--- Log opened Sat Nov 30 00:00:46 2013 | ||
-!- naywhayare [~ryan@spoon.lugatgt.org] has joined #shogun | 01:12 | |
shogun-buildbot_ | build #633 of nightly_default is complete: Failure [failed test notebooks] Build details are at http://buildbot.shogun-toolbox.org/builders/nightly_default/builds/633 | 05:02 |
---|---|---|
-!- travis-ci [~travis-ci@ec2-54-242-159-176.compute-1.amazonaws.com] has joined #shogun | 11:47 | |
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/14731258 | 11:47 |
-!- travis-ci [~travis-ci@ec2-54-242-159-176.compute-1.amazonaws.com] has left #shogun [] | 11:47 | |
-!- sonne|osx [~sonne@f053036245.adsl.alicedsl.de] has joined #shogun | 13:12 | |
-!- sonne|osx [~sonne@f053036245.adsl.alicedsl.de] has quit [Quit: sonne|osx] | 13:19 | |
-!- sonne|osx [~sonne@f053036245.adsl.alicedsl.de] has joined #shogun | 16:23 | |
-!- sonne|osx [~sonne@f053036245.adsl.alicedsl.de] has quit [Quit: sonne|osx] | 16:52 | |
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun | 18:20 | |
shogun-notifier- | shogun: Soeren Sonnenburg :develop * e47602f / src/interfaces/modular/SGBase.i: https://github.com/shogun-toolbox/shogun/commit/e47602f2e9ac3e65f7c5797cf9efe8977e1354e2 | 18:20 |
shogun-notifier- | shogun: change order of includes to fix compile error | 18:20 |
-!- Saurabh7 [~Saurabh7@115.248.130.148] has joined #shogun | 18:47 | |
-!- travis-ci [~travis-ci@ec2-54-237-116-165.compute-1.amazonaws.com] has joined #shogun | 18:59 | |
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/14740299 | 18:59 |
-!- travis-ci [~travis-ci@ec2-54-237-116-165.compute-1.amazonaws.com] has left #shogun [] | 18:59 | |
-!- iglesiasg [~iglesiasg@206.Red-83-61-168.dynamicIP.rima-tde.net] has joined #shogun | 19:09 | |
-!- mode/#shogun [+o iglesiasg] by ChanServ | 19:09 | |
shogun-buildbot_ | build #2061 of deb3 - modular_interfaces is complete: Failure [failed test python modular test csharp modular] Build details are at http://buildbot.shogun-toolbox.org/builders/deb3%20-%20modular_interfaces/builds/2061 blamelist: Soeren Sonnenburg <sonne@debian.org> | 19:19 |
lisitsyn | iglesiasg: o la la welcome back to espana | 19:25 |
-!- iglesiasg [~iglesiasg@206.Red-83-61-168.dynamicIP.rima-tde.net] has quit [Ping timeout: 245 seconds] | 19:37 | |
@sonney2k | lisitsyn, look things compile now :) | 19:56 |
@sonney2k | lisitsyn, but now all tests fail http://buildbot.shogun-toolbox.org/builders/deb3%20-%20modular_interfaces/builds/2061/steps/test%20python%20modular/logs/stdio | 19:57 |
lisitsyn | sonney2k: ah still this issue? | 19:59 |
lisitsyn | sonney2k: I can't see what's changed? | 19:59 |
lisitsyn | is that %feature ("ref") before include? | 20:00 |
@sonney2k | lisitsyn, yes it is before the include | 20:30 |
lisitsyn | sonney2k: usual swig magic | 20:31 |
lisitsyn | sonney2k: you could write a book 'real life swig' | 20:31 |
lisitsyn | just one page | 20:32 |
lisitsyn | 'if something doesn't work just change the order of something or add some forward declaration' | 20:32 |
@sonney2k | lisitsyn, yeah | 20:32 |
@sonney2k | ok lets try to put the feature("ref") after the include stuff | 20:34 |
@sonney2k | lisitsyn, how is life btw? still overloaded? | 20:42 |
@sonney2k | or any improvement in sight? | 20:42 |
lisitsyn | sonney2k: oh well yeah something is getting better actually | 20:44 |
lisitsyn | sonney2k: what about you? | 20:44 |
@sonney2k | lisitsyn, I wonder whether you will ever have time for shogun again ... | 20:44 |
lisitsyn | sonney2k: I am stucked accuracy wise with some detector, once I train it properly I am done :D | 20:44 |
