--- Log opened Mon May 27 00:00:19 2013 | ||
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun | 00:40 | |
shogun-notifier- | shogun: Kevin Hughes :develop * 7d14502 / src/shogun/multiclass/QDA.cpp: https://github.com/shogun-toolbox/shogun/commit/7d14502376e2cee7d0b2fdcdc0605b51f4802145 | 00:40 |
---|---|---|
shogun-notifier- | shogun: fixed crash in QDA covar calculation | 00:40 |
shogun-notifier- | shogun: Soeren Sonnenburg :develop * bfae181 / src/shogun/multiclass/QDA.cpp: https://github.com/shogun-toolbox/shogun/commit/bfae18199f8dd2f1e609a4c7faf30a6aa92f02ab | 00:40 |
shogun-notifier- | shogun: Merge pull request #1131 from pickle27/qda | 00:40 |
shogun-notifier- | shogun: | 00:40 |
shogun-notifier- | shogun: fixed crash in QDA covar calculation | 00:40 |
-!- travis-ci [~travis-ci@ec2-107-22-28-98.compute-1.amazonaws.com] has joined #shogun | 01:01 | |
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/7519558 | 01:01 |
-!- travis-ci [~travis-ci@ec2-107-22-28-98.compute-1.amazonaws.com] has left #shogun [] | 01:01 | |
shogun-buildbot | build #1166 of deb3 - modular_interfaces is complete: Failure [failed test python_modular] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/1166 blamelist: Kevin Hughes <kevin@kevin-M3920.(none)> | 01:21 |
shogun-buildbot | build #1167 of deb3 - modular_interfaces is complete: Failure [failed test python_modular] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/1167 blamelist: Soeren Sonnenburg <sonne@debian.org> | 01:49 |
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout] | 03:40 | |
shogun-buildbot | build #408 of nightly_default is complete: Failure [failed test] Build details are at http://www.shogun-toolbox.org/buildbot/builders/nightly_default/builds/408 | 03:44 |
-!- nube [~rho@49.244.8.172] has quit [Quit: Leaving.] | 03:44 | |
-!- nube [~rho@116.90.239.3] has joined #shogun | 05:51 | |
-!- nube [~rho@116.90.239.3] has quit [Ping timeout: 240 seconds] | 06:57 | |
-!- nube [~rho@116.90.239.3] has joined #shogun | 07:02 | |
gsomix | good morning | 07:04 |
gsomix | sonney2k, good. tnx | 07:05 |
-!- foulwall [~foulwall@li379-21.members.linode.com] has joined #shogun | 07:52 | |
-!- gsomix [~gsomix@188.168.2.227] has quit [Quit: Leaving] | 08:17 | |
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun | 08:37 | |
shogun-notifier- | shogun-data: Soeren Sonnenburg :master * ca1d0b6 / testsuite/tests/ (2 files): https://github.com/shogun-toolbox/shogun-data/commit/ca1d0b696150224cc280f9daff612124852d5025 | 08:37 |
shogun-notifier- | shogun-data: update qda data | 08:37 |
shogun-notifier- | shogun: Soeren Sonnenburg :develop * 363d6c2 / data,tests/integration/python_modular/generator.py: https://github.com/shogun-toolbox/shogun/commit/363d6c273b60f538948589e9ae9b6f9130983f70 | 08:38 |
shogun-notifier- | shogun: update data and add director examples to blacklist | 08:38 |
@sonney2k | foulwall, good morning! | 08:39 |
shogun-notifier- | shogun: Soeren Sonnenburg :develop * 93df87f / src/shogun/statistics/MMDKernelSelectionComb.cpp: https://github.com/shogun-toolbox/shogun/commit/93df87f1f531ab6cb03107092e7d9b740b60595b | 08:39 |
shogun-notifier- | shogun: fix typo in warning msg | 08:39 |
foulwall | sonney2k: morning, I'll have my first final this afternoon. | 08:42 |
@sonney2k | foulwall, oh good luck then and lets talk again tomorrow once google sent out the notifications! | 08:43 |
@sonney2k | pickle27, I've regenerated the qda integration test data | 08:44 |
foulwall | sonney2k: thanks~ | 08:44 |
@sonney2k | so finally tests should all pass | 08:44 |
@sonney2k | hushell, done with nips paper :D | 08:44 |
@sonney2k | oops gtg | 08:45 |
shogun-buildbot | build #876 of cyg1 - libshogun is complete: Failure [failed configure] Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/876 blamelist: Soeren Sonnenburg <sonne@debian.org> | 08:58 |
-!- travis-ci [~travis-ci@ec2-184-72-84-14.compute-1.amazonaws.com] has joined #shogun | 08:59 | |
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/7525685 | 08:59 |
-!- travis-ci [~travis-ci@ec2-184-72-84-14.compute-1.amazonaws.com] has left #shogun [] | 08:59 | |
-!- vgorbati [5f6fea63@gateway/web/freenode/ip.95.111.234.99] has joined #shogun | 09:05 | |
-!- sonne|work [~sonnenbu@sams-office-nat.tomtomgroup.com] has joined #shogun | 09:13 | |
sonne|work | Re | 09:13 |
hushell | sonney2k: :D not as quick as I expected, but all results are obtained, I am writing introduction ... | 09:17 |
sonne|work | hehe | 09:17 |
-!- travis-ci [~travis-ci@ec2-184-72-84-14.compute-1.amazonaws.com] has joined #shogun | 09:18 | |
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/7525713 | 09:18 |
-!- travis-ci [~travis-ci@ec2-184-72-84-14.compute-1.amazonaws.com] has left #shogun [] | 09:18 | |
hushell | sonney2k: gsoc starts next week? | 09:18 |
hushell | writing is painful | 09:19 |
sonne|work | hushell: today will be notifications | 09:19 |
sonne|work | of acceptance/reject | 09:19 |
sonne|work | from then on everyone can start but official start is slightly later | 09:19 |
hushell | !!!exciting | 09:20 |
hushell | also nervous! | 09:21 |
shogun-buildbot | build #1168 of deb3 - modular_interfaces is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/1168 | 09:21 |
sonne|work | june 17 is the official start - but it sure won't hurt to start before | 09:22 |
hushell | I am expecting for the gsoc for a long time, writing is so boring | 09:23 |
-!- nube [~rho@116.90.239.3] has quit [Ping timeout: 245 seconds] | 09:23 | |
-!- nube [~rho@116.90.239.3] has joined #shogun | 09:24 | |
-!- iglesiasg [d58f3247@gateway/web/freenode/ip.213.143.50.71] has joined #shogun | 09:32 | |
-!- mode/#shogun [+o iglesiasg] by ChanServ | 09:32 | |
-!- iglesiasg [d58f3247@gateway/web/freenode/ip.213.143.50.71] has quit [Client Quit] | 09:36 | |
-!- lambday [67157d36@gateway/web/freenode/ip.103.21.125.54] has joined #shogun | 09:41 | |
shogun-buildbot | build #1169 of deb3 - modular_interfaces is complete: Failure [failed test python_modular] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/1169 blamelist: Soeren Sonnenburg <sonne@debian.org> | 09:44 |
-!- lisitsyn [~lisitsin@mxs.kg.ru] has joined #shogun | 09:44 | |
-!- mode/#shogun [+o lisitsyn] by ChanServ | 09:44 | |
@lisitsyn | ha-ha | 09:44 |
@lisitsyn | sonne|work: I am back from no-internet-connection-land! | 09:45 |
sonne|work | lisitsyn: siberia? | 09:45 |
sonne|work | had fun playing with your grizzly bears? | 09:45 |
sonne|work | or whatever you call them up there? | 09:45 |
shogun-notifier- | shogun: Soeren Sonnenburg :develop * d6cb133 / tests/integration/python_modular/generator.py: https://github.com/shogun-toolbox/shogun/commit/d6cb133556c7e82ec4f1641f2fab4dcb5e77cca1 | 09:49 |
shogun-notifier- | shogun: fix typo in blacklist | 09:49 |
sonne|work | lisitsyn: anyways just in time to watch the test to go red -> green | 09:49 |
-!- van51 [~van51@athedsl-320452.home.otenet.gr] has joined #shogun | 09:52 | |
sonne|work | van51: welcome back | 09:52 |
van51 | sonne|work: hello | 09:53 |
@lisitsyn | sonne|work: no electricity due to some fire :) | 10:07 |
-!- travis-ci [~travis-ci@ec2-107-22-28-98.compute-1.amazonaws.com] has joined #shogun | 10:11 | |
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/7527064 | 10:11 |
-!- travis-ci [~travis-ci@ec2-107-22-28-98.compute-1.amazonaws.com] has left #shogun [] | 10:11 | |
shogun-buildbot | build #877 of cyg1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/877 | 10:16 |
-!- thoralf [~thoralf@enki.zib.de] has joined #shogun | 10:25 | |
thoralf | Good morning | 10:25 |
shogun-buildbot | build #1170 of deb3 - modular_interfaces is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/1170 | 10:34 |
-!- vgorbati [5f6fea63@gateway/web/freenode/ip.95.111.234.99] has quit [Ping timeout: 250 seconds] | 10:35 | |
@lisitsyn | sonne|work: did you know ray dassen? | 10:41 |
sonne|work | no | 10:41 |
sonne|work | who was/is he | 10:41 |
sonne|work | ? | 10:42 |
@lisitsyn | debian developer | 10:42 |
@lisitsyn | passed away yesterday | 10:42 |
@lisitsyn | oh not yesterday | 10:42 |
@lisitsyn | 23th | 10:42 |
@lisitsyn | sonne|work: http://www.debian.org/News/2013/20130523 | 10:42 |
