--- Log opened Fri Aug 03 00:00:02 2012 | ||
--- Day changed Fri Aug 03 2012 | ||
blackburn | n4nd0: will you be at barcelona in february? :D | 00:00 |
---|---|---|
n4nd0 | blackburn: no, I don't think so, why? | 00:00 |
n4nd0 | NIPS there? | 00:00 |
blackburn | n4nd0: no some shity conference I hope to go through | 00:00 |
n4nd0 | I think I remember that there was going to be a conference taken place there | 00:00 |
blackburn | but I have no paper actually and have only one month | 00:00 |
n4nd0 | what's the name of it? | 00:00 |
blackburn | ICPRAM | 00:01 |
@sonney2k | n4nd0, we need to talk a bit about release/stabilizing/documentation/nice examples and future involvement at that meeting | 00:01 |
blackburn | I am pretty interested who hates us and do not want to continue and who loves and want to | 00:02 |
blackburn | :) | 00:02 |
n4nd0 | haha | 00:02 |
@sonney2k | blackburn, | 00:02 |
@sonney2k | hah | 00:02 |
@sonney2k | hate love relationship :D | 00:02 |
blackburn | we mostly support binary you know | 00:03 |
blackburn | so I start to think binary too | 00:03 |
n4nd0 | blackburn: btw, what is it that should be prepared as a code sample for google? | 00:03 |
@sonney2k | n4nd0, bern4rd did some great job with the website btw... I hope I will have some more time to finish it up :D | 00:03 |
blackburn | n4nd0: ah there will be some script around | 00:04 |
@sonney2k | n4nd0, don't worry blackburn has some script for that :D | 00:04 |
blackburn | so you just pack everything you've done | 00:04 |
blackburn | would be better if they ask for blood sample | 00:04 |
blackburn | or your foot | 00:04 |
blackburn | no idea how to use code samples | 00:04 |
@sonney2k | luckily only foot | 00:04 |
@sonney2k | not finger | 00:04 |
@sonney2k | otherwise coding gets tough *eg* | 00:04 |
n4nd0 | lol | 00:05 |
blackburn | why not? we need you to send us your foot by mail to exhibit you have done something | 00:05 |
blackburn | I found that fair | 00:05 |
blackburn | sonney2k: n4nd0: did you know one gsoc student made that to london 2012? | 00:06 |
blackburn | not the foot thingy I mean | 00:06 |
blackburn | iirc he will participate in some 20k wlak | 00:06 |
blackburn | walk | 00:06 |
n4nd0 | blackburn: wow coder + professional sportsman/woman | 00:07 |
n4nd0 | nice | 00:07 |
blackburn | trevor barron is the name | 00:07 |
blackburn | moreover he has some illness - I'd say that's how one live | 00:08 |
blackburn | sonney2k: have you seen FREAK descriptor thing? | 00:09 |
@sonney2k | blackburn, yes | 00:09 |
blackburn | did you ever try? | 00:10 |
blackburn | I am going to | 00:10 |
@sonney2k | blackburn, I meant olympics... | 00:10 |
blackburn | heh | 00:10 |
@sonney2k | blackburn, FREAK - no | 00:10 |
@sonney2k | n4nd0, so you do you love or hate us :D | 00:10 |
@sonney2k | ... or do you need a real valued output :D | 00:10 |
blackburn | sonney2k: I am curious whether one can use one of string kernels to compare freaks | 00:11 |
@sonney2k | -inf ... +inf :D | 00:11 |
blackburn | cause it is binary string | 00:11 |
@sonney2k | probably - but these may not look for the right patterns... | 00:11 |
blackburn | sonney2k: well everyone uses hamming | 00:12 |
n4nd0 | sonney2k: I want to continue, sure | 00:13 |
n4nd0 | I plan to do CRFs | 00:13 |
n4nd0 | for the moment :) | 00:13 |
@sonney2k | blackburn, at least the weighted degree kernel does too | 00:14 |
blackburn | sonney2k: does what? | 00:14 |
@sonney2k | blackburn, lets send lots of love or rose ponies to n4nd0 | 00:14 |
blackburn | sonney2k: we've exchanged postcards btw | 00:14 |
n4nd0 | sonney2k: but I will need some backup support for theory aspects and such | 00:14 |
n4nd0 | I don't know if Nico or any other will be somewhat available for it? | 00:15 |
blackburn | n4nd0: best way is to collaborate on research | 00:15 |
blackburn | then he will have time for you :) | 00:15 |
n4nd0 | I understand | 00:16 |
blackburn | that's how I bother ckwidmer_ | 00:16 |
n4nd0 | I just need a bit of guidance, as long as a I am not completely alone, it will be ok | 00:17 |
blackburn | n4nd0: I did help georg zeller with some data two weeks ago btw | 00:18 |
blackburn | we are all the big family :D | 00:18 |
n4nd0 | I have really enjoyed receiving help from Georg | 00:18 |
@sonney2k | n4nd0, georg and nico continue research on that subject - so they will certainly help | 00:19 |
@sonney2k | ahh and if you get COFFIN into SO then that might even be a paper... | 00:19 |
n4nd0 | he have had a couple of dicussions on skype, he made lot of useful comments | 00:19 |
n4nd0 | sonney2k: that sounds interesting | 00:19 |
n4nd0 | blackburn: I count on you for that too, right? | 00:20 |
blackburn | I am in in everything | 00:20 |
n4nd0 | :) | 00:20 |
@sonney2k | n4nd0, http://memegenerator.net/instance/24349213?browsingOrder=New&browsingTimeSpan=Today | 00:21 |
n4nd0 | haha | 00:22 |
n4nd0 | sonney2k: thank you :) | 00:22 |
@sonney2k | blackburn, big question is how we get shogun ready for release | 00:23 |
blackburn | hardly hard | 00:23 |
@sonney2k | and I mean stable :D | 00:23 |
blackburn | recalling that m$ advertisement | 00:24 |
blackburn | we are in pretty soft state | 00:24 |
@sonney2k | blackburn, I am a hard guy you know ... | 00:25 |
blackburn | sonney2k: good for ya :) | 00:26 |
-!- cronor [~cronor@g229081189.adsl.alicedsl.de] has quit [Read error: Connection reset by peer] | 00:30 | |
-!- cronor_ [~cronor@g229081189.adsl.alicedsl.de] has joined #shogun | 00:30 | |
-!- zxtx [~zv@c-24-6-91-131.hsd1.ca.comcast.net] has joined #shogun | 00:49 | |