lisitsyn | sounds easy | 20:44 |
lisitsyn | sonney2k: I think so | 20:45 |
@sonney2k | lisitsyn, it is really bad that heiko is still off for another month | 20:45 |
lisitsyn | sonney2k: I have one major thing I want to change | 20:45 |
lisitsyn | it can't get out of my mind | 20:45 |
lisitsyn | these parameters | 20:45 |
lisitsyn | :D | 20:45 |
lisitsyn | always on my mind | 20:45 |
@sonney2k | heh | 20:46 |
lisitsyn | sonney2k: yeah heiko is earth mover | 20:46 |
lisitsyn | but well it is somewhat good | 20:46 |
@sonney2k | I want more cool demos / examples | 20:46 |
@sonney2k | and debian etc packages | 20:46 |
lisitsyn | sonney2k: demos-wise I am much more prepared than before | 20:46 |
@sonney2k | I really hope besser82 finishes his cmake endeavour sound | 20:47 |
@sonney2k | soon | 20:47 |
lisitsyn | I use d3/whatever js stack at my job | 20:47 |
@sonney2k | lisitsyn, cool | 20:47 |
@sonney2k | so we should add some more cool ones | 20:47 |
-!- iglesiasg [~iglesiasg@206.Red-83-61-168.dynamicIP.rima-tde.net] has joined #shogun | 20:47 | |
-!- mode/#shogun [+o iglesiasg] by ChanServ | 20:47 | |
@sonney2k | lisitsyn, your road sign thing or some other obj detection stuff | 20:47 |
@sonney2k | or whaterver | 20:47 |
@sonney2k | I actually really think we should do all the teaching examples that people normally do in ML classes | 20:48 |
lisitsyn | sonney2k: I want to understand what ML moves to and follow that | 20:49 |
lisitsyn | sonney2k: I had a lot of fun with some changes I see | 20:49 |
lisitsyn | deep learning is not everywhere anymore | 20:49 |
@sonney2k | lisitsyn, and btw while we are at it - I am thinking of maybe we should have a weekly meeting - whoever has tiem can attend | 20:49 |
@sonney2k | lisitsyn, just follow jmlr / nips / icml | 20:50 |
lisitsyn | well we can do that less official | 20:50 |
lisitsyn | just talk :D | 20:50 |
@sonney2k | yes | 20:50 |
lisitsyn | whenever we can | 20:50 |
@sonney2k | like a beer hour or sth | 20:50 |
lisitsyn | this normally worked before actually | 20:50 |
@sonney2k | well no a fixed day with a fixed time | 20:50 |
@sonney2k | so people know that everyone *tries* to be there | 20:50 |
lisitsyn | sonney2k: jmlr/nips/icml wise | 20:51 |
lisitsyn | you know I am not a ML researcher so can't really follow it deeply | 20:51 |
@sonney2k | e.g some 30 mins slot | 20:51 |
lisitsyn | but I also can't get out of a feeling | 20:51 |
lisitsyn | that 95% is useless waste of paper | 20:51 |
@sonney2k | lisitsyn, well you are ML guyish enough | 20:51 |
-!- iglesiasg [~iglesiasg@206.Red-83-61-168.dynamicIP.rima-tde.net] has quit [Ping timeout: 240 seconds] | 20:51 | |
@sonney2k | lisitsyn, yes that is normal :D | 20:51 |
lisitsyn | I can't filter out things that are needed | 20:52 |
lisitsyn | and ones that are not | 20:52 |
@sonney2k | lisitsyn, but useless depends also an what you want tod od | 20:52 |
@sonney2k | to do | 20:52 |
lisitsyn | oh yeah that makes me crazy | 20:52 |
lisitsyn | it is unresolved in my mind still | 20:53 |
@sonney2k | lisitsyn, I think science is always like this | 20:53 |
@sonney2k | it explores all possible ways | 20:53 |
@sonney2k | and there is of course some mainstream trend | 20:53 |
@sonney2k | like say music genres | 20:53 |
@sonney2k | or whatever mainstream movies | 20:54 |
@sonney2k | and then the experimental stuff | 20:54 |
lisitsyn | sonney2k: music wise - I am into that now http://instagram.com/p/gqYIMWzfYs/ | 20:54 |
lisitsyn | :D | 20:54 |
@sonney2k | yeah you are young and ahve time :) | 20:54 |
@sonney2k | lisitsyn, btw I also did some d3 at work | 20:55 |
