:40 |
| @lisitsyn | better/ | 20:40 |
| @lisitsyn | ? | 20:40 |
| @iglesiasg | sure I am going for sure | 20:40 |
| @sonney2k | lisitsyn, http://www.ostel.eu/ | 20:41 |
| @sonney2k | if you want east german style :) | 20:41 |
| @sonney2k | very close | 20:41 |
| @iglesiasg | hostel yaaay | 20:41 |
| dvalcarce | sonney2k, i am working on range values right now | 20:41 |
| @lisitsyn | iglesiasg: I do not feel the difference between hostel and hotel | 20:42 |
| @lisitsyn | :D | 20:42 |
| @iglesiasg | price :) | 20:42 |
| @lisitsyn | and? | 20:42 |
| @lisitsyn | iglesiasg: it can't be just price right? :) | 20:43 |
| @iglesiasg | lisitsyn: well I think in a hostel it is normal that food is not included | 20:43 |
| @lisitsyn | ahh | 20:43 |
| @iglesiasg | there is normally a small kitchen or so | 20:43 |
| @iglesiasg | it looks like that place has nothing available those days unfortunately | 20:44 |
| @iglesiasg | there must me others too of course | 20:44 |
| @lisitsyn | sonney2k: if I still be with that alfa or enjoyHotel - what time it would take to get to the place? | 20:44 |
| -!- gsomix [~gsomix@85.26.233.11] has quit [Quit: Leaving] | 20:46 | |
| @iglesiasg | lisitsyn: if I am checking the stuff right, it seems to be about 3km far | 20:46 |
| @lisitsyn | sonney2k: are you living kinda 20 km far away from your job place? | 20:50 |
| abhi_iit_10691 | Hi I am interested in the project "Large Scale Learning & Dimension reduction using hashing". The mentor is Olivier Chapelle. How do I get in touch with him ? Also we are required to submit a patch as part of our application. Can that be anything or does it have to be related to the project I am interested in ? | 20:52 |
| @sonney2k | lisitsyn, I think my place is more close to the hotel than alfa :) | 20:53 |
| -!- sumit [73f91219@gateway/web/freenode/ip.115.249.18.25] has joined #shogun | 20:53 | |
| @lisitsyn | sonney2k: to the what? | 20:53 |
| @sonney2k | abhi_iit_10691, yeah well mail him and cc me :) | 20:53 |
| @sonney2k | abhi_iit_10691, and if you want to start working - we need the basics | 20:54 |
| @lisitsyn | well that's ok if it far away - is your public transport ok in berlin, sonney2k? | 20:54 |
| @sonney2k | abhi_iit_10691, e.g. write some function that can split up a text document based on some tokens | 20:54 |
| @sonney2k | abhi_iit_10691, so we can later on-the-fly compute the hash of the words or n-grams | 20:55 |
| @sonney2k | lisitsyn, c-base | 20:55 |
| @lisitsyn | sonney2k: I see | 20:55 |
| @sonney2k | lisitsyn, sometimes it is | 20:55 |
| abhi_iit_10691 | sonney2k : his email-ID?...and I am sorry I didn't get your second point....did you mean I need to get familiar with the source code and understand some examples ? | 20:56 |
| @sonney2k | lisitsyn, what is wrong with ostel? | 20:56 |
| @sonney2k | isn't that hotel sufficient? | 20:56 |
| @lisitsyn | sonney2k: nothing - didn't check carefully yet | 20:56 |
| -!- gsomix [~gsomix@85.26.233.11] has joined #shogun | 20:56 | |
| @sonney2k | rgirish28, what do you want to know? | 20:56 |
| @sonney2k | just ask | 20:56 |
| rgirish28 | hey sonney2k, just wanted to know what kind of deliverables are you expecting in the fast food project | 20:57 |
| vgorbati | lisitsyn: hi, I have just updated the pull request (https://github.com/shogun-toolbox/shogun/pull/990) (as we discussed) - could please take a look, when you have some time? | 20:57 |
| @lisitsyn | vgorbati: alright | 20:57 |
| @lisitsyn | sonney2k: hmm ostel costs pretty the same | 20:58 |
| @sonney2k | rgirish28, well there are a couple of things one can do to get faster training for kernel classifiers | 20:58 |
| @sonney2k | rgirish28, one is to use a linear svm (usually very fast!) | 20:58 |
| rgirish28 | yeh but linear sum only work for linear data | 20:59 |
| rgirish28 | svms | 20:59 |
| @sonney2k | and compute some approximation of features (aka kernel expansion) | 20:59 |