sonne|work | we all die... | 10:43 |
sonne|work | wow around for 19 years in debian | 10:44 |
@lisitsyn | yes quite a long period | 10:44 |
@lisitsyn | that's I am curious whether he was famous around debianeers | 10:44 |
-!- HeikoS [~heiko@nat-191-4.internal.eduroam.ucl.ac.uk] has joined #shogun | 10:51 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 10:51 | |
-!- van51 [~van51@athedsl-320452.home.otenet.gr] has left #shogun ["QUIT :Leaving."] | 10:58 | |
-!- iglesiasg [c1934d16@gateway/web/freenode/ip.193.147.77.22] has joined #shogun | 11:09 | |
-!- mode/#shogun [+o iglesiasg] by ChanServ | 11:10 | |
-!- lambday [67157d36@gateway/web/freenode/ip.103.21.125.54] has quit [Ping timeout: 250 seconds] | 11:13 | |
-!- lambday [67157d37@gateway/web/cgi-irc/kiwiirc.com/ip.103.21.125.55] has joined #shogun | 11:14 | |
lambday | HeikoS: hi | 11:15 |
@HeikoS | lambday: hi! | 11:15 |
lambday | got your mail | 11:15 |
lambday | I did't understand what you meant by an extension of shogun framework | 11:15 |
lambday | for integrating and testing new algorithms in shogun? | 11:16 |
lambday | as in, in future? | 11:16 |
@HeikoS | lambday: A way of generalising independent computation tasks might be very useful for other parts | 11:16 |
@HeikoS | take for example CMachine model-selection via grod-search | 11:16 |
@HeikoS | this is a stupid algorithm, but it is used by many people | 11:16 |
@HeikoS | you basically train/test machines with different parameters | 11:17 |
@HeikoS | all these tasks are independent | 11:17 |
lambday | okay | 11:17 |
lambday | hmm | 11:17 |
@HeikoS | you we could formulise them as a Task as in the class diagram | 11:17 |
@lisitsyn | hah stupid == working HeikoS ;) | 11:17 |
@HeikoS | then pass them to an engine that solves them in some way | 11:17 |
-!- lambday [67157d37@gateway/web/cgi-irc/kiwiirc.com/ip.103.21.125.55] has quit [Client Quit] | 11:17 | |
@HeikoS | lisitsyn: what? | 11:17 |
@HeikoS | ah | 11:18 |
@HeikoS | well | 11:18 |
@lisitsyn | HeikoS: grid search is stupid but it is a working algorithm | 11:18 |
@lisitsyn | :) | 11:18 |
@HeikoS | lisitsyn: it is a ridiculus thing in my eyes, but what can you do :) | 11:18 |
@HeikoS | people just like simple algorithms | 11:18 |
@HeikoS | brute force stuff :D | 11:19 |
@lisitsyn | HeikoS: but how would you approach it | 11:19 |
@lisitsyn | I mean how do you search C then? | 11:19 |
@HeikoS | there are many way to do this, all described early 2000s | 11:19 |
@HeikoS | similar to the way its done with GPs | 11:19 |
@HeikoS | gradient descent on some upper bound of the error | 11:19 |
@lisitsyn | well | 11:19 |
@lisitsyn | you'd have to have a great API to make that common | 11:20 |
@lisitsyn | HeikoS: it is easy to reproduce brute solution | 11:20 |
@lisitsyn | and they always work for you | 11:21 |
@HeikoS | all true | 11:21 |
@HeikoS | but we have this for GPs | 11:21 |
@HeikoS | just need to implement some of these error functions | 11:21 |
-!- lambday [67157e4e@gateway/web/cgi-irc/kiwiirc.com/ip.103.21.126.78] has joined #shogun | 11:23 | |
@HeikoS | lisitsyn: sorry I dont want to dizz this stuff, just my thoughts | 11:23 |
@lisitsyn | HeikoS: what do you mean? :) | 11:23 |
@HeikoS | its already pretty good that we have grid-serach for arbritary machines | 11:23 |
@HeikoS | I mean its fine as it is | 11:23 |
@lisitsyn | HeikoS: I just wanted to note for stupid engineers like me it is working solution | 11:24 |
lambday | sorry I got disconnected :( | 11:24 |
@HeikoS | lambday: no worries | 11:24 |
@lisitsyn | I don't give a heck how to approach it else :) | 11:24 |
@lisitsyn | probably being stupid is a problem of mine though :D | 11:24 |
@HeikoS | haha, I wouldnt agree ;) | 11:25 |
@lisitsyn | HeikoS: btw recalling that syntax I was pushing | 11:25 |
@lisitsyn | are we still up for it? | 11:26 |
@HeikoS | yes totally | 11:26 |
@lisitsyn | hmm who could do that then | 11:26 |
@HeikoS | lisitsyn: btw there is also a way of speeding the std-grid search up quite a bit through some statistical tricks | 11:27 |
@HeikoS | sonney2k: pointed this out at some point | 11:27 |
@HeikoS | might be nice to have this | 11:27 |
@lisitsyn | HeikoS: like MCMC or so? | 11:27 |
@HeikoS | no, MCMC and the word speed never go together :D | 11:27 |
-!- van51 [~van51@195.251.235.71] has joined #shogun | 11:27 | |
@lisitsyn | HeikoS: you can't touch this | 11:27 |
@HeikoS | haha :) | 11:28 |
-!- gsomix [~gsomix@37.61.179.240] has joined #shogun | 11:28 | |
gsomix | hello! | 11:28 |
@lisitsyn | can't touch this | 11:28 |
@lisitsyn | HeikoS: agaaain | 11:28 |
@HeikoS | http://blog.mikiobraun.de/2011/12/fast-cross-validation.html | 11:28 |
@lisitsyn | whoop | 11:28 |
@lisitsyn | ahhh | 11:28 |
@lisitsyn | HeikoS: I recall some core-sh*t from NIPS | 11:28 |
@lisitsyn | sonne|work told us about it | 11:28 |
@lisitsyn | core sets or so | 11:28 |
@lisitsyn | I don't recall | 11:29 |
@HeikoS | you drop parameters which are statistically significant loosers against others | 11:29 |
@lisitsyn | sounds genetic | 11:29 |
@lisitsyn | HeikoS: can't touch this | 11:29 |
@HeikoS | not really | 11:29 |
@lisitsyn | jk | 11:29 |
@lisitsyn | :) | 11:29 |
@HeikoS | anyway, we should have a look at this at some point | 11:30 |
@HeikoS | problem is that its not really parallelisable anymore | 11:30 |
sonne|work | gsomix: finaly a function that is readable :) | 11:31 |
@HeikoS | or that to that level | 11:31 |
@HeikoS | but other things first | 11:31 |
@lisitsyn | HeikoS: I would go for stupid but parallel solution actually | 11:31 |
@HeikoS | lisitsyn: having both would be nice | 11:31 |
@HeikoS | if you have many cores, go for the easy, if not, go for the clever | 11:31 |
@lisitsyn | our CPUs are fast and our RAM is going infinite :) | 11:31 |
gsomix | sonne|work, horray! I also got yours comments. | 11:31 |
@lisitsyn | HeikoS: you can't touch this | 11:33 |
@lisitsyn | :D | 11:33 |
@lisitsyn | srry | 11:33 |
@HeikoS | haha :D | 11:33 |
@lisitsyn | HeikoS: this should be an anthem of shogun | 11:34 |
@lisitsyn | HeikoS: when you are entering an website | 11:34 |
@lisitsyn | you should hearr that | 11:34 |
@HeikoS | hehe | 11:35 |
@HeikoS | lambday: btw the Task.sove method should better be named .compute() | 11:35 |
@lisitsyn | HeikoS: are you going to push that mcmc hammer to shogun? | 11:35 |
lambday | HeikoS: okay, changing | 11:36 |
lambday | :) | 11:36 |
@HeikoS | lisitsyn: probably | 11:36 |
@HeikoS | we got a name clash unfortunately :) | 11:36 |
@HeikoS | have to rename it | 11:36 |
@lisitsyn | HeikoS: I am submitting tapkee paper *today* so I am going to join your game with tasks at some point | 11:36 |
@lisitsyn | HeikoS: no way, whyyy | 11:36 |
@HeikoS | google for it | 11:36 |
@HeikoS | mcmc hammer | 11:36 |
@lisitsyn | hahah | 11:36 |
@HeikoS | lisitsyn: ah shit, I never gave you my comments did I? | 11:36 |
@HeikoS | lisitsyn: cool, that would help a lot | 11:36 |
@lisitsyn | HeikoS: hmm you did I think | 11:37 |
@lisitsyn | I don't remember | 11:37 |
@HeikoS | really, good :) | 11:37 |
@lisitsyn | anyway nevermind I can't wait much longer :) | 11:37 |
@HeikoS | me neither | 11:37 |
@HeikoS | ok | 11:37 |
sonne|work | gsomix: you should write a test for this function though | 11:37 |
@lisitsyn | I've been working on it for years | 11:37 |
@lisitsyn | and now I am 62 and no submission still yet | 11:38 |
sonne|work | gsomix: we have enough ascii files to try :D | 11:38 |
@lisitsyn | I had one ascii file somewhere | 11:38 |
@lisitsyn | HeikoS: https://twitter.com/Rick_Moyer/status/278641295285698560 LOL | 11:38 |
@HeikoS | lambday: could you write a seperate diagram for all the computation task stuff (without the implementations, just base classes) | 11:38 |
@HeikoS | lisitsyn: thats it | 11:39 |
@HeikoS | lambday: will help discussions with Sergey | 11:39 |
lambday | HeikoS: okay...you mean, computation task and computation task result classes? | 11:39 |
@HeikoS | yeah just those for now, will be a small diagram and you can do this with copy paste | 11:40 |
@lisitsyn | can't touch this diagram | 11:40 |
lambday | HeikoS: okay... mailing | 11:40 |
gsomix | sonne|work, really? I need one ~#N GB sized file. :) | 11:40 |
@lisitsyn | ohohohoho | 11:40 |