lisitsyn | sonney2k: well I get back to home and have things to do with other work but just can't do anything | 20:55 |
lisitsyn | so I thought it would be good to learn some stuff haha | 20:55 |
lisitsyn | sonney2k: yeah I remember you said about that | 20:55 |
lisitsyn | you were secret as usual so no idea | 20:56 |
lisitsyn | we use it to draw some time-space graph of railroad | 20:56 |
lisitsyn | in 24h scale | 20:56 |
@sonney2k | lisitsyn, well yeah I signed some contract right? So I better not break the rules... | 20:56 |
lisitsyn | sonney2k: indeed :) | 20:57 |
lisitsyn | sonney2k: don't you feel all that web stuff is much more funnier that anything else? | 20:59 |
@sonney2k | lisitsyn, yeah it is cool | 20:59 |
@sonney2k | aha | 20:59 |
lisitsyn | I actually have some explanation | 20:59 |
@sonney2k | /home/sonne/shogun/src/shogun/base/SGObject.h:102: Warning 401: Base class 'SGRefObject' undefined. | 20:59 |
lisitsyn | you see feedback instantly | 20:59 |
lisitsyn | the cycle is much shorter | 21:00 |
@sonney2k | lisitsyn, and it is eye candy | 21:00 |
lisitsyn | you just change something and you see it with your eyes | 21:00 |
@sonney2k | lisitsyn, problem though is that you cannot properly test things | 21:00 |
lisitsyn | no that's not that impossible | 21:00 |
lisitsyn | there are a lot of testing things for js | 21:01 |
-!- iglesiasg [~iglesiasg@206.Red-83-61-168.dynamicIP.rima-tde.net] has joined #shogun | 21:01 | |
-!- mode/#shogun [+o iglesiasg] by ChanServ | 21:02 | |
@iglesiasg | lisitsyn, thanks! :) | 21:02 |
@sonney2k | iglesiasg, I guess it is the best decision you could make weather wise! | 21:08 |
@iglesiasg | sonney2k, it actually snew in Spain a couple of days ago, but no snow in Stockholm yet this year, crazy! | 21:08 |
@sonney2k | iglesiasg, that reminds me of a visit of Barcelona a few years back where there were like 10cm snow | 21:09 |
@sonney2k | I think in october / november | 21:09 |
@iglesiasg | wow in Barcelona | 21:09 |
@iglesiasg | beach place, so even more unlikely I guess! | 21:10 |
@sonney2k | iglesiasg, traffic broke down that day | 21:11 |
@sonney2k | but snow was more a mush then real snow | 21:12 |
@sonney2k | and I was 0% prepared for that | 21:12 |
-!- ChanServ changed the topic of #shogun to: Shogun Machine Learning Toolbox | Devs are usually around between 10am-10pm UTC | When asking questions please be patient or use the mailing list. We always answer but it might take a while. | Channel logs http://bit.ly/1eBecWl | 21:14 | |
@sonney2k | iglesiasg, nice that you will continue the readme*md endaevour | 21:16 |
@iglesiasg | everything broke down here when it snows, even if it is very very little | 21:16 |
@iglesiasg | sonney2k, sure | 21:16 |
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout] | 21:20 | |
@sonney2k | lisitsyn, seen that http://www.p01.org/releases/wolfensteiny/ ? | 21:37 |
lisitsyn | sonney2k: ha! no | 21:37 |
@sonney2k | http://www.p01.org/releases/tea_storm/tea_storm.htm | 21:38 |
@sonney2k | I guess this guy would write shogun in .js in 10kb ;) | 21:39 |
lisitsyn | well | 21:39 |
lisitsyn | no :) | 21:39 |
@sonney2k | lisitsyn, do you have any idea how I could in small incremental steps get rid of SGString* crap? | 21:42 |
@sonney2k | and replace it with SGVector* goodness? | 21:42 |
lisitsyn | sonney2k: incremental hmm | 21:42 |
@sonney2k | I did a couple of attempts | 21:42 |
@sonney2k | but they were all trying too much | 21:42 |
lisitsyn | sonney2k: what about typedef? | 21:42 |
@sonney2k | so no way of every finishing | 21:42 |
@sonney2k | lisitsyn, ?? | 21:43 |