-!- zxtx [~zv@c-24-6-91-131.hsd1.ca.comcast.net] has quit [Ping timeout: 245 seconds] | 00:58 | |
n4nd0 | good night guys | 01:08 |
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving] | 01:09 | |
CIA-18 | shogun: Sergey Lisitsyn master * r78c48e4 / (20 files in 5 dirs): Grand multitask refactoring - https://github.com/shogun-toolbox/shogun/commit/78c48e4ec97627f93ab118deee429be4cf8cc3eb | 01:54 |
-!- blackburn [~blackburn@109.226.100.87] has quit [Quit: Leaving.] | 02:00 | |
-!- cronor [~cronor@g229081189.adsl.alicedsl.de] has joined #shogun | 02:08 | |
-!- cronor_ [~cronor@g229081189.adsl.alicedsl.de] has quit [Read error: Connection reset by peer] | 02:08 | |
-!- pluskid [6f782742@gateway/web/freenode/ip.111.120.39.66] has joined #shogun | 03:13 | |
pluskid | sonney2k: around? | 03:14 |
-!- cronor_ [~cronor@g229081189.adsl.alicedsl.de] has joined #shogun | 03:58 | |
-!- cronor [~cronor@g229081189.adsl.alicedsl.de] has quit [Read error: Connection reset by peer] | 03:58 | |
-!- cronor_ is now known as cronor | 03:58 | |
-!- cronor_ [~cronor@g229081189.adsl.alicedsl.de] has joined #shogun | 05:49 | |
-!- cronor [~cronor@g229081189.adsl.alicedsl.de] has quit [Read error: Connection reset by peer] | 05:49 | |
-!- cronor_ is now known as cronor | 05:49 | |
-!- cronor_ [~cronor@g229081189.adsl.alicedsl.de] has joined #shogun | 06:04 | |
-!- cronor [~cronor@g229081189.adsl.alicedsl.de] has quit [Read error: Connection reset by peer] | 06:04 | |
-!- cronor_ is now known as cronor | 06:04 | |
-!- pluskid [6f782742@gateway/web/freenode/ip.111.120.39.66] has quit [Quit: Page closed] | 06:05 | |
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 06:57 | |
-!- gry [gryllida@freenode/staff/gry] has joined #shogun | 07:20 | |
-!- gry [gryllida@freenode/staff/gry] has left #shogun [] | 07:20 | |
CIA-18 | shogun: Michal Uricar master * r57b2791 / (4 files): libbmrm ICP removal fixed - https://github.com/shogun-toolbox/shogun/commit/57b2791ac900e58014f99e456b7b3a6877381758 | 07:34 |
CIA-18 | shogun: Michal Uricar master * r4f7d733 / (3 files): DualLibQPBMSOSVM support for new bm solver added - https://github.com/shogun-toolbox/shogun/commit/4f7d73337f2faa6ac252a215c91a90fc5d112c87 | 07:34 |
CIA-18 | shogun: Michal Uricar master * rabdc4fb / (2 files): libppbm - Proximal Point BMRM solver - https://github.com/shogun-toolbox/shogun/commit/abdc4fb60ceb36329d1914bfb34fa15106080bd4 | 07:34 |
CIA-18 | shogun: Soeren Sonnenburg master * r19672f3 / (6 files): Merge pull request #685 from uricamic/BM_SOL_EXAMPLE - https://github.com/shogun-toolbox/shogun/commit/19672f38bbe8935156363b1cb2d673c38ff707df | 07:34 |
-!- pluskid [~pluskid@108.171.196.83] has joined #shogun | 07:37 | |
pluskid | sonney2k: around? | 07:37 |
CIA-18 | shogun: puffin444 master * re101512 / (3 files in 2 dirs): Added FITC and visual demos. - https://github.com/shogun-toolbox/shogun/commit/e1015126bfe22765b3b4de82b136ad4ccc2eff50 | 07:38 |
CIA-18 | shogun: Soeren Sonnenburg master * r4279e27 / (9 files in 5 dirs): Merge pull request #683 from puffin444/master - https://github.com/shogun-toolbox/shogun/commit/4279e27bb88b3074b4dcf537745285adffc29041 | 07:38 |
CIA-18 | shogun: puffin444 master * re64e571 / (6 files in 3 dirs): FITC computations are now accurate! - https://github.com/shogun-toolbox/shogun/commit/e64e57183990457736d4d6991a45cf716f6f895c | 07:38 |
CIA-18 | shogun: puffin444 master * r1606ac2 / (7 files in 4 dirs): Cleaned up some of the code - https://github.com/shogun-toolbox/shogun/commit/1606ac2aa79e7b41df74852681f452fdbd5b7c4f | 07:38 |
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving] | 07:46 | |
-!- cronor_ [~cronor@g229081189.adsl.alicedsl.de] has joined #shogun | 08:06 | |
-!- cronor [~cronor@g229081189.adsl.alicedsl.de] has quit [Read error: Connection reset by peer] | 08:06 | |
-!- cronor_ is now known as cronor | 08:06 | |
shogun-buildbot | build #235 of deb1 - libshogun is complete: Failure [failed compile] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/235 blamelist: puffin444 <walke434@msu.edu>, Soeren Sonnenburg <sonne@debian.org> | 08:44 |
-!- cronor_ [~cronor@g229081189.adsl.alicedsl.de] has joined #shogun | 08:45 | |
-!- cronor [~cronor@g229081189.adsl.alicedsl.de] has quit [Read error: Connection reset by peer] | 08:45 | |
-!- cronor_ is now known as cronor | 08:45 | |
-!- uricamic [~uricamic@2001:718:2:1634:39be:d411:577:692b] has joined #shogun | 08:55 | |
@sonney2k | pluskid, now yes | 10:07 |
@sonney2k | pluskid, you too? | 10:07 |
pluskid | sonney2k: yeah | 10:07 |
-!- blackburn [~blackburn@109.226.100.87] has joined #shogun | 10:26 | |
-!- heiko [~heiko@host86-181-156-213.range86-181.btcentralplus.com] has joined #shogun | 11:02 | |
pluskid | blackburn: around? | 11:06 |
blackburn | pluskid: yeah | 11:06 |
pluskid | blackburn: how do we paste code in shogun-tutorial with syntax highlighting? | 11:07 |
blackburn | hmm | 11:07 |
pluskid | personally, I like the minted package: http://code.google.com/p/minted/ | 11:07 |
pluskid | but it requires pygments (an external python package) | 11:07 |
pluskid | and you need to pass -shell-escape to latex when compiling | 11:08 |
pluskid | it use external program to render the coloring | 11:08 |
blackburn | heiko: what do you think? | 11:08 |
blackburn | I don't mind to use it | 11:08 |
pluskid | the result looks beautiful but slows down the compile time a little | 11:08 |
blackburn | the other alternative is listings | 11:08 |
heiko | pluskid is it hard to install? | 11:08 |
pluskid | blackburn: I'm using it for some of my other docs, but I'm not sure other authors are wish to install pygments for this | 11:09 |