@lisitsyn | I recalled | 11:40 |
@lisitsyn | wiking: you made random() const! | 11:41 |
@lisitsyn | bad bad bad boy | 11:41 |
sonne|work | *lol* | 11:41 |
sonne|work | gsomix: you could use data/asp/Human.dat.bz2 | 11:42 |
sonne|work | it is ~80M | 11:42 |
@lisitsyn | let us have | 11:42 |
@lisitsyn | uint64_t a = random64(); | 11:42 |
@lisitsyn | uint64_t b = random64(); | 11:42 |
sonne|work | err 100M | 11:42 |
@lisitsyn | the question is | 11:42 |
@lisitsyn | whether a==b | 11:42 |
@lisitsyn | I'd say -O3 makes it *true* | 11:42 |
gsomix | sonne|work, ok. | 11:42 |
@lisitsyn | okay not really sure but it could | 11:42 |
@lisitsyn | sonne|work: ^ | 11:43 |
sonne|work | lisitsyn: sorry btu why | 11:43 |
@lisitsyn | sonne|work: because it looks like a pure function if it is const | 11:43 |
@lisitsyn | why to recompute its value twice if it has no side-effects | 11:43 |
@lisitsyn | which cause changing its value | 11:43 |
@lisitsyn | that's the meaning of const | 11:43 |
sonne|work | but only inside the CRandom class | 11:43 |
@lisitsyn | yes but we have to check | 11:44 |
@lisitsyn | compiler is magic | 11:44 |
lambday | HeikoS: just these two base classes? but these contains almost nothing here, their subclasses do | 11:44 |
@lisitsyn | random function is not const | 11:44 |
sonne|work | lisitsyn: well tests work so you are wrong | 11:44 |
@lisitsyn | it is as mutable as it can be | 11:44 |
sonne|work | but if tests are compiled with -O0 then :) | 11:44 |
@lisitsyn | sonne|work: vgorbati told me we have an issue with calling normal only once | 11:44 |
@lisitsyn | I didn't check due to blackout | 11:44 |
@lisitsyn | and now recalled | 11:45 |
@lisitsyn | will check tonight | 11:45 |
@HeikoS | lambday: maybe an example of a class that has generate_tasks / aggregate results | 11:45 |
sonne|work | lisitsyn: normal? | 11:45 |
@lisitsyn | normal_random I mean | 11:45 |
@lisitsyn | sonne|work: anyway it is not const | 11:45 |
sonne|work | hmmhh, we have to move the other gaussian generator from CMath | 11:45 |
@HeikoS | lambday: I am sorry, that was asked too quickly. Just leave it for now | 11:45 |
lambday | HeikoS: okay.. | 11:46 |
@HeikoS | lambday: we will use the full diagram for discussions | 11:46 |
@HeikoS | its clearer, you are right | 11:46 |
@lisitsyn | const tells me calling this function sequentially twice will return the same value | 11:46 |
lambday | HeikoS: yes its pretty clear when we see this whole picture, and log-det as an example of that | 11:47 |
lambday | this task idea was awesome!! :) | 11:47 |
@HeikoS | lambday: lets hope it is possible ;) | 11:48 |
lambday | HeikoS: implementation? | 11:49 |
@HeikoS | lambday: as lisitsyn pointed out, there are always difficulties if one wants to write those things in a general way | 11:49 |
@lisitsyn | can't touch this | 11:50 |
lambday | HeikoS: but its good to have this general framework, no? | 11:50 |
@HeikoS | lambday: yes, it would be very useful | 11:50 |
@lisitsyn | lambday: I just mean we can't foresee all the stuff | 11:50 |
lambday | lisitsyn: true that | 11:51 |
@lisitsyn | there is a wall right forward actually | 11:51 |
@lisitsyn | we are moving to | 11:51 |
@lisitsyn | :D | 11:51 |
@HeikoS | lisitsyn, lambday: sonne|work pointed out that we can even have a qsub-based implementaton for the engine | 11:51 |
lambday | lisitsyn: all in all its just another brick in the wall, then :P | 11:51 |
@lisitsyn | once we crashed into the wall we start refactoring | 11:51 |
lambday | qsub-based implementation? | 11:52 |
@lisitsyn | curing our pain in the ass :D | 11:52 |
@HeikoS | lambday: yes for qsub based clusters | 11:52 |
lambday | wow! | 11:53 |
@lisitsyn | HeikoS: we will become graphlab you have to live with that | 11:53 |
@lisitsyn | :D | 11:53 |
@HeikoS | lisitsyn: we will be fine! all independent | 11:53 |
lambday | HeikoS: i'll shortly be back.. starving.. gotta eat something... about the work, the picture is quite clear now... what should be the next step? adding all the implementations? | 11:58 |
lambday | HeikoS: and we don't need the linear system class then | 11:58 |
@HeikoS | lambday: indeed, only the linear solver, which takes a linear operator and a vector | 11:59 |
lambday | HeikoS: okay.. | 11:59 |
@HeikoS | lambday: once everything is clear, I would write all the base classes | 11:59 |
@HeikoS | to get a feeling whether things actually work | 11:59 |
@HeikoS | make sure to start a new feature branch with git-flow | 12:00 |
@HeikoS | if those are done, refine class diagram | 12:00 |
@HeikoS | then some simple implementations | 12:00 |
lambday | HeikoS: okay.. | 12:01 |
lambday | so, it may be good to have a diagram with the base classes only | 12:02 |
@HeikoS | lambday: and could you send the current class diagram to lisitsyn? | 12:02 |
@HeikoS | lambday: yes, totally | 12:02 |
@HeikoS | also without any overhead in the diagram | 12:02 |
@HeikoS | just plain classes | 12:02 |
@HeikoS | it gets too big otherwise | 12:02 |
lambday | HeikoS: okay.. | 12:03 |
lambday | HeikoS: oh I almost forgot... the CRationalApproximationTaskResults inherits from both CIndependentComputationTaskResults and CIndependentComputationTask... I didn't understand this | 12:06 |
@lisitsyn | naming reminds me java | 12:07 |
@lisitsyn | :D | 12:07 |
lambday | lisitsyn: :D | 12:07 |
lambday | but sounds good and clear | 12:07 |
@lisitsyn | let me find the longest class name in my current proj | 12:07 |
@lisitsyn | hmm I am not a java programmer | 12:08 |
@lisitsyn | the longest is AgentMessagesFactory :( | 12:08 |
sonne|work | lambday: I don't understand what this class does though so the name is not that helpful | 12:08 |
sonne|work | lisitsyn: but I know what you mean :D | 12:09 |
lambday | sonne|work: umm.... | 12:09 |
@lisitsyn | sonne|work: AbstractBusinessLayerCompositorDecompositorBuilderMasterCollaboratorFactory | 12:09 |
@HeikoS | lambday: ehm, that might have been my mistake | 12:09 |
lambday | lisitsyn: whoa! :o | 12:09 |
lambday | HeikoS: hmmm... okay then removing it | 12:10 |
@HeikoS | results doesnt inherit from task | 12:10 |
sonne|work | HeikoS: couldn't we use 'Job' ? | 12:10 |
@lisitsyn | yes Job is better | 12:10 |
lambday | yes I thought so but wanted to make sure | 12:10 |
@HeikoS | sonne|work: yes totally | 12:10 |
sonne|work | it is short | 12:10 |
@HeikoS | make it job | 12:10 |
@lisitsyn | handjob | 12:10 |
@lisitsyn | scnr | 12:10 |
@HeikoS | tztz | 12:10 |
lambday | lisitsyn: lol man :D | 12:10 |
sonne|work | lisitsyn: a few hours without power got you in the mood heh? | 12:11 |
@HeikoS | lambday: CIndependentJob, but the subclasses dont get the "Independent" | 12:11 |
lambday | HeikoS: okay | 12:11 |
lambday | HeikoS: I'll change it after lunch and mail it back soon.. also adding sample_without_averaging method you told about | 12:12 |
@lisitsyn | sonne|work: yes sure! | 12:12 |
@HeikoS | lambday: yes, maybe there is also a better name for this :) | 12:12 |
@HeikoS | lambday: enjoy your lunch | 12:13 |
lambday | be right back | 12:13 |
-!- nube [~rho@116.90.239.3] has quit [Quit: Leaving.] | 12:14 | |
@lisitsyn | HeikoS: why to have such a class like independent job? | 12:22 |
@HeikoS | algorithms can generate them under a unified interface | 12:23 |
@HeikoS | computation class can call compute() on the interface | 12:23 |
@HeikoS | computation class does not need to worry whats going under underneath | 12:23 |
@lisitsyn | HeikoS: may be just a list<Job> dependencies? | 12:23 |
@lisitsyn | if it is empty you have no dependencies | 12:24 |
@HeikoS | I would like to go without dependencies | 12:24 |
@HeikoS | the implementation of the job can still operate on say shared memory | 12:24 |
@HeikoS | or serialized files | 12:25 |
@HeikoS | but this can be done inside | 12:25 |
@HeikoS | for every job | 12:25 |
@HeikoS | I have the fear that adding this to the framework might make things complicated | 12:25 |
@HeikoS | and it is task-dependent anyways | 12:25 |
@HeikoS | lisitsyn: whats your opinion on this? | 12:28 |
@HeikoS | how would you formulate dependencies and what would they be? | 12:28 |
@lisitsyn | HeikoS: that's the question of this day | 12:28 |
@lisitsyn | :D | 12:28 |