lisitsyn | typedef SGVector SGString | 21:43 |
@sonney2k | haha | 21:43 |
lisitsyn | with C++11 you can do template typedef | 21:43 |
@sonney2k | that kaboom | 21:43 |
lisitsyn | yes of course | 21:43 |
lisitsyn | but you could change SGVector to support the same interface then | 21:43 |
@sonney2k | well I would want to have something less drastic | 21:43 |
@sonney2k | lisitsyn, hmmhh | 21:43 |
@sonney2k | I think there are 2 major steps | 21:44 |
@sonney2k | 1) is derive it from SGReferencedData | 21:44 |
lisitsyn | ahhhhhhh | 21:44 |
lisitsyn | SGString is not sgdata | 21:44 |
lisitsyn | sure sure | 21:44 |
@sonney2k | 2) switch over to ref'd | 21:44 |
lisitsyn | now I understand | 21:44 |
@sonney2k | err vector | 21:44 |
@sonney2k | yeah | 21:44 |
@sonney2k | it is the only data type we didn't finish | 21:44 |
@sonney2k | converting | 21:44 |
lisitsyn | many brave souls died during the first phase | 21:45 |
lisitsyn | remember I rolled things back | 21:45 |
lisitsyn | :D | 21:45 |
@sonney2k | lisitsyn, unfortunately | 21:45 |
@sonney2k | you should have kept things | 21:45 |
lisitsyn | well it wasn't working at all | 21:46 |
lisitsyn | like you have everything failing | 21:46 |
@sonney2k | maybe it is impossible to do 1) incrementally | 21:46 |
lisitsyn | sonney2k: http://2.bp.blogspot.com/-41v6n3Vaf5s/UeRN_XJ0keI/AAAAAAAAN2Y/YxIHhddGiaw/s1600/css.gif | 21:46 |
lisitsyn | like that | 21:46 |
@sonney2k | yeah it certainly is like that | 21:47 |
lisitsyn | sonney2k: well may be some solution would be | 21:52 |
lisitsyn | do 1) but don't delete | 21:52 |
lisitsyn | so you won't get terrible failures yet | 21:52 |
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun | 21:56 | |
shogun-notifier- | shogun: Soeren Sonnenburg :develop * 5844f48 / / (14 files): https://github.com/shogun-toolbox/shogun/commit/5844f480a1b588e5fccfef90d1df096579467d19 | 21:56 |
shogun-notifier- | shogun: don't include SGStringList in SGObject.h | 21:56 |
@sonney2k | lisitsyn, yeah maybe I can split it up even | 21:58 |
@sonney2k | lisitsyn, first make SGStringList 'refable' | 21:58 |
@sonney2k | and then later sgstring | 21:58 |
@sonney2k | lisitsyn, haha | 21:59 |
@sonney2k | SGStringList is already! | 21:59 |
lisitsyn | sonney2k: one step done :D | 22:00 |
@sonney2k | *lol* | 22:00 |
@sonney2k | lisitsyn, actually one way to 'fix' things would be to make SGString 'refable' and then use only the 'disable refcount' constructor | 22:00 |
@sonney2k | lisitsyn, and then gradually switch over to use the refcount | 22:01 |
lisitsyn | sonney2k: yes makes sense | 22:01 |
lisitsyn | but I don't see gradually here | 22:01 |
lisitsyn | :D | 22:01 |
lisitsyn | you just turn it on and then go to hospital to heal wounds | 22:01 |
-!- iglesiasg [~iglesiasg@206.Red-83-61-168.dynamicIP.rima-tde.net] has quit [Quit: Leaving] | 22:06 | |
-!- travis-ci [~travis-ci@ec2-54-242-242-147.compute-1.amazonaws.com] has joined #shogun | 22:39 | |
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/14746073 | 22:39 |
-!- travis-ci [~travis-ci@ec2-54-242-242-147.compute-1.amazonaws.com] has left #shogun [] | 22:39 | |
shogun-buildbot_ | build #2062 of deb3 - modular_interfaces is complete: Failure [failed test python modular test ruby modular test java modular test csharp modular test lua modular test octave modular test r modular] Build details are at http://buildbot.shogun-toolbox.org/builders/deb3%20-%20modular_interfaces/builds/2062 blamelist: Soeren Sonnenburg <sonne@debian.org> | 23:45 |
--- Log closed Sun Dec 01 00:00:47 2013 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!