heiko | The thing is: do we really need code in the tutorial? | 11:09 |
heiko | wouldnt well documented examples be better? | 11:09 |
pluskid | heiko: if there's no code, then what is tutorial for? | 11:09 |
heiko | for explaining concepts | 11:09 |
heiko | And refering to examples | 11:09 |
pluskid | heiko: I thought shogun-tutor is for explaining concepts *within shogun* ? | 11:10 |
heiko | But actually, you are right, one could add examples in the tutorial that one explains a bit deepter | 11:10 |
pluskid | heiko: "referring to examples" means to examples/undocumented/ directory? | 11:10 |
heiko | yeah, I mean there should be a larger example of all things that one can do, and one could describe them in the tutorial from a greater distance, but the comments in the sourcecode are also important for this | 11:12 |
heiko | but as said, I agree that code snippets are a good thing in the tutorial | 11:12 |
blackburn | probably we would need | 11:13 |
blackburn | snippets | 11:13 |
blackburn | for describing some concepts | 11:13 |
pluskid | heiko: btw, it requires pygments (usually can be installed with easy_install Pygments), and the package it self involves copying the code to some texmf dir I guess (not sure since I use my system package manager to install it) | 11:13 |
blackburn | like multiclass machines and etc | 11:13 |
heiko | ok, well then its fine for me | 11:13 |
blackburn | heiko: pygments is in ubuntu repo so no need to use easy_install | 11:13 |
heiko | we should have a live version of the pdf online anyway | 11:13 |
heiko | (for people that dont like to install this stuff) | 11:14 |
pluskid | heiko: good idea | 11:14 |
pluskid | heiko: but where to put it and how to keep it sync-ed ? | 11:14 |
heiko | pluskid, just make the buildbot compile it everytime its updated and then have a link to the latest pdf | 11:15 |
heiko | just like the shogun doxygen | 11:16 |
pluskid | heiko: OK, I see | 11:16 |
heiko | pluskid, could you add some example on how to use this highlighting? | 11:20 |
pluskid | heiko: yes, I'm writing the multiclass part, I'll make a PR later today with some examples | 11:20 |
heiko | cool | 11:21 |
@sonney2k | pluskid, heiko - I would suggest to put examples only in the examples folder and then have some magic to insert these examples directly in the doc | 11:46 |
@sonney2k | if it is only short snippets of code | 11:46 |
@sonney2k | then things are different of course | 11:46 |
@sonney2k | (not full examples) | 11:46 |
@sonney2k | thunderstorm again! | 11:47 |
pluskid | sonney2k: I'm trying to use a full example and explain part by part in the tutorial, any suggestion for this? | 11:47 |
@sonney2k | pluskid, I tried the same in shogun doc on the website... | 11:48 |
@sonney2k | it partially includes examples from shogun/examples | 11:48 |
@sonney2k | and partially not | 11:48 |
@sonney2k | no idea how to do it better | 11:48 |
pluskid | hmm | 11:49 |
@sonney2k | problem with external tutorial is that it is likely to become outdated if not regularly updated | 11:49 |
pluskid | external? | 11:49 |
@sonney2k | uricamic, please fix the warning in http://shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/235/steps/compile/logs/stdio | 11:50 |
pluskid | internal tutorial will also get out-dated | 11:50 |
@sonney2k | pluskid, not in code | 11:50 |
pluskid | oh, get it | 11:50 |
@sonney2k | pluskid, that is why we use doxygen and shogun/examples | 11:50 |
@sonney2k | this requires just the author of the code to update stuff and the buildbot will tell us about any breakage / errors | 11:51 |
@sonney2k | that is a problem in the tutorial | 11:51 |
pluskid | sonney2k: I thought (and I guess this is also Cheng's idea) shogun/examples are mainly simple test cases to let buildbot to check the code | 11:51 |
@sonney2k | machine learning 'theory' is luckily not affected :D | 11:51 |
@sonney2k | pluskid, nope | 11:51 |
@sonney2k | these are examples showing how to minimally use a certain machine | 11:51 |
@sonney2k | but they return some objects at the end such that we can use them for regression tests too | 11:52 |
@sonney2k | (checking if results stay the same over versions) | 11:52 |
pluskid | sonney2k: Cheng suggest that tutorial code work on some real world examples (like USPS), I don't think it will be afardable for build-bot to constantly run those examples, isn't it? | 11:52 |
@sonney2k | we cannot run it on every commit - maybe weekly or sth | 11:53 |
@sonney2k | pluskid, we even have shogun/applications | 11:53 |
@sonney2k | and already that gets outdated (because not run in tests...) | 11:53 |
pluskid | oh | 11:53 |
pluskid | I see | 11:54 |
pluskid | history is a good lesson | 11:54 |
@sonney2k | I guess if we had 100 developers then things might be better | 11:54 |
@sonney2k | ...or more chaotic :D | 11:54 |
pluskid | then you'll have no time with your kids any more :D | 11:54 |
pluskid | OK, then hope the build-bot for tutorial code could be set up soon | 11:55 |
pluskid | brb | 11:55 |
-!- blackburn [~blackburn@109.226.100.87] has quit [Quit: Leaving.] | 11:56 | |
-!- blackburn [~blackburn@109.226.100.87] has joined #shogun | 11:56 | |
@sonney2k | blackburn, any ideas why the buildbot fails - what does one need to include for eigen3? | 12:03 |
-!- blackburn [~blackburn@109.226.100.87] has quit [Ping timeout: 240 seconds] | 12:05 | |
-!- gsomix [~gsomix@95.153.188.241] has joined #shogun | 12:08 | |
gsomix | good day | 12:08 |
@sonney2k | gsomix, hey! | 12:08 |
-!- blackburn [~blackburn@109.226.80.43] has joined #shogun | 12:09 | |