@lisitsyn | I don't know tbh | 12:28 |
@HeikoS | lisitsyn: so an example | 12:28 |
@HeikoS | a matrix | 12:29 |
@HeikoS | large one | 12:29 |
@HeikoS | all jobs need this readonly | 12:29 |
@HeikoS | what one could do is to add a pointer to the data to each job | 12:29 |
@lisitsyn | HeikoS: I'd go for something generic step by step like read only proxies for these matrices | 12:29 |
@HeikoS | why? | 12:30 |
@lisitsyn | well that's safe | 12:30 |
@HeikoS | if we want to go multiple-machine, we *have* to serialize it anyway | 12:30 |
@HeikoS | so every job will load this matrix from disc | 12:30 |
@HeikoS | or network | 12:30 |
@lisitsyn | oh I can't touch this | 12:30 |
@lisitsyn | :D | 12:30 |
@HeikoS | and in multicore case, these things have to be read-only | 12:30 |
@lisitsyn | sounds like implementing hadoop | 12:30 |
@HeikoS | lisitsyn: this is not meant for millions of nodes | 12:31 |
@HeikoS | just a few hundred max | 12:31 |
@lisitsyn | MR in particular | 12:31 |
@HeikoS | so we can rely on the nfs | 12:31 |
@lisitsyn | uh uh I don't like doing that at all | 12:31 |
@HeikoS | this is not large-scale stuff | 12:31 |
@HeikoS | whats the problem (in our context)? | 12:32 |
@lisitsyn | I am afraid of NIH | 12:33 |
@HeikoS | whats that? | 12:33 |
@lisitsyn | never invented here syndrome ;) | 12:33 |
@HeikoS | I am not following you | 12:34 |
@lisitsyn | I mean we'd rather stick to something supported by anyone else | 12:34 |
@lisitsyn | than writing that thing | 12:34 |
@HeikoS | still not following ;) | 12:34 |
@lisitsyn | HeikoS: I mean doing it with our own serialization etc | 12:34 |
@lisitsyn | reproduces what map reduce does | 12:35 |
@HeikoS | we are much simpler | 12:35 |
@HeikoS | but yes its similar | 12:36 |
@HeikoS | so whats the alternative then? | 12:36 |
@lisitsyn | may be employ something supported | 12:37 |
@lisitsyn | HeikoS: what I am most worried is that we have to support all the stuff | 12:37 |
@lisitsyn | we already have no power/will to support all the code we have | 12:37 |
@lisitsyn | such IO routines | 12:37 |
@lisitsyn | and other infrastructure sounds painful | 12:37 |
@HeikoS | agreed, but thats a general problem | 12:37 |
@HeikoS | lisitsyn: so the multi-machine stuff is far away | 12:38 |
@HeikoS | if this worked for multi-core, it would already be very helpful | 12:38 |
@HeikoS | even if the engine is just a single core loop its helpful | 12:38 |
@HeikoS | to cast algorithms in terms of these jobs | 12:38 |
@HeikoS | since it makes transition easier | 12:38 |
@HeikoS | for the log-dets for example | 12:39 |
@HeikoS | there is no point in even doing this without this job stuff | 12:39 |
@lisitsyn | HeikoS: I can imagine we do that engines stuff | 12:39 |
@lisitsyn | but we have to take it really seriously | 12:40 |
@HeikoS | sure | 12:40 |
@lisitsyn | basically it would become a supported core while algorithms get experimental | 12:40 |
@HeikoS | I wouldnt change the existing algorithms for now | 12:40 |
@HeikoS | just try to write new ones in that style | 12:40 |
@HeikoS | if it works, fine, maybe change existing ones | 12:41 |
@HeikoS | if it doesnt (too hard in whatever way), we can just plug in the sequential engine and no problems | 12:41 |
@lisitsyn | HeikoS: I have a suggestion | 12:41 |
@lisitsyn | once you get with some solution on log-det | 12:41 |
@lisitsyn | let lambday change some existing algorithm to new 'paradigm' | 12:42 |
@HeikoS | lisitsyn: yes good idea | 12:42 |
@lisitsyn | and we will see something | 12:42 |
@HeikoS | grid-search maybe | 12:42 |
@HeikoS | or even simpler things | 12:42 |
@HeikoS | kernel matrix | 12:42 |
@HeikoS | one job for a block | 12:42 |
@HeikoS | then we wouldnt have to do this ultra messy pthread stuff anymore | 12:43 |
@HeikoS | pure horror that thing (at least to me :) | 12:43 |
@lisitsyn | HeikoS: well jobs may be pthread powered still | 12:43 |
@HeikoS | lisitsyn: yes, but with a central implementation | 12:43 |
@lisitsyn | yes | 12:43 |
@HeikoS | rather than 100 local once | 12:43 |
@HeikoS | ones | 12:43 |
@lisitsyn | yes these pthread runners allthrough the code are scary | 12:44 |
@lisitsyn | HeikoS: I am quite interested in developing it actually | 12:46 |
@HeikoS | lisitsyn: very welcome to do this | 12:47 |
@lisitsyn | HeikoS: yeah lets see if I get enough time to join you | 12:48 |
@HeikoS | lisitsyn: we dont have time pressure on this | 12:48 |
@HeikoS | the log-det will be written in that framework anyway | 12:49 |
@HeikoS | engine implementation will be sequential at first | 12:49 |
@HeikoS | and if someone has time, we can do cooler things | 12:49 |
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout] | 12:49 | |
@lisitsyn | HeikoS: btw are you done with your nips thing? | 12:50 |
@HeikoS | working on it | 12:50 |
lambday | back | 12:51 |
lambday | HeikoS: task looked better than job :( | 12:51 |
@HeikoS | why? | 12:51 |
lambday | job is fine though.. | 12:52 |
lambday | changed it.. added that method | 12:52 |
lambday | mailing to both of you | 12:52 |
@lisitsyn | except to blowjob collision job sounds better | 12:52 |
@lisitsyn | hah | 12:52 |
lambday | lisitsyn: I was just about to say - let's hope we don't blow this job : | 12:53 |
lambday | :D | 12:53 |
lambday | lisitsyn: I don't have your email id | 12:53 |
lambday | :( | 12:53 |
@lisitsyn | lambday: lisitsyn.s.o@gmail.com | 12:53 |
lambday | lisitsyn: ty | 12:55 |
@HeikoS | lambday: looks good, maybe even call aggregate results "reduce_results" or something | 12:58 |
lambday | HeikoS: umm... yes that's more generic name | 12:59 |
@HeikoS | looking good apart from that, we could now start writing code for the base classes | 12:59 |
@HeikoS | so one more thing that we have to be careful about | 12:59 |
@HeikoS | the generic linear operators and operator functions | 12:59 |
@HeikoS | I dont really know how we should do this | 13:00 |
@HeikoS | shogun does not have the type Complex | 13:00 |
@HeikoS | lisitsyn: what do you think on this? | 13:00 |
lambday | std::complex should be fine, no? | 13:00 |
@HeikoS | lambday: still | 13:01 |
@HeikoS | might be a problem | 13:01 |
lambday | HeikoS: implement complex things for shogun then? | 13:02 |
lambday | but we need complex for only a few parts here | 13:02 |
@HeikoS | yes, internally this is fine | 13:02 |
@HeikoS | but if we start having complex variables, we might get into trouble | 13:03 |
@HeikoS | so I am not sure whether to have a generic CLinearOperator | 13:03 |
@HeikoS | or CDenseMatrix | 13:03 |
@HeikoS | CComplexDenseMatrix | 13:03 |
@HeikoS | etc | 13:03 |
@HeikoS | where do the complex values appear? | 13:04 |
@HeikoS | only in the matrix of the shifted system right? | 13:04 |
lambday | in the weights and shifts only | 13:04 |
lambday | yes | 13:04 |
@HeikoS | so then we could even completely hide this | 13:04 |
@HeikoS | and say a linear operator just applies to a real vector | 13:05 |
lambday | and we'll never really use direct solver | 13:05 |
lambday | HeikoS: yes | 13:05 |
@HeikoS | so then we could have an implementation CComplexSparseMatrixOperator | 13:05 |
@HeikoS | which represents the shifted stuff | 13:06 |
@HeikoS | and CSparseMatrixOperator for the other | 13:06 |
@HeikoS | then we dont need any generic stuff anywhere | 13:06 |
lambday | umm.. okay then | 13:06 |
@HeikoS | sorry for all those changes, just remembered that I made a mental note of this | 13:07 |
lambday | but having a base linear operator class would have been helpful, no? | 13:07 |
@HeikoS | we have to try to avoid that any complex numbers are in function interfaces | 13:07 |
@HeikoS | lambday: yes, but not generic | 13:07 |
lambday | HeikoS: no problem.. I myself am forgetting so many things | 13:07 |
lambday | HeikoS: only for real, you mean | 13:07 |
lambday | no complex stuff | 13:08 |
@HeikoS | lambday: ah sorry | 13:08 |
@HeikoS | let me think | 13:08 |
@HeikoS | so the abstract base might be generic | 13:08 |
@HeikoS | float64, float32 etc | 13:08 |
@HeikoS | not complex though | 13:08 |
lambday | cool | 13:09 |
@HeikoS | and the sublcass fixes T=float64_t | 13:09 |
lambday | float32_t would be used later? | 13:09 |
lambday | yes... okay | 13:09 |