| @lisitsyn | sonney2k: ok when you will get me to hotel by car, okay? ;) hahah | 20:59 |
| @iglesiasg | lisitsyn: U must be subway in the map | 20:59 |
| @iglesiasg | tunnelbanet ~ tunnelbana | 21:00 |
| rgirish28 | right | 21:00 |
| @lisitsyn | is that ok to get drunk and ride through the city in your country sonney2k? | 21:00 |
| @lisitsyn | may be some gunfire | 21:00 |
| @lisitsyn | iglesiasg: ^ that's how some ethnicities do (remember our talk) :D | 21:01 |
| dvalcarce | sonney2k, i've send a pull request | 21:01 |
| @iglesiasg | lisitsyn: omg you like hot topics :) | 21:02 |
| dvalcarce | i think its fixed | 21:02 |
| @sonney2k | lisitsyn, yes sure in particular you should try it when leaving c-base | 21:02 |
| @sonney2k | (which is very close to the chinese - well guarded - embassy ) | 21:02 |
| @lisitsyn | sonney2k: when do we leave c-base? | 21:02 |
| abhi_iit_10691 | sonney2k : can I get your email-ID as well as Olivier's ? | 21:03 |
| gsomix | sonney2k, you're hot today. :) | 21:04 |
| @iglesiasg | O_O | 21:05 |
| @lisitsyn | haha | 21:05 |
| @lisitsyn | gsomix: that was lewdly | 21:05 |
| gsomix | :( | 21:06 |
| -!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has quit [Ping timeout: 245 seconds] | 21:06 | |
| gsomix | I mean "very active". | 21:06 |
| @lisitsyn | haha | 21:06 |
| gsomix | bad, bad multitran dictionary | 21:06 |
| @sonney2k | gsomix, but I am hot too | 21:06 |
| @iglesiasg | hahaha | 21:06 |
| @sonney2k | so no worries :P | 21:06 |
| @lisitsyn | gay jokes back | 21:06 |
| @lisitsyn | sonney2k: so are you in pants today? | 21:07 |
| @lisitsyn | are you hot enough to get rid of that ridiculous wearing? | 21:07 |
| dvalcarce | im going to dinner, see you later | 21:08 |
| -!- dvalcarce [5b757556@gateway/web/freenode/ip.91.117.117.86] has left #shogun [] | 21:08 | |
| rgirish28 | sonney2k: i understand the problem and i am familiar with kernel expansion and it's deficiencies. just wanted to understand what you thought of the problem and what are the implementation ideas that yu had. | 21:10 |
| @iglesiasg | you scared dvalcarce lisitsyn :( | 21:10 |
| gsomix | lisitsyn, bad boy. | 21:13 |
| @lisitsyn | haha | 21:14 |
| @lisitsyn | gsomix: you sound lewdly again | 21:14 |
| @sonney2k | lisitsyn, no pants as usual | 21:14 |
| gsomix | I know, I know... | 21:14 |
| @sonney2k | rgirish28, well just implement the various expansions | 21:14 |
| @sonney2k | from random fourier features, over a couple of random kitchen sinks to fast food | 21:15 |
| @sonney2k | that would be the paln | 21:15 |
| @sonney2k | plan | 21:15 |
| @lisitsyn | the pain | 21:15 |
| rgirish28 | fair enough. any demo ideas for speed comparison? datasets | 21:15 |
| @sonney2k | and everything as dot features for real-valued (dense) and if sanely possible sparse inputs | 21:15 |
| rgirish28 | ? | 21:15 |
| -!- gsomix [~gsomix@85.26.233.11] has quit [Quit: Leaving] | 21:16 | |
| -!- gsomix [~gsomix@85.26.233.11] has joined #shogun | 21:16 | |
| @sonney2k | rgirish28, you could just use mnist and train a gaussian kernel svm on that and then compare e.g. liblinear/sgd/ocas with these 'new' features | 21:16 |
| @iglesiasg | sonney2k: Berlin question: is it easy to get lost in the subway and train? I mean changing lines and stuff like that | 21:17 |
| @iglesiasg | or is it well indicated | 21:17 |
| rgirish28 | right, i think i'll work out a proposal and see how it goes. Thanks! | 21:19 |
| @lisitsyn | iglesiasg: come to samara it is impossible to get lost here | 21:19 |
| @lisitsyn | in subway I mean | 21:19 |
| @lisitsyn | iglesiasg: now the million dollar question - why? | 21:19 |
| @iglesiasg | lisitsyn: but your streets eat cars | 21:19 |
| @lisitsyn | 3 attempts | 21:19 |
| @lisitsyn | iglesiasg: nahh that's ok | 21:19 |
| @lisitsyn | and your first attempt is ..? ;) | 21:19 |
| @iglesiasg | none uses it? | 21:20 |