-!- cronor [~cronor@g229081189.adsl.alicedsl.de] has quit [Quit: cronor] | 12:11 | |
gsomix | sonney2k, my rest at sea is over. | 12:13 |
@sonney2k | gsomix, thunderstorm in progress again here... | 12:13 |
@sonney2k | gsomix, hope you enjoyed it... | 12:13 |
gsomix | woohoo :) | 12:13 |
-!- blackburn [~blackburn@109.226.80.43] has quit [Client Quit] | 12:13 | |
@sonney2k | gsomix, and I am looking forward to your PRs with the zero-memory-copy for python modular and label python buffer stuff | 12:14 |
gsomix | sonney2k, I will be in the train 4-5 August. And then I will be absolutely free. | 12:15 |
gsomix | ufm, I'll send this PR at 6 August. | 12:15 |
CIA-18 | shogun: Soeren Sonnenburg master * r02d643e / src/shogun/regression/gp/FITCInferenceMethod.h : include eigen3 libs to fix compile failure - https://github.com/shogun-toolbox/shogun/commit/02d643e65ab24b36b73fa8cf8cfaa11f954be9e1 | 12:15 |
-!- blackburn [~blackburn@109.226.80.43] has joined #shogun | 12:16 | |
@sonney2k | gsomix, long train rides... | 12:16 |
gsomix | yep, over 42 hours | 12:17 |
gsomix | in cool train w/o air conditioning | 12:17 |
blackburn | sonney2k: you said your deutchebahns suck? | 12:18 |
@sonney2k | blackburn, we sometimes have air con | 12:18 |
@sonney2k | but for sure no windows to open | 12:18 |
@sonney2k | so if it is 30 C outside - pray! | 12:19 |
blackburn | sonney2k: do you have to stay for 7 hours in a city for no reason? | 12:19 |
blackburn | :) | 12:19 |
blackburn | chelyabinsk - st.petersburg train does it | 12:19 |
@sonney2k | blackburn, in germany that is not necessary - we just walk home ;-) | 12:19 |
pluskid | I once was on a train for 33 hours (without a ticket for seat) with people all around | 12:20 |
blackburn | what a bad boy | 12:20 |
pluskid | I got on that train through the window actually | 12:20 |
blackburn | bad^2 | 12:20 |
shogun-buildbot | build #236 of deb1 - libshogun is complete: Failure [failed compile] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/236 blamelist: Soeren Sonnenburg <sonne@debian.org> | 12:20 |
blackburn | pluskid: doesn't it lead to execution in china? :D | 12:21 |
@sonney2k | pluskid, you would have to ride on the roof of the ICE | 12:21 |
@sonney2k | (in germany) | 12:21 |
@sonney2k | but if it runs 250km/h you certainly don't need to worry about air con :D | 12:21 |
pluskid | blackburn: no, you can get on the train and then buy a ticket (a ticket without a seat of course), especially at the crowd time | 12:21 |
blackburn | I see | 12:22 |
pluskid | there're soooooooooooo many people | 12:22 |
pluskid | it's a nightmare to buy ticket for summer/winter holiday | 12:22 |
@sonney2k | I can imagine | 12:22 |
@sonney2k | people already complain here | 12:23 |
pluskid | I once queued outside for 12 hours overnight for a ticket. | 12:23 |
blackburn | I bet small city in china is a city with 'only' 2 millions of people | 12:23 |
pluskid | blackburn: haha, I have no sense of numbers | 12:23 |
@sonney2k | blackburn, almost like the capital of germany :D | 12:23 |
blackburn | sonney2k: I thought it is sth like 5M.. | 12:24 |
@sonney2k | blackburn, same order of magnitude :D | 12:24 |
blackburn | hmm 3.5 | 12:24 |
blackburn | pretty small | 12:24 |
pluskid | all of my friends who has been abroad would say: " there's almost no one in the street there" :D | 12:24 |
@sonney2k | blackburn, buildbot fails due to not having proper ifdef have_eigen3 stuff | 12:24 |
@sonney2k | the class is not known | 12:24 |
blackburn | sonney2k: CHALLENGE ACCEPTED | 12:25 |
blackburn | :D | 12:25 |
@sonney2k | pluskid, when visiting china only hope would be that chinese live in the 1.60-1.70m height so I can breathe fresh air above them :D | 12:26 |
blackburn | hmm interesting | 12:26 |
pluskid | sonney2k: sure :D | 12:26 |
blackburn | I can't :( | 12:26 |
gsomix | blackburn, oh no, I always remember pic with Sasha Gray and title "CHALLENGE ACCEPTED" | 12:26 |
blackburn | gsomix: like pics with sasha grey? | 12:27 |
blackburn | :D | 12:27 |
blackburn | sonney2k: I am having a strange issue with eigen3 | 12:27 |
@sonney2k | CFITCInferenceMethod is unknown if eigen3 is not available | 12:28 |
blackburn | it produces a lot of warnings when included via -I flag | 12:28 |
@sonney2k | and our class list detects it | 12:28 |
blackburn | why does it detect it then? | 12:28 |
@sonney2k | so either we make some empty class available | 12:28 |
gsomix | blackburn, ooops, Grey, of course | 12:28 |
gsomix | ^^' | 12:28 |
pluskid | btw: is the meeting in half an hour or actually tomorrow? | 12:29 |
@sonney2k | gsomix, taht thing http://beerandstupidity.blogspot.de/2011/07/challenge-accepted.html | 12:30 |
@sonney2k | pluskid, in 30 mins yes | 12:30 |
uricamic | sonney2k: ok, I will fix it with next PR | 12:30 |
gsomix | sonney2k, yes! :D | 12:31 |
CIA-18 | shogun: Soeren Sonnenburg master * r5e4b294 / src/shogun/base/class_list.cpp.py : make class list generator script be aware of HAVE_EIGEN3 - https://github.com/shogun-toolbox/shogun/commit/5e4b294c8dba439741de7c98693018e9d3aa8e85 | 12:32 |
@sonney2k | blackburn, next attempt | 12:32 |
@sonney2k | pluskid, btw maintaining all the interfaces is not that much work with swig (and no library / std stuff being used in *headers*) | 12:33 |
@sonney2k | actually close to none | 12:33 |
pluskid | that's cool | 12:33 |
blackburn | sonney2k: I have a strange issue with eigen | 12:35 |
blackburn | sonney2k: do you have it installed? | 12:35 |
@sonney2k | pluskid, that is why I always insist that no one uses magic in headers ... | 12:35 |
@sonney2k | blackburn, locally yes | 12:35 |