@HeikoS | I dont know | 13:09 |
@HeikoS | maybe leave it as it is | 13:10 |
@HeikoS | so CLinearOperator<T> | 13:10 |
@HeikoS | CoperatorFunction<T> | 13:10 |
-!- lambday [67157e4e@gateway/web/cgi-irc/kiwiirc.com/ip.103.21.126.78] has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client] | 13:11 | |
@HeikoS | and then CRationalApproximation<T=float64_t> | 13:11 |
@HeikoS | lisitsyn: what do you think? | 13:11 |
@HeikoS | should these things be made available for types other that float64? | 13:11 |
@HeikoS | we dont need others | 13:11 |
-!- lambday [67157e4e@gateway/web/cgi-irc/kiwiirc.com/ip.103.21.126.78] has joined #shogun | 13:13 | |
lambday | HeikoS: sorry I got dc again :( my net connection sucks! | 13:13 |
@HeikoS | lambday: see mail, leave the abstract bases generic it should be fine | 13:13 |
@HeikoS | lambday: just one more sublety, we should not use SGVector<Complex> | 13:13 |
@HeikoS | rather have to SGVector<float64_t> for real and imaginary parts | 13:14 |
lambday | so, array then? | 13:14 |
lambday | okay | 13:14 |
lambday | but that makes it bad to look at :-/ | 13:14 |
@HeikoS | the Complex type is not supported by shogun, so nothing that can be seen from outside a class should have this type | 13:14 |
lambday | HeikoS: alright... yes | 13:14 |
@HeikoS | and linear operators you use all have T=float64_T | 13:14 |
@HeikoS | because this is the type of the vector it is applied to | 13:15 |
lambday | implementations? | 13:15 |
@HeikoS | ah shit | 13:15 |
@HeikoS | so in the shifted system | 13:15 |
@HeikoS | we compute the result of (C + i)^(-1) v and then use the imaginary part only or something right? | 13:15 |
lambday | HeikoS: yes | 13:16 |
lambday | after computing the system, only the imaginary part | 13:16 |
lambday | I mean, solving systems | 13:16 |
@HeikoS | so actually, the result of the linear operator applied to a vector is complex | 13:16 |
lambday | HeikoS: yep, but we'll take the imaginary part of that | 13:16 |
@HeikoS | this is bad | 13:17 |
lambday | :( | 13:17 |
@HeikoS | we run in all sorts of issues with that | 13:17 |
lambday | like? | 13:17 |
@HeikoS | see, first we set T=float64_t in CRationalApprox | 13:17 |
lambday | yes | 13:18 |
@HeikoS | so the type of the base classes will also be like this | 13:18 |
lambday | yes | 13:18 |
@HeikoS | which is fine | 13:18 |
@HeikoS | since its a real matrix | 13:18 |
@HeikoS | ok | 13:18 |
lambday | hmm | 13:18 |
@HeikoS | sorry, this works :) | 13:18 |
@HeikoS | next thing is | 13:18 |
@HeikoS | when the jobs are created | 13:18 |
@HeikoS | we create LinearOperators which have to return complex value | 13:19 |
@HeikoS | s | 13:19 |
lambday | yes | 13:19 |
@HeikoS | and we cannot hide this away | 13:19 |
lambday | cocg_m takes care of that | 13:19 |
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has joined #shogun | 13:19 | |
@HeikoS | but then is has a field m_operator where T=Complex | 13:19 |
@HeikoS | which we cannot do | 13:19 |
lambday | we don't have to make operator complex | 13:20 |
@HeikoS | if it is real valued though | 13:20 |
@HeikoS | apply() will return a real valued vector | 13:20 |
@HeikoS | am I correct? | 13:22 |
lambday | HeikoS: that's fine, I guess? | 13:23 |
lambday | https://github.com/lambday/KRYLSTAT/blob/master/cocg_m/eigen/eigen_cocg_m.h#L234 | 13:23 |
@HeikoS | whats that? | 13:24 |
lambday | linear operator applies on real, returns real vector.. in cocg_m | 13:25 |
@HeikoS | ah thats in COCG_M | 13:25 |
lambday | that's where we need this apply(), no? | 13:26 |
@HeikoS | so this solves a real system and then internally loops over the complex shifts | 13:26 |
lambday | hmm | 13:26 |
@HeikoS | yes, should be fine in COCG_M then | 13:27 |
@HeikoS | but what about solving all systems seperately | 13:27 |
lambday | HeikoS: that's problematic :-/ | 13:27 |
lambday | will we ever use a complex direct solver? that was just for testing purpose, right? | 13:28 |
lambday | testing whether the framework works or not | 13:28 |
lambday | confused :( | 13:28 |
@HeikoS | lambday yes, however, a) we still want this and b) also for cg, the matrix you multiply is compelx | 13:28 |
@HeikoS | so you get a complex result | 13:29 |
@HeikoS | lambday: btw complex direct solvers are no problem | 13:29 |
lambday | hmmm | 13:29 |
@HeikoS | the problem is how to store the matrices without adding a new type to shogun | 13:29 |
@HeikoS | we can currnetly only have non-complex types | 13:30 |
@HeikoS | so the operator applied to a vector cannot return somthing complex | 13:30 |
@HeikoS | I see two possible solutions: | 13:30 |
@HeikoS | 1) add complex type to shogun (extremely messy but possible) | 13:31 |
@HeikoS | 2) make the complex linear operator operator one vectors where the first half is for real and the second half is for imag. Internally, we can then construct the complex systems. Even more messy | 13:31 |
@HeikoS | sonne|work, lisitsyn what do you think about adding a new type complex_t to shogun? | 13:32 |
@HeikoS | how much work would that be? | 13:32 |
@HeikoS | typemaps, serialisation, lots of lots of switch statements, what else? | 13:32 |
sonne|work | the typedef is just 3 lines. supporting it natively for typemaps etc means writing it for each interface | 13:33 |
sonne|work | all doable | 13:33 |
@HeikoS | sonne|work: and feasible? did you ever consider complex types? | 13:33 |
@HeikoS | lambday: maybe we should do this first then | 13:34 |
sonne|work | HeikoS: no never needed it but it is not hard to do. however problem is that we might want to limit ourselves to float64_t for complex | 13:34 |
sonne|work | to not have even more types | 13:34 |
lambday | adding complex... hmm | 13:35 |
@HeikoS | sonne|work: yes I would also go for complex64_t | 13:35 |
@HeikoS | or complex128_t ? I dont know how I would call it | 13:35 |
lambday | HeikoS: by the way, only the complex shifts makes the system complex, right? those are taken care separately in COCG_M, no? apply only applies the real matrix part - or may be I misunderstood | 13:36 |
lambday | :( | 13:36 |
sonne|work | HeikoS: 64_t | 13:36 |
@HeikoS | lambday: you are right | 13:36 |
@HeikoS | for COCG_M it is fine | 13:36 |
@HeikoS | but what if we want to solve all systems seperately? | 13:37 |
lambday | HeikoS: got your point | 13:37 |
@HeikoS | then we need a cg solver for a complex system | 13:37 |
lambday | yes | 13:37 |
@HeikoS | lambday: but since I would like to see this included (one can precondition each system seperately and gain speed) | 13:37 |
@HeikoS | we have to think about it | 13:37 |
@HeikoS | lambday: but lets start with the COCG_M | 13:38 |
@HeikoS | the other one only makes sense once we have a cluster implementation | 13:38 |
lambday | HeikoS: what have we decided regarding the complex then? | 13:38 |
lambday | add to shogun, or use separate vectors for real and imag | 13:38 |
@HeikoS | lambday: haha, sorry for all the confusion :) | 13:38 |
@HeikoS | so we will add complex64_t to shogun | 13:38 |
@HeikoS | but later | 13:39 |
@HeikoS | once COCG_M works, we can work on this | 13:39 |
lambday | HeikoS: no problem :D | 13:39 |
lambday | HeikoS: no I mean, even for COCG_M, we need shift vectors | 13:39 |
lambday | which are complex | 13:39 |
@HeikoS | lambday, right | 13:39 |
@HeikoS | hmm | 13:39 |
@HeikoS | so: either two float64_t vectors and you construct the complex numbers internally | 13:40 |
@HeikoS | or add the complex64_t now | 13:40 |
@HeikoS | if we go the first way, we have to change it later | 13:40 |
@HeikoS | but it might be easier for now | 13:40 |
lambday | now is better :D cause handling all real and imag in arithmatic operation will be deadly :( | 13:40 |
@HeikoS | lambday: you can just construct a vector of std::complex internally | 13:41 |
@HeikoS | just hide it to the outside world | 13:41 |
@HeikoS | so no problem | 13:41 |
@HeikoS | and easy to change later | 13:41 |
lambday | like the one I did in elliptic curve functions? or you mean, no typedef at all | 13:41 |
lambday | oh that's not for vector | 13:42 |
lambday | here we need vectors - sorry :( | 13:42 |
@lisitsyn | HeikoS: uh a lot | 13:42 |
@HeikoS | lambday: this needs to be changed later | 13:43 |