| -!- rgirish28 [~rgirish28@61.17.21.190] has left #shogun [] | 21:20 | |
| @lisitsyn | iglesiasg: yes but it doesn't mean you can't get lost | 21:20 |
| @lisitsyn | one more! | 21:20 |
| @iglesiasg | there are just two stations? | 21:20 |
| @lisitsyn | iglesiasg: you are near! | 21:20 |
| @iglesiasg | there is no subway at all | 21:21 |
| @lisitsyn | haha | 21:21 |
| @lisitsyn | no | 21:21 |
| @lisitsyn | iglesiasg: there are may be 8 station | 21:21 |
| @lisitsyn | s | 21:21 |
| @lisitsyn | but one line | 21:21 |
| @lisitsyn | :D | 21:21 |
| @lisitsyn | so you can't get lost on the one line | 21:21 |
| @iglesiasg | ha! let me try | 21:21 |
| @lisitsyn | iglesiasg: moscow subway is crazy though | 21:22 |
| @lisitsyn | iglesiasg: okay should I let you know if I settle with any hotel/hostel? | 21:24 |
| @iglesiasg | lisitsyn: inded | 21:24 |
| @iglesiasg | indeed | 21:24 |
| gsomix | moscow subway... *scared* | 21:24 |
| @iglesiasg | please | 21:24 |
| @lisitsyn | iglesiasg: will you consider the same place? | 21:24 |
| @iglesiasg | lisitsyn: yeah | 21:25 |
| @lisitsyn | iglesiasg: alrighto | 21:25 |
| @lisitsyn | iglesiasg: my gf didn't like hostel idea though :) | 21:25 |
| @iglesiasg | hehe I see | 21:26 |
| @lisitsyn | iglesiasg: well if it is just 20 minutes by subway | 21:26 |
| @iglesiasg | it sounds worse than it is in reality, but I understand | 21:26 |
| @lisitsyn | I am totally ok | 21:26 |
| @lisitsyn | iglesiasg: it takes the same time to get me to the job so it is just phew | 21:27 |
| -!- sumit [73f91219@gateway/web/freenode/ip.115.249.18.25] has quit [Ping timeout: 245 seconds] | 21:27 | |
| abhi_iit_10691 | sonney2k : Can you give me your and Olivier's mail-ID so that I can contact you via mail ? and regarding the splitting the text document thing you basically want me to find all words in the document and their frequencies right ?Also should I write this function in plain C++ ? | 21:27 |
| -!- mdragan [8d55007a@gateway/web/freenode/ip.141.85.0.122] has quit [Quit: Page closed] | 21:29 | |
| @iglesiasg | abhi_iit_10691: you can get the mails in the ideas page | 21:31 |
| abhi_iit_10691 | iglesiasg : the ideas page doesn't have any email-IDs all it shows for Olivier is this "http://ulivier.chapelle.cc/" . | 21:37 |
| @iglesiasg | abhi_iit_10691: really? let me check | 21:38 |
| @iglesiasg | abhi_iit_10691: I checked that page personally but I might be wrong | 21:38 |
| @iglesiasg | abhi_iit_10691: hehe just change uilivier to olivier :) | 21:39 |
| abhi_iit_10691 | iglesiasg : the hyperlink on the mentor's name goes to "http://ulivier.chapelle.cc/" | 21:40 |
| @lisitsyn | ulivier? | 21:40 |
| @lisitsyn | oops | 21:40 |
| @iglesiasg | abhi_iit_10691: yeah I see that | 21:40 |
| @lisitsyn | olivier | 21:40 |
| @iglesiasg | abhi_iit_10691: just go to http://olivier.chapelle.cc/ instead | 21:40 |
| @lisitsyn | heh | 21:40 |
| @lisitsyn | okay lets consider that a trap | 21:40 |
| @iglesiasg | haha | 21:41 |
| @iglesiasg | it is the first task for GSoC candidates | 21:41 |
| abhi_iit_10691 | iglesiasq : haha....thanx... | 21:41 |
| @iglesiasg | lisitsyn: let me fix the link | 21:41 |
| @lisitsyn | iglesiasg: well may be that trap is worth to stay hahah | 21:41 |
| @iglesiasg | lisitsyn, abhi_iit_10691 : it should be ok now | 21:42 |
| abhi_iit_10691 | iglesiasg sonney2k : thanx also it would be great if you could address my question related to the document splitting task sonney2k told me to do or should I just contact Olivier ? | 21:44 |
| @iglesiasg | abhi_iit_10691: remind me what sonney2k told you | 21:44 |
| abhi_iit_10691 | he told me to write a function to split a document based on tokens....my question is should I just write in plain C++ ? Basically I need some task which I can submit as a patch which is required as part of the application. | 21:47 |
| @iglesiasg | abhi_iit_10691: mmm I guess it should be in C++ but maybe there is already something done that makes that, I wouldn't be surprised | 21:50 |