shogun-buildbot | build #237 of deb1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/237 | 12:35 |
pluskid | sonney2k: OK, I see | 12:35 |
blackburn | sonney2k: and you see no warnings, right? | 12:35 |
@sonney2k | blackburn, tons of warnings | 12:36 |
blackburn | sonney2k: from eigen? | 12:36 |
@sonney2k | hey a successful build | 12:36 |
@sonney2k | yes | 12:36 |
@sonney2k | eigen has lots of shadowed variables | 12:36 |
blackburn | sonney2k: they go away when I include eigen3/Eigen/Dense | 12:36 |
blackburn | instead of Eigen/Dense | 12:36 |
blackburn | I do not understand why | 12:36 |
@sonney2k | blackburn, I once got bitten by a variable naming clash (shadowed variable) ... was hunting that bug for quite a while. since then this warning is turned on | 12:39 |
@sonney2k | blackburn, do we get other warnings from eigen | 12:39 |
@sonney2k | would the eigen people 'fix' that? | 12:39 |
@sonney2k | or should we disable this warning again... | 12:39 |
blackburn | sonney2k: but why it goes away with different way of include? | 12:40 |
shogun-buildbot | build #143 of bsd1 - libshogun is complete: Failure [failed test] Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/143 blamelist: Soeren Sonnenburg <sonne@debian.org> | 12:41 |
@sonney2k | blackburn, no clue | 12:46 |
-!- cheng [~cheng@115-64-111-17.tpgi.com.au] has joined #shogun | 12:48 | |
-!- shogun-toolbox [~shogun@7nn.de] has quit [Ping timeout: 244 seconds] | 12:52 | |
--- Log closed Fri Aug 03 12:52:32 2012 | ||
--- Log opened Fri Aug 03 12:52:44 2012 | ||
-!- shogun-t1olbox [~shogun@7nn.de] has joined #shogun | 12:52 | |
-!- Irssi: #shogun: Total of 15 nicks [1 ops, 0 halfops, 0 voices, 14 normal] | 12:52 | |
-!- Irssi: Join to #shogun was synced in 9 secs | 12:52 | |
-!- shogun-toolbox [~shogun@7nn.de] has quit [Ping timeout: 244 seconds] | 12:53 | |
-!- blackburn [~blackburn@109.226.80.43] has quit [Ping timeout: 240 seconds] | 12:53 | |
-!- ckwidmer_ [8ca3fe9d@gateway/web/freenode/ip.140.163.254.157] has quit [Ping timeout: 245 seconds] | 12:53 | |
-!- shogun-buildbot [~shogun-bu@7nn.de] has quit [Ping timeout: 244 seconds] | 12:53 | |
wiking | meeting in 10? | 12:53 |
cheng | I'm ready. | 12:54 |
-!- shogun-buildbot_ [~shogun-bu@7nn.de] has joined #shogun | 12:55 | |
-!- alexlovesdata [6d2d0040@gateway/web/freenode/ip.109.45.0.64] has joined #shogun | 12:55 | |
-!- n4nd0 [~nando@192.36.80.8] has joined #shogun | 12:56 | |
n4nd0 | hey! | 12:56 |
n4nd0 | nice internet in the airport :) | 12:57 |
-!- vojtech [9320543b@gateway/web/freenode/ip.147.32.84.59] has joined #shogun | 12:57 | |
n4nd0 | and free! | 12:57 |
pluskid | n4nd0: cool | 12:57 |
pluskid | n4nd0: which country are you in now? | 12:57 |
n4nd0 | pluskid: Sweden | 12:57 |
gsomix | mimimi | 12:58 |
@sonney2k | n4nd0, wow you made it? | 12:59 |
@sonney2k | who is missing? | 12:59 |
-!- puffin444 [62e3926e@gateway/web/freenode/ip.98.227.146.110] has joined #shogun | 12:59 | |
n4nd0 | blackburn? | 12:59 |
@sonney2k | puffin444, welcome | 12:59 |
-!- ckwidmer [45c9b18c@gateway/web/freenode/ip.69.201.177.140] has joined #shogun | 12:59 | |
@sonney2k | chris | 12:59 |
-!- cronor [~cronor@141.23.76.80] has joined #shogun | 12:59 | |
puffin444 | hello | 13:00 |
@sonney2k | heiko, will arthur join? | 13:00 |
ckwidmer | hi all | 13:00 |
@sonney2k | n4nd0, how much time do you have at the airport? | 13:00 |
heiko | sonney2k, unfortunately I have no response so far, sent a text and an email, I feel bad that this always happens since we talk every few days, but I cannot do anything about it | 13:01 |
n4nd0 | sonney2k: I think the boarding will start in about 1h20min | 13:01 |
n4nd0 | so as long as the connection holds, I think I can make it for the whole meeting | 13:01 |
@sonney2k | wiking, what about alex? | 13:02 |
@sonney2k | alexlovesdata, there | 13:02 |
@sonney2k | ok | 13:02 |
@sonney2k | then lets just start | 13:02 |
@sonney2k | Welcome everyone to our 4th meeting. In the last weeks we have seen various PRs from everybody and things have been progressing nicely. | 13:02 |
@sonney2k | That said, the program is coming to an end rather soon. | 13:02 |
@sonney2k | We are very proud what you all have achieved and naturally want to keep you as contributors (maybe even becoming core developers like blackburn/heiko) after GSoC. | 13:02 |
@sonney2k | (it would be blackburns part here to say yes yes yes we want you) | 13:02 |
n4nd0 | :) | 13:03 |
@sonney2k | for some reason he got disconnected few minutes ago | 13:03 |
@sonney2k | so heiko you have to take over this part :D | 13:03 |
heiko | lol :) | 13:03 |
heiko | yes yes yes we want you, | 13:03 |
heiko | join the team, its great fun! | 13:03 |
@sonney2k | we all know real live will take over after gsoc | 13:04 |
@sonney2k | so one cannot commit a lot of time but just occassionally | 13:04 |
heiko | its pretty cool seeing that your stuff is being used | 13:06 |
heiko | and that usually takes some time, so stay, stay! :D | 13:06 |
@sonney2k | heiko did that last year and eventually became a core developer | 13:06 |
@sonney2k | (we cannot mention this often enough) | 13:06 |
@sonney2k | For sure we want to have all of your code in the upcoming shogun 2.0 release. | 13:06 |
@sonney2k | However, to be able to do a release we need to shift focus from features to stability, documentation and examples. | 13:06 |
@sonney2k | So please *everyone* help us to get the buildbot to build on all archs with close to 0 warnings and create nice examples (graphical too) such that we can illustrate the work you did. | 13:06 |
heiko | and btw it opens a lot of doors in uni when you work on an open-source project | 13:06 |