@HeikoS | as said, we cannot have Real or Complex in any interface, only shogun types | 13:43 |
@HeikoS | otherwise we need to hide those to the modular/static interfaces | 13:43 |
@HeikoS | ah they are high precision anyway | 13:44 |
@HeikoS | so we need to hide them | 13:44 |
@HeikoS | they can stay as they are | 13:44 |
@HeikoS | lambday: okay, lets add the complex64_t now. Then this discussion is over | 13:46 |
lambday | HeikoS: okay | 13:46 |
lambday | then we can have SGVector<complex64_t> | 13:46 |
@HeikoS | yep | 13:46 |
lambday | cool | 13:46 |
lambday | :) | 13:46 |
@lisitsyn | HeikoS: recalling porn things your project is really deep penetration | 13:47 |
@HeikoS | lisitsyn: good stuff :) | 13:47 |
@HeikoS | lambday: have a look into the type stuff | 13:47 |
@HeikoS | its a bit messy but should be possible | 13:47 |
@lisitsyn | HeikoS: I didn't follow your conversation sorry - are there unanswered questions? | 13:47 |
lambday | HeikoS: okay, I'm checking | 13:47 |
@HeikoS | lisitsyn: solution to all of them: we will add complex64_t to shogun, | 13:48 |
lambday | where do we define this types in shogun? | 13:51 |
-!- vgorbati [~anonymous@77.47.132.39] has joined #shogun | 13:52 | |
@HeikoS | lambday: common.h | 13:52 |
lambday | so, we need to add typedef std::complex<float64_t> complex64_t | 13:53 |
@HeikoS | lambday: and DataType.h | 13:53 |
@HeikoS | lambday: yes first thing | 13:54 |
@HeikoS | then the new SG primitive tye | 13:54 |
@HeikoS | which might cause problems | 13:54 |
@HeikoS | add to your todo list: serialization, modular typemaps, check all switch over ptype | 13:55 |
lambday | I need to understand how the existing ones work first :( | 13:55 |
@HeikoS | lambday: yes, take your time | 13:55 |
@HeikoS | so we have this shogun type object | 13:55 |
@HeikoS | TSGDataType | 13:56 |
lambday | hmm | 13:56 |
lambday | add it to primitive type, right? | 13:56 |
@HeikoS | contains primitive type, container type(vector,matrix), struct type (string,sparse) | 13:56 |
@HeikoS | yes | 13:56 |
lambday | hmmm | 13:56 |
@HeikoS | TSGDataType is used by shogun parameter framework to represent things | 13:57 |
@HeikoS | for example for serialisation, model-selection, equals method | 13:57 |
@HeikoS | one has to register parameters with the SG_ADD macro | 13:57 |
@HeikoS | which calls m_parameters->add of the current object | 13:57 |
@HeikoS | in there, we need new methods for complex | 13:57 |
@HeikoS | should be a few | 13:58 |
@HeikoS | in the same style as the existing ones | 13:58 |
@HeikoS | then we have to check serialisation | 13:58 |
@HeikoS | which writes types to files | 13:58 |
@HeikoS | then modular maps which map the type the the type of the language (eg python complex) | 13:59 |
@HeikoS | lambday: browse through the code a bit, take your time | 13:59 |
lambday | HeikoS: yes... and by new methods, we'll be needing all the operators, right? | 13:59 |
@HeikoS | I will have lunch now and have to do some work afterwards, but will be around during the day, so ask if anything uclear | 13:59 |
lambday | overloaded | 13:59 |
@HeikoS | lambday: what do you mean by new operator? | 13:59 |
-!- vgorbati [~anonymous@77.47.132.39] has quit [Ping timeout: 256 seconds] | 14:00 | |
@HeikoS | + - * += etc? | 14:00 |
lambday | yes yes | 14:00 |
@HeikoS | we get that from std::complex dont we? | 14:00 |
lambday | yes | 14:00 |
lambday | okay so nothing to do for that in shogun | 14:00 |
@HeikoS | no thats all fine | 14:00 |
@HeikoS | its very easy to add the type | 14:01 |
@HeikoS | a bit more work to support it properly :) | 14:01 |
@HeikoS | man, I realise that I type way too much :) | 14:01 |
@HeikoS | so will get lunch now, see you in a bit | 14:01 |
lambday | HeikoS: hmm.. I'm checking.. :) | 14:02 |
lambday | HeikoS: lol I do that too | 14:02 |
lambday | not here though cause most of the time I'm confused :D | 14:02 |
lambday | yes have lunch | 14:02 |
lambday | see you :) | 14:02 |
lambday | i too am going for some tea | 14:03 |
lambday | be right back | 14:03 |
@HeikoS | chai :) | 14:03 |
lambday | HeikoS: yes garam chai :D | 14:05 |
lambday | garam=hot | 14:05 |
lambday | ciao :) | 14:05 |
-!- nube [~rho@49.244.111.24] has joined #shogun | 14:10 | |
-!- iglesiasg [c1934d16@gateway/web/freenode/ip.193.147.77.22] has quit [Ping timeout: 250 seconds] | 14:12 | |
-!- iglesiasg [c1934d16@gateway/web/freenode/ip.193.147.77.22] has joined #shogun | 14:13 | |
-!- mode/#shogun [+o iglesiasg] by ChanServ | 14:13 | |
-!- van51 [~van51@195.251.235.71] has quit [Quit: Leaving.] | 14:46 | |
@HeikoS | lambday: back | 14:55 |
-!- foulwall [~foulwall@li379-21.members.linode.com] has quit [Ping timeout: 256 seconds] | 15:11 | |
-!- foulwall_ [~foulwall@2001:da8:215:503:4907:15c7:758:7cf5] has joined #shogun | 15:17 | |
-!- foulwall_ [~foulwall@2001:da8:215:503:4907:15c7:758:7cf5] has quit [Remote host closed the connection] | 15:17 | |
-!- foulwall [~foulwall@li379-21.members.linode.com] has joined #shogun | 15:20 | |
-!- iglesiasg [c1934d16@gateway/web/freenode/ip.193.147.77.22] has quit [Quit: Page closed] | 15:24 | |
-!- van51 [~van51@athedsl-320452.home.otenet.gr] has joined #shogun | 15:25 | |
-!- iglesiasg [c1934d16@gateway/web/freenode/ip.193.147.77.22] has joined #shogun | 15:53 | |
-!- mode/#shogun [+o iglesiasg] by ChanServ | 15:53 | |
-!- zxtx [~zv@ool-457e751d.dyn.optonline.net] has quit [Ping timeout: 248 seconds] | 15:56 | |
-!- zxtx [~zv@ool-457e751d.dyn.optonline.net] has joined #shogun | 15:59 | |
-!- goldbug [~goldbug@global-2-1.nat.csx.cam.ac.uk] has joined #shogun | 16:07 | |
lambday | HeikoS: back | 16:15 |
-!- goldbug [~goldbug@global-2-1.nat.csx.cam.ac.uk] has quit [Ping timeout: 240 seconds] | 16:25 | |
-!- van51 [~van51@athedsl-320452.home.otenet.gr] has quit [Quit: Leaving.] | 16:30 | |
-!- gsomix [~gsomix@37.61.179.240] has quit [Quit: Leaving] | 16:32 | |
-!- van51 [~van51@athedsl-320452.home.otenet.gr] has joined #shogun | 16:36 | |
-!- iglesiasg [c1934d16@gateway/web/freenode/ip.193.147.77.22] has quit [Ping timeout: 250 seconds] | 16:48 | |
-!- iglesiasg [c1934d16@gateway/web/freenode/ip.193.147.77.22] has joined #shogun | 16:50 | |
-!- foulwall [~foulwall@li379-21.members.linode.com] has quit [Remote host closed the connection] | 16:52 | |
@sonney2k | van51, around? | 17:08 |
van51 | sonney2k: yea | 17:09 |
van51 | but on a skype call atm | 17:09 |
@sonney2k | van51, ahh k | 17:10 |
@sonney2k | van51, once you are free please say so :) | 17:10 |
@sonney2k | shogun-buildbot, force build nightly_default | 17:11 |
shogun-buildbot | build forced [ETA 53m10s] | 17:11 |
shogun-buildbot | I'll give a shout when the build finishes | 17:11 |
@sonney2k | hushell, have a look at http://shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/1170/steps/compile%20python_modular/logs/warnings%20%2817%29 | 17:12 |
-!- foulwall [~foulwall@2001:da8:215:c252:addf:5916:dba8:873b] has joined #shogun | 17:12 | |
@sonney2k | hushell, there still are some warnings in MulticlassOneVsRestStrategy that you might want to fix | 17:12 |
@sonney2k | foulwall, so how did it go? superhero? | 17:12 |
pickle27 | sonney2k: I saw that the test failed thats the same result I got on my system, for some reason there are 2 differences in the classification with the Eigen3 QDA | 17:13 |
pickle27 | it passes the synthetic test though | 17:13 |
foulwall | sonney2k: not bad~ | 17:14 |
@sonney2k | pickle27, which tests fail? | 17:17 |
@sonney2k | foulwall, congrats then! | 17:17 |
foulwall | sonney2k: thanks, I can do some work tonight~~ | 17:18 |
@sonney2k | foulwall, when do you have your next exam? | 17:18 |
pickle27 | the python modular one on the fm_train_real.dat has a slightly different output | 17:18 |
pickle27 | sonney2k: but the unit test I wrote passes | 17:18 |
@sonney2k | pickle27, compared to? | 17:18 |
foulwall | sonney2k: 1 month later, 3 exams left | 17:19 |
pickle27 | sonney2k: the unit test is synthetic perfectly separated data and it is tested on the training input | 17:19 |
@sonney2k | pickle27, I don't see anything that fails https://travis-ci.org/shogun-toolbox/shogun | 17:19 |
@sonney2k | http://shogun-toolbox.org/buildbot/waterfall | 17:19 |