| @iglesiasg | you can also wait, sonney2k will answer you sooner or later | 21:51 |
| @iglesiasg | abhi_iit_10691: or you can select another task from the entrance tasks in github | 21:51 |
| -!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout] | 21:56 | |
| abhi_iit_10691 | iglesiasg : thanx I'll look up the entrance tasks and wait for sonney2k's reply | 21:57 |
| -!- jaypanda [0e8b5206@gateway/web/freenode/ip.14.139.82.6] has joined #shogun | 22:10 | |
| -!- abhi_iit_10691 [7ab0c13e@gateway/web/freenode/ip.122.176.193.62] has quit [Quit: Page closed] | 22:14 | |
| -!- utsavjain [~tiripS@14.139.82.6] has left #shogun [] | 22:26 | |
| saket | iglesiasg: lisitsyn sonney2k There are two files with two different implementations of v_array namely lib/v_array.h and /lib/JLCoverTreePoint.h | 22:28 |
| @lisitsyn | saket: I'd even say it appears 3 times :D | 22:29 |
| saket | omg! | 22:29 |
| saket | I am getting redefinition error in make :'( | 22:29 |
| @lisitsyn | sounds like a bug | 22:30 |
| saket | a possible solution I am currently thinking is refactoring ...but that will too cumbersome to do for the complete project .. | 22:32 |
| saket | is there a better way ? | 22:32 |
| -!- gsomix [~gsomix@85.26.233.11] has quit [Quit: Leaving] | 22:33 | |
| @lisitsyn | saket: yes, put one of them to some namespace deeper | 22:34 |
| saket | okay ...but won't i needed to write 'using namespace deeper' wherever the other one is used ? | 22:35 |
| saket | *won't i be needed to | 22:35 |
| @lisitsyn | by deeper I meant some namespace that is 'deeper' than shogun | 22:37 |
| @lisitsyn | like v_array namespace inside shogun namespace | 22:37 |
| @lisitsyn | and yes you'd have to do that | 22:37 |
| saket | okay ... I will do it .. | 22:38 |
| wiking | bu | 22:54 |
| -!- pararth [744988b1@gateway/web/freenode/ip.116.73.136.177] has quit [Quit: Page closed] | 22:59 | |
| @lisitsyn | wiking: not scaried yet! | 23:00 |
| wiking | lisitsyn: hehehe | 23:00 |
| saket | adding a deeper namespace is adding to woes :( ../shogun/lib/JLCoverTreePoint.h:61:11: error: reference to 'v_array' is ambiguous | 23:00 |
| @lisitsyn | saket: de-ambiguate it! ;) | 23:00 |
| saket | by putting shogun::v_array ? | 23:01 |
| @lisitsyn | yeah | 23:01 |
| saket | okay ... | 23:02 |
| -!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has joined #shogun | 23:16 | |
| -!- vikram360 [~vikram360@117.192.175.188] has quit [Ping timeout: 248 seconds] | 23:22 | |
| saket | lisitsyn: that worked ..thanks :-) | 23:23 |
| saket | but just out of curiosity and general discussion, wasn't it incorrect to have 2/3 classes having the same name in the same namepsace? | 23:24 |
| @lisitsyn | saket: of course | 23:24 |
| -!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has quit [Ping timeout: 245 seconds] | 23:24 | |
| @lisitsyn | saket: it is in shogun/lib/tapkee/... because tapkee is kind of 'separate' library | 23:24 |
| @lisitsyn | saket: as for shogun/lib/v_array.h - yes it should be removed I guess | 23:25 |
| saket | ohh | 23:27 |
| saket | sorry to bug again ...but i need a little help ...make runs fine in /src/shogun ... but exits with error in /src/ | 23:39 |
| saket | 'make' | 23:39 |
| saket | " /lib/v_array.h:40:25: error: candidates are: template<class T> class shogun::shogun_v_array::v_array" is the error and v_array.h:40 is the definition of class | 23:40 |
| saket | and if put "template<class T> class shogun::shogun_v_array::v_array" as the definition then it gives "../shogun/lib/v_array.h:41:1: error: qualified name does not name a class before '{' token" ... | 23:43 |
| saket | any help? | 23:48 |
| -!- saket [0e8b5206@gateway/web/freenode/ip.14.139.82.6] has quit [Quit: Page closed] | 23:52 | |
| --- Log closed Mon Apr 22 00:00:28 2013 | ||
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!