@sonney2k | ideally you use shogun to do baseline experiments and then from time to time contribute some extensions | 13:07 |
-!- blackburn [~blackburn@109.226.80.43] has joined #shogun | 13:07 | |
heiko | sonney2k, it would be great if we could sort out the testing-branch issues in order to maybe have some tests for the release | 13:07 |
@sonney2k | that works well in academia and even industry | 13:07 |
shogun-buildbot_ | build #247 of deb3 - modular_interfaces is complete: Failure [failed test libshogun] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/247 blamelist: Soeren Sonnenburg <sonne@debian.org> | 13:08 |
wiking | heiko: second that... i think we should discuss it today on the meeting to have a consensus about that | 13:08 |
n4nd0 | haha that was very appropriate from shogun-buildbot_ | 13:08 |
@sonney2k | heiko, wiking ok noted down | 13:08 |
pluskid | buildbot wants to join the meeting | 13:08 |
@sonney2k | pluskid, yeah it complains about my merge of puffin444's code :D | 13:08 |
blackburn | ehm my connection sucks | 13:08 |
@sonney2k | alright | 13:09 |
@sonney2k | Mentors - again who wants to go to mountain view on Oct 20/21 to join the mentors summit? That event is a good place to get to know core developers from other organisations (like swig, orange, opencv, ...). | 13:09 |
blackburn | ckwidmer wants I am sure :) | 13:09 |
ckwidmer | sonney2k, same as last time: interested, but let others go if anyone else wants to | 13:10 |
shogun-buildbot_ | build #205 of deb2 - static_interfaces is complete: Failure [failed test libshogun] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/205 blamelist: Soeren Sonnenburg <sonne@debian.org> | 13:10 |
blackburn | ckwidmer: it should be easier for you this time hehe | 13:10 |
ckwidmer | blackburn yes, true :) | 13:11 |
heiko | sonney2k, no students allowed there right? | 13:11 |
@sonney2k | heiko, exactly | 13:11 |
n4nd0 | :( | 13:11 |
@sonney2k | only mentors | 13:11 |
@sonney2k | become a mentor next year and you two can go :D | 13:11 |
@sonney2k | ckwidmer, I guess it will be us two then | 13:12 |
ckwidmer | heiko, mentors get paid in cool rather than money ;) | 13:12 |
ckwidmer | sonney2k, I'm not complaining :) | 13:12 |
heiko | ckwidmer, yeah :) | 13:12 |
@sonney2k | Now as usual, please every mentor/student team - give us a status update, telling us what you did since the last meeting, whether you are behind or on schedule and what has already been merged into shogun and what will be the final steps. | 13:12 |
blackburn | ckwidmer: sonney2k said he is a hard boy so be careful | 13:12 |
ckwidmer | heh | 13:13 |
@sonney2k | blackburn, it is not the first time I share a room with ckwidmer - so he knows pretty well ;-) | 13:13 |
blackburn | I don't want to know what happened last time! | 13:13 |
blackburn | :D | 13:13 |
ckwidmer | yes, sonney2k snores | 13:13 |
heiko | lol :) | 13:14 |
@sonney2k | ckwidmer, that is how I code at night! | 13:14 |
@sonney2k | so who wants to start? | 13:14 |
ckwidmer | heh | 13:14 |
@sonney2k | ckwidmer, blackburn you? | 13:14 |
blackburn | why not | 13:14 |
ckwidmer | sonney2k, sure | 13:14 |
blackburn | okay what we've done | 13:14 |
ckwidmer | blackburn, fire away | 13:14 |
blackburn | I did not prepare the speech lol | 13:14 |
blackburn | there are a few multitask algorithms available now | 13:15 |
@sonney2k | maybe n4nd0 next and then vojtech/uricamic | 13:15 |
blackburn | some time was spent on making it work with xval | 13:15 |
ckwidmer | blackburn, just make a rap video from your commits and upload them to youtube | 13:15 |
blackburn | we actually use these algos to predict on ovarian cancer data | 13:15 |
blackburn | namely, algorithms are L1/Lq task tree/group logistic regression/least squares regression | 13:16 |
blackburn | L1/L2 following argyriou probably popular thing | 13:16 |
blackburn | trace norm regularized logistic regression and clustered multitask logistic regression | 13:16 |
blackburn | SLEP-based algorithms do not support xval yet but it is on the way | 13:16 |
blackburn | moreover I am working on kernel mean matching now | 13:17 |
blackburn | that's something from domain adaptation I want to implement | 13:17 |
@sonney2k | blackburn, so what do you want to have in shogun 2.0? | 13:17 |
ckwidmer | so seems like we have most of the popular MTL methods (non deep learning) stuff in shogun now, which is really nice | 13:17 |
@sonney2k | (and what you would like to do after?) | 13:17 |
blackburn | sonney2k: after gsoc I will get back to edrt and made yet another attempt | 13:18 |
blackburn | as for MTL | 13:18 |
blackburn | well I hope ckwidmer will catch me on the train so we will continue to play around with data | 13:18 |
blackburn | so if something cool appears I can extend MTL part | 13:19 |
blackburn | actually MTL is not so useless sonney2k | 13:19 |
blackburn | it appears to be best classifier on the data we work on now | 13:19 |
ckwidmer | blackburn, never give up hope ;) | 13:19 |
@sonney2k | blackburn, ckwidmer alright thanks - so what is missing then is documentaion / examples mostly right? | 13:20 |
blackburn | sonney2k: I actually want to polish these things because they are not user-friendly yet | 13:20 |
blackburn | yes, doc, examples and safety | 13:20 |
ckwidmer | the good thing is, quite a bunch of ppl here at MSK get interested in this stuff, when realizing that their data set could benefit from these methods | 13:20 |
blackburn | if you pass something wrong it could kaboom | 13:20 |
@sonney2k | recall that we at least intended to release Sept 1. | 13:21 |