pickle27 | the python example is on real data and doesn't produce 100 percent results the old and new version just make 2 different mistakes | 17:20 |
@sonney2k | foulwall, ok nice! | 17:20 |
pickle27 | it was on the build bot | 17:20 |
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun | 17:20 | |
shogun-notifier- | shogun: Soeren Sonnenburg :develop * 819f7de / src/shogun/mathematics/Random.h: https://github.com/shogun-toolbox/shogun/commit/819f7de7bb608a8d057e9ac6a6567cba6a392066 | 17:20 |
shogun-notifier- | shogun: add some documentation to the random() functions | 17:20 |
pickle27 | sonney2k: http://shogun-toolbox.org/buildbot/builders/nightly_default/builds/408/steps/test/logs/stdio | 17:21 |
@sonney2k | pickle27, ohh but this is from last night where everything was still kaboom | 17:21 |
@sonney2k | pickle27, the current one is the python_modular one and it is happy | 17:21 |
pickle27 | sonney2k: huh okay great | 17:22 |
@sonney2k | pickle27, so just do you performance tests - and let the buildbot do its work :D | 17:22 |
pickle27 | sonney2k: right you wanted a benchmark | 17:22 |
@sonney2k | foulwall, what I consider most important is to get the structure of the demo's right | 17:23 |
pickle27 | sonney2k: you want between lapack and the new version? | 17:23 |
@sonney2k | foulwall, I mean general enough such that we don't have much code duplication | 17:23 |
@sonney2k | pickle27, yeah and in general | 17:23 |
pickle27 | sonney2k: okay! | 17:23 |
foulwall | sonney2k: I'm also worrying about that | 17:23 |
@sonney2k | pickle27, I think we should provide benchmarks for each and every example in shogun but hey | 17:24 |
pickle27 | sonney2k: yeah I'll put the results in a comment block | 17:24 |
@sonney2k | foulwall, so for data it is pretty clear: it can come from a number of sources like artificially generated (datagenerator class, could be loaded from a file or input by the user via just clicks) | 17:24 |
foulwall | sonney2k: I'll have a try? seperate the demos into modules and reuse them | 17:24 |
@sonney2k | then we have the coordinate system for drawing points - that is shared for everything | 17:25 |
@sonney2k | and then the task specific machines, e.g. regression machines, binary classifiers, multiclass classifiers | 17:26 |
@sonney2k | clustering (each one is rather specific...) | 17:26 |
@sonney2k | then dimension reduction (aka converters - we can use lisitsyn's example there) | 17:26 |
van51 | sonney2k: ok, I'm free now | 17:27 |
@sonney2k | and then some illustrations for kernels, distances, loss, preprocessors, evaluations | 17:27 |
@sonney2k | foulwall, if we get the structure right adding one more method is like writing 5-10 lines | 17:27 |
foulwall | sonney2k: ok~~ I'll rewrite the structures for reuse. | 17:27 |
@sonney2k | foulwall, or make a plan and discuss it before doing the actual work | 17:28 |
@sonney2k | van51, we have to talk later I have to leave the train now | 17:28 |
@sonney2k | hope you will be around in the evening | 17:28 |
van51 | sonney2k: I'll be around for 2-3 more hours for sure | 17:29 |
van51 | and then by night again | 17:29 |
foulwall | sonney2k: so just give me some time and let me come up with a plan. | 17:29 |
foulwall | sonney2k: the exams are easy to deal, I'll start work for 6 hours/day from tomorrow if i'm accepted. | 17:31 |
shogun-buildbot | build #878 of cyg1 - libshogun is complete: Failure [failed configure] Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/878 blamelist: Soeren Sonnenburg <sonne@debian.org> | 17:36 |
-!- iglesiasg [c1934d16@gateway/web/freenode/ip.193.147.77.22] has quit [Ping timeout: 250 seconds] | 17:38 | |
-!- iglesiasg [c1934d16@gateway/web/freenode/ip.193.147.77.22] has joined #shogun | 17:41 | |
-!- mode/#shogun [+o iglesiasg] by ChanServ | 17:41 | |
shogun-buildbot | build #409 of nightly_default is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/nightly_default/builds/409 | 17:46 |
-!- lambday [67157e4e@gateway/web/cgi-irc/kiwiirc.com/ip.103.21.126.78] has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client] | 18:12 | |
-!- goldbug [~goldbug@global-1-26.nat.csx.cam.ac.uk] has joined #shogun | 18:22 | |
-!- goldbug [~goldbug@global-1-26.nat.csx.cam.ac.uk] has quit [Client Quit] | 18:25 | |
-!- goldbug [~goldbug@global-1-26.nat.csx.cam.ac.uk] has joined #shogun | 18:25 | |
-!- lisitsyn [~lisitsin@mxs.kg.ru] has left #shogun [] | 18:25 | |
-!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has joined #shogun | 18:27 | |
-!- goldbug [~goldbug@global-1-26.nat.csx.cam.ac.uk] has quit [Quit: Leaving] | 18:34 | |
-!- goldbug [~goldbug@global-1-26.nat.csx.cam.ac.uk] has joined #shogun | 18:34 | |
-!- goldbug [~goldbug@global-1-26.nat.csx.cam.ac.uk] has quit [Client Quit] | 18:37 | |
-!- goldbug [~goldbug@global-1-26.nat.csx.cam.ac.uk] has joined #shogun | 18:37 | |
-!- vgorbati_ [d4029f22@gateway/web/freenode/ip.212.2.159.34] has joined #shogun | 19:28 | |
-!- lisitsyn [~blackburn@188.168.13.232] has joined #shogun | 19:30 | |
-!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has quit [Ping timeout: 250 seconds] | 19:30 | |
lisitsyn | ahaha forgot *both* my room keys and notebook AC adapter | 19:30 |
-!- vgorbati_ is now known as vgorbati | 19:31 | |
lisitsyn | I am probably with you gsocers tonight | 19:31 |
lisitsyn | not* | 19:31 |
lisitsyn | not with you :D | 19:31 |
-!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has quit [Ping timeout: 250 seconds] | 19:46 | |
@sonney2k | lisitsyn, poor guy! | 19:54 |
lisitsyn | sonney2k: yes I got a bad luck | 19:54 |
@sonney2k | I am probably too dead to stay awake long :/ | 19:54 |
lisitsyn | sonney2k: in an hour | 19:54 |
@sonney2k | accumulated lack of sleep | 19:54 |
@sonney2k | no idea if I survive that long | 19:55 |
lisitsyn | my battery won't survive for sure | 19:55 |
@sonney2k | HeikoS, around? | 19:55 |
@HeikoS | sonney2k: yes | 19:55 |
@sonney2k | HeikoS, it seems to be more problematic to organize rooms for the hands-on workshop at TU | 19:56 |
@HeikoS | okay? | 19:56 |
@HeikoS | TUB is so big, how can that be? | 19:56 |
@sonney2k | maybe we should try to sit at c-base | 19:56 |
lisitsyn | lets sit at tiergarten! | 19:57 |
lisitsyn | with vodka! | 19:57 |
@sonney2k | HeikoS, it is lots of stuff like they want money, each participant by name | 19:57 |
@sonney2k | etc etc | 19:57 |
@HeikoS | annoying | 19:57 |
@HeikoS | and now? | 19:57 |
@HeikoS | c-base? | 19:57 |
@HeikoS | did you ask them? | 19:57 |
lisitsyn | tiergarten! | 19:57 |
lisitsyn | :D | 19:57 |
@HeikoS | haha :D :) | 19:57 |
lisitsyn | I've never been to tiergarten | 19:58 |
lisitsyn | is it nice? | 19:58 |
lisitsyn | alright I've never been ANYWHER | 19:58 |
lisitsyn | :D | 19:58 |
@HeikoS | Germany? Europe? NotRussia(tm)? :D | 20:01 |
@sonney2k | HeikoS, no I am planning to though. I guess the group will be downsized quite a bit anyway on the post-official workshop day | 20:01 |
lisitsyn | HeikoS: never been anywhere abroad | 20:01 |
@HeikoS | sonney2k: in that case, should be fine | 20:01 |
@HeikoS | if cbase agrees | 20:01 |
@iglesiasg | how many applicants for the workshop this far? | 20:02 |
@sonney2k | I should have done the shogun & gsoc thing back when I was in academia. then one could have done the post conference sprints | 20:02 |
@HeikoS | well thats life :) | 20:03 |
@sonney2k | HeikoS, there are other suitable spots in berlin - star bucks in the worst case :D | 20:03 |
@HeikoS | haha :d please not | 20:05 |
@sonney2k | I will ask around | 20:05 |
@sonney2k | HeikoS, I mean places like http://hackerspaces.org/wiki/Berlin | 20:06 |
@sonney2k | Raumfahrtagentur would be something fancy too :D | 20:06 |
@HeikoS | yes, that sounds cool | 20:06 |
-!- iglesiasg [c1934d16@gateway/web/freenode/ip.193.147.77.22] has quit [Quit: Page closed] | 20:07 | |
-!- van51 [~van51@athedsl-320452.home.otenet.gr] has quit [Quit: Leaving.] | 20:07 | |
@sonney2k | HeikoS, the big issue really is that I am away on that day so that makes it difficult | 20:07 |
@HeikoS | I see | 20:08 |
@HeikoS | so we should set beforehand right | 20:08 |
lisitsyn | batterryyyyy dyyyying | 20:08 |
@HeikoS | then ask cbase first and then we will see | 20:08 |
@HeikoS | lisitsyn: bye :) | 20:08 |
lisitsyn | I am alive yeet | 20:08 |
lisitsyn | noooo | 20:08 |
-!- lisitsyn [~blackburn@188.168.13.232] has quit [Quit: Leaving.] | 20:09 | |
@sonney2k | HeikoS, we should really think about spending some money on gsoc'ers maybe some can join | 20:10 |
@HeikoS | the new ones? | 20:10 |
@HeikoS | would be cool | 20:10 |
@HeikoS | but might be expensive | 20:10 |
@HeikoS | for the non-europeans | 20:10 |
@sonney2k | yeah :/ | 20:10 |
-!- goldbug [~goldbug@global-1-26.nat.csx.cam.ac.uk] has quit [Ping timeout: 245 seconds] | 20:14 | |
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout] | 20:20 | |
-!- deerishi [c649b206@gateway/web/freenode/ip.198.73.178.6] has joined #shogun | 20:24 | |
-!- lambday [67157e4f@gateway/web/freenode/ip.103.21.126.79] has joined #shogun | 20:25 | |
lambday | HeikoS: hi... | 20:25 |
@sonney2k | ohh man I really have to sleep now everyone relax. those of you that get accepted for gsoc - congrats. for those of you that didn't get in this year. don't take it personally we have 8 slots (which is a lot in gsoc terms) but had >50 students applying. if you are still interested in working with us we certainly would be more than happy to work with you too. | 20:39 |
@sonney2k | and if you want to know more about our decision process (why you did / did not get in) ask me personally | 20:40 |
-!- gsomix [~gsomix@185.6.239.144] has joined #shogun | 20:47 | |
gsomix | good evening | 20:47 |
hushell | sonney2k: I should check these warning msg myself :( I'll fix it quickly | 20:56 |
-!- iglesiasg [d58f32c1@gateway/web/freenode/ip.213.143.50.193] has joined #shogun | 21:17 | |
-!- mode/#shogun [+o iglesiasg] by ChanServ | 21:19 | |
@iglesiasg | congrats to everyone accepted! | 21:19 |
gsomix | sonney2k, sonne|work hey. just got mail. thanks! | 21:19 |
hushell | cong!!! guys | 21:20 |
-!- gsomix is now known as lisitsyn | 21:22 | |
lisitsyn | iglesiasg, hushell, votjakovr, lambday, foulwall, pickle27 congrats | 21:22 |
@iglesiasg | thanks lisitsyn! :) | 21:23 |
pickle27 | lisitsyn: thanks! | 21:23 |
foulwall | thanks lisitsyn | 21:23 |
votjakovr | lisitsyn: thanks :) | 21:23 |
lisitsyn | you all have done a great work pre-season | 21:23 |
hushell | thank you! lisitsyn | 21:23 |
lisitsyn | iglesiasg, hushell, votjakovr, lambday, foulwall, pickle27 let us have a meeting later - we will let you know | 21:25 |
@iglesiasg | sure, nice idea | 21:25 |
lisitsyn | if you are curious who is the 8th one it is van51 | 21:26 |
lisitsyn | I am sorry to have no student this year | 21:26 |
lisitsyn | but that's how we vote | 21:26 |
pickle27 | thats awesome, I'mm really looking forward to it! | 21:26 |
@iglesiasg | lisitsyn: I am sorry too :( | 21:27 |
pickle27 | lisitsyn: aren't I working with you? | 21:27 |
@iglesiasg | lisitsyn: but then you will have more time to co-mentor me :) | 21:27 |
lisitsyn | pickle27, yes sure I'll will review your code | 21:27 |
@iglesiasg | and the others too | 21:27 |
lisitsyn | pickle27, but I would have to learn BSS just like you :) | 21:28 |
pickle27 | lisitsyn: ah right you were co-mentor or something for me too | 21:28 |
lisitsyn | I know about dimension reduction a bit more | 21:28 |
pickle27 | lisitsyn: me too! | 21:28 |
lisitsyn | we really hope that the most of you are here to stay | 21:29 |
lisitsyn | that's really important for us to have guys like iglesiasg | 21:29 |
pickle27 | lisitsyn: yeah I deffs hope to stick around! | 21:29 |
lisitsyn | anyway we are going to have a dense summer | 21:30 |
lisitsyn | we will schedule a meeting one of next days | 21:30 |
hushell | really happy to working with you guys! | 21:30 |
@iglesiasg | definetely, lot of new and cool stuff coming up | 21:31 |
pickle27 | yeah let me know about the meeting! | 21:31 |
-!- deerishi [c649b206@gateway/web/freenode/ip.198.73.178.6] has quit [Quit: Page closed] | 21:31 | |
pickle27 | I am going to be at a conference for a few days so email might be more reliable but I'll try and be on IRC too | 21:31 |
lisitsyn | pickle27, yeah that | 21:32 |
lisitsyn | that's for sure | 21:32 |
lisitsyn | guys about some rules | 21:32 |
lisitsyn | once we start | 21:32 |
lisitsyn | you will have to send a weekly report | 21:32 |
lisitsyn | say monday or so | 21:32 |
lisitsyn | I'd like to note that we have to make it mandatory | 21:33 |
lisitsyn | there you would have just to say what you have done last week and what do you plan to do next week | 21:33 |
lisitsyn | just to clarify - if you had some issues and had no time - it is still muuuch better to send a report about that | 21:34 |
lisitsyn | iglesiasg, hushell, votjakovr, lambday, foulwall, pickle27 alright? :) | 21:34 |
@iglesiasg | lisitsyn: of course | 21:34 |
foulwall | lisitsyn: gotcha | 21:35 |
hushell | sure, that's the way I'd like to schedule myself | 21:35 |
lisitsyn | we will have a few meetings through the summer | 21:35 |
lisitsyn | near milestones | 21:35 |
lisitsyn | so we will have one in next days | 21:36 |
lisitsyn | then once we started | 21:36 |
lisitsyn | then mid-term | 21:36 |
lisitsyn | then final meeting | 21:36 |
foulwall | lisitsyn: ok | 21:36 |
lisitsyn | there we can discuss what we are heading to | 21:36 |
lisitsyn | iglesiasg, hushell, votjakovr, lambday, foulwall, pickle27 do you guys have any questions? | 21:36 |
foulwall | lisitsyn: ok | 21:36 |
hushell | I am good | 21:37 |
@iglesiasg | lisitsyn: yep, are we going to stick to IRC for these meetings this year or use something more interactive like google hangout? | 21:37 |
pickle27 | yeah that sounds good! | 21:37 |
lisitsyn | iglesiasg, I don | 21:37 |
lisitsyn | t know | 21:37 |
lisitsyn | what others think | 21:37 |
@iglesiasg | ok | 21:37 |
pickle27 | might be cool to do one hangout but IRC is probably easier for regular meetings | 21:37 |
@iglesiasg | let's see what sonne|work, sonney2k thinks too | 21:38 |
@iglesiasg | I agree with pickle27 | 21:38 |
lisitsyn | yeah he will have some opinion on that for sure | 21:38 |
lisitsyn | alright | 21:38 |
hushell | let everybody talk also a great idea | 21:38 |
lisitsyn | will pass gsomix's notebook to gsomix back then :) | 21:38 |
lisitsyn | congrats again guys | 21:39 |
-!- lisitsyn is now known as gsomix | 21:39 | |
gsomix | spirit of lisitsyn. huh | 21:40 |
gsomix | so, congrats too | 21:40 |
pickle27 | thanks! | 21:41 |
foulwall | go sleeping now, it's 3:xx am here... too late, thanks to all my friends, cu | 21:42 |
pickle27 | alright guys I am out for now too, talk to you later! | 21:42 |
-!- pickle27 [~kevin@rcv3-lab-pc.ee.queensu.ca] has quit [Quit: Leaving] | 21:42 | |
-!- goldbug [~goldbug@global-2-11.nat.csx.cam.ac.uk] has joined #shogun | 21:42 | |
-!- goldbug [~goldbug@global-2-11.nat.csx.cam.ac.uk] has quit [Client Quit] | 21:42 | |
-!- goldbug [~goldbug@global-2-11.nat.csx.cam.ac.uk] has joined #shogun | 21:43 | |
-!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has joined #shogun | 21:43 | |
hushell | good nite | 21:45 |
-!- lambday [67157e4f@gateway/web/freenode/ip.103.21.126.79] has quit [Ping timeout: 250 seconds] | 21:58 | |
-!- lambday [67157e4f@gateway/web/freenode/ip.103.21.126.79] has joined #shogun | 22:04 | |
-!- lambday [67157e4f@gateway/web/freenode/ip.103.21.126.79] has quit [Client Quit] | 22:05 | |
-!- goldbug [~goldbug@global-2-11.nat.csx.cam.ac.uk] has quit [Ping timeout: 252 seconds] | 22:33 | |
-!- lambday [67157e4f@gateway/web/freenode/ip.103.21.126.79] has joined #shogun | 22:34 | |
lambday | guys, I had an urgent sysad job in my lab :( missed the chat | 22:34 |
lambday | checking the log | 22:35 |
-!- goldbug [~goldbug@global-2-11.nat.csx.cam.ac.uk] has joined #shogun | 22:35 | |
gsomix | nite | 22:46 |
lambday | night | 22:56 |
lambday | :) | 22:56 |
-!- vgorbati [d4029f22@gateway/web/freenode/ip.212.2.159.34] has quit [Ping timeout: 250 seconds] | 23:03 | |
-!- goldbug [~goldbug@global-2-11.nat.csx.cam.ac.uk] has quit [Quit: Leaving] | 23:14 | |
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has quit [Quit: Leaving] | 23:18 | |
-!- lambday [67157e4f@gateway/web/freenode/ip.103.21.126.79] has quit [Ping timeout: 250 seconds] | 23:32 | |
--- Log closed Tue May 28 00:00:20 2013 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!