ckwidmer | will help us to polish | 13:21 |
@sonney2k | ckwidmer, users are the best testers o_O | 13:21 |
ckwidmer | :) | 13:21 |
ckwidmer | they will complain | 13:21 |
@sonney2k | alright thanks | 13:21 |
@sonney2k | I think we have to continue | 13:21 |
@sonney2k | n4nd0, could you please? | 13:21 |
n4nd0 | sure | 13:22 |
n4nd0 | I implemented the HM-SVM based on the hmsvm toolbox. I have compared the model learnt | 13:22 |
n4nd0 | by my implementation with the one in the toolbox and the results are correct. However, it | 13:22 |
n4nd0 | just works with discrete data and I am currently working on a extension to handle | 13:22 |
n4nd0 | real data for the observations using PLiFs. | 13:22 |
n4nd0 | After this I will focus on improve the documentation and the examples. | 13:22 |
@sonney2k | n4nd0, so plifs will still be done in gsoc and doc / example right? | 13:22 |
@sonney2k | gsoc timeframe | 13:22 |
n4nd0 | yes | 13:22 |
n4nd0 | right now the plifs are taking a bit more time because I am trying to use what it is already in SHOGUN | 13:23 |
@sonney2k | and in case you stick around (which is what we all hope) coffin, etc | 13:23 |
n4nd0 | exactly | 13:23 |
-!- alexlovesdata [6d2d0040@gateway/web/freenode/ip.109.45.0.64] has quit [Ping timeout: 245 seconds] | 13:23 | |
n4nd0 | Unfortunately, it has not been possible to implement conditional random fields. This part was included on the initial schedule. | 13:23 |
@sonney2k | n4nd0, ok thanks - questions to that anyone | 13:23 |
blackburn | actually KMM would work with SO too | 13:24 |
blackburn | so we can have domain adaptation structured output learning | 13:24 |
@sonney2k | blackburn, hehe post shogun 2.0 | 13:24 |
blackburn | a few more words and it would sound like something cool | 13:24 |
wiking | n4nd0: i'd like to have some changes in SO for supporting COFFIN fw. i dont know whether we should discuss it now or later...? | 13:24 |
n4nd0 | wiking: this will be something to do after gsoc | 13:25 |
@sonney2k | wiking, n4nd0 after the meeting I guess | 13:25 |
wiking | sonney2k n4nd0 ok | 13:25 |
n4nd0 | wiking: but we can discuss about it at any moment :) | 13:25 |
wiking | n4nd0: hehe yeah let's do it after as that is what i need as well in latent... anyways after the meeting then if u have time | 13:25 |
@sonney2k | alright next uricamic/vojtech? | 13:26 |
vojtech | ok | 13:26 |
uricamic | ok | 13:26 |
vojtech | he work Michal did since the last meeting: | 13:26 |
vojtech | i. He implemented better cutting plane management for both BMRM and BM. ii. He implemented our variant of Bundle Method for risk minimization. iii. Finally, he has been implementing multiple-cutting plane model improvement (similar to n-slack of Joachims). He will finish in 3-4 days. And that's it. The final steps: to collect some nice benchmarks for StructSVM, testing, documentation, examples. | 13:26 |
vojtech | sorry for the bad formating | 13:26 |
@sonney2k | vojtech, ok nice - I think we really need some nice/cool example for that otherwise people won't know that this is actually cool stuff | 13:27 |
vojtech | I agree | 13:27 |
@sonney2k | n4nd0, blackburn holds for you too | 13:27 |
n4nd0 | uricamic, vojtech : I got a question here | 13:27 |
vojtech | go ahead | 13:28 |
uricamic | so far I have just the multiclass example, I have tried and compared to matlab implementation on MNIST data | 13:28 |
@sonney2k | vojtech, uricamic any chance of continuing afterwards doing libqp business (exctracting things like pr loqo etc)? | 13:28 |
n4nd0 | could I use the models I have worked in my project, the HMSVMModel or the MulticlassModel, with one of your algorithms directly? | 13:28 |
wiking | n4nd0: you need a risk function... | 13:28 |
-!- alexlovesdata [~binder@goldenezahl.ml.tu-berlin.de] has joined #shogun | 13:28 | |
n4nd0 | wiking: aham, I see | 13:29 |
blackburn | ah btw vojtech I patched libqp_gsmo a little | 13:29 |
blackburn | may be you would want to push it into your version | 13:29 |
vojtech | sonney2k: I'm continously improving LIBQP and LIBOCAS if there is a substantial motivation, e.g. recently I added positivity constraint to LIBOCAS which was requested by some computer vision people | 13:29 |
uricamic | n4nd0 yes, but you have to create the RiskFunction for that, anyway in that function it should be possible to use your stuff | 13:29 |
wiking | vojtech: oh that's great news! i'm looking forward to use that extension in latent methods ;) | 13:29 |
n4nd0 | uricamic: nice, thank you! | 13:29 |
uricamic | but you have to wait for the next PR, where I will add support for user data | 13:29 |
vojtech | blackburn, thanks I'll look at it | 13:30 |
@sonney2k | ckwidmer, could you please continue to lead the meeting | 13:30 |
n4nd0 | uricamic: ok | 13:30 |
@sonney2k | (my kids just woke up...) | 13:30 |
ckwidmer | sure | 13:30 |
vojtech | wiking: you can already downloaded | 13:30 |
ckwidmer | I guess I have to wake up, too then (around 7AM) ;) | 13:30 |
wiking | uricamic: remember the chat we had 2 weeks ago, i'll give u some changes i've done to your code to support latent variable in bmrm_data_t | 13:30 |
vojtech | wiking: http://cmp.felk.cvut.cz/~xfrancv/ocas/html/index.html | 13:30 |
wiking | vojtech: ok i'll then try to merge the code into shogun! | 13:30 |
blackburn | vojtech: basically one needs to initialize solution that doesn't violate \sum constraint | 13:30 |
cheng | ckwidmer: I had to cut my Friday beers short. ;-) | 13:30 |
wiking | blackburn: we can do that straight away right? | 13:31 |
blackburn | wiking: what? | 13:31 |
ckwidmer | cheng haha. Ill probably fall asleep in mine | 13:31 |
uricamic | wiking: yep, I remember and I am almost finished with new design to support whatever user data you want | 13:31 |
wiking | blackburn: just take the new release of libocas and update the code in shogun | 13:31 |
wiking | uricamic: great let's see how our two ways collide ;) | 13:31 |
blackburn | wiking: sure why not just push everything | 13:32 |
wiking | blackburn: okie | 13:32 |
ckwidmer | ok | 13:32 |
blackburn | who's next? | 13:32 |
ckwidmer | who wants to go next? | 13:32 |
wiking | i can go | 13:32 |
wiking | i mean alexlovesdata and me | 13:32 |
ckwidmer | ok | 13:32 |
wiking | if that's ok | 13:32 |
ckwidmer | wikiing, sure go ahead | 13:32 |
wiking | ok so basically 2 things happened ever since the last meeting | 13:32 |
uricamic | wiking: but anyways I would like to see your version supporting latend variables, maybe then I will modify the current version to meet better you requirements | 13:33 |
wiking | a) most importantly latent implementations has been PR-ed and merged into shogun master branch. that includes LatentSVM and a preliminary LatentSOSVM. i've added a very simple example but we've agreed with alexlovesdata that we should do a more complex and nicer example | 13:34 |
ckwidmer | (the others can maybe start preparing some short summary) | 13:34 |
wiking | thus i'm currently working on finishing a full blown example for latentSVM and LatentSOSVM based on image-net data | 13:34 |
wiking | basically it'll be a computer vision based problem and i think that'll be a very nice example of how to use these two methods and what are they capable of | 13:35 |
wiking | what is still missing is to finish this example and add modular interface for both of the classifiers | 13:35 |
wiking | and that i add support for bmrm based LatentSOSVM | 13:35 |
blackburn | wiking: I've been studying freak a little | 13:36 |
wiking | b) with heiko i've created a unit testing branch on my repo to add unit testing for the c++ code in shogun because we both thought that it would be great to have it for shogun | 13:36 |
alexlovesdata | and I owe you still the Hog code viking ;) | 13:36 |
blackburn | it seems that it could be a good example of latent model | 13:36 |
blackburn | should pretty fast actually I mean | 13:36 |
wiking | currently the PR for that attempt is at https://github.com/shogun-toolbox/shogun/pull/667 | 13:36 |
ckwidmer | wiking, ok thank you | 13:36 |
wiking | thank you very much for all the comments | 13:36 |
wiking | BUT | 13:36 |
@sonney2k | vojtech, I would very much prefer to have all the qp solver stuff in your libqp and only use that from shogun via some standard interface - then we can (from shogun) just switch solvers on the fly for any of the chunking svms etc | 13:36 |
wiking | ckwidmer: just as a last thing! | 13:37 |
ckwidmer | wiking, sure | 13:37 |
wiking | so i think | 13:37 |
heiko | blackburn, sonney2k, could you have a look on the current approach so that we can decide how to proceed with the testing today? | 13:37 |
wiking | that we should really discuss how to proceed with the unit testing | 13:37 |
wiking | because it's there i've added examples for both simple unit testing and mocking | 13:37 |
blackburn | let me check | 13:37 |
vojtech | sonney2k: ok, I have increased the priority counter for LIBQP by one | 13:37 |
wiking | now we should really try to agree on how to add this into shogun | 13:37 |
wiking | before the 2.0 release | 13:38 |
wiking | naturally i want to add both latentSVM and latentSOSVM into 2.0 release of shogun | 13:38 |
wiking | and i'm comfident that that will happen | 13:38 |
wiking | i'm more 'concerned' about the unit testing | 13:38 |
blackburn | one problem I see | 13:38 |
wiking | since that should be done by all of us together | 13:38 |
wiking | ;) | 13:38 |
wiking | and yeas | 13:38 |
@sonney2k | wiking, sure we do that as submodule :) | 13:38 |
wiking | i've forgot | 13:38 |
blackburn | we wouldn't be able to complete all the tests | 13:39 |
blackburn | before release | 13:39 |
heiko | blackburn, but this is incremental | 13:39 |
wiking | sonney2k: i've given my reasons why i think submoduling will be really bad for the whole concept of unit testing | 13:39 |
blackburn | yeah | 13:39 |
@sonney2k | blackburn, that is not a problem | 13:39 |
@sonney2k | wiking, why? | 13:39 |
blackburn | sonney2k: wiking: yes, submoduling can be a problem | 13:39 |
heiko | so, it can be included in the release | 13:39 |
blackburn | because they are not synchronized | 13:39 |
@sonney2k | they are | 13:39 |
wiking | sonney2k: "my concern is here more about the whole method of how to add new code + corresponding unit tests. | 13:39 |
wiking | that's going to be at least 2 commits + 2 pushes and one extra pulls to get in sync with the submodule..." | 13:39 |
@sonney2k | thats true | 13:40 |
@sonney2k | though I don't see a problem with that | 13:40 |
wiking | and yeah we've try to use shogun with blackburn on a large scale image categorisation challenge | 13:40 |
wiking | it's part of this years Pascal VOC | 13:40 |
@sonney2k | I think we need to postpone the discussion until after this meeting though | 13:40 |
wiking | so if u feel like joining the team of shogun in this | 13:40 |
wiking | please come and contribute because we need you! | 13:41 |
blackburn | wiking: yeah we need to start some progresson that | 13:41 |
wiking | blackburn: indeed :) | 13:41 |
heiko | wiking, what are deadlines and prices? :) | 13:41 |
alexlovesdata | OMG, viking you should have told me about pascal VOC | 13:41 |
blackburn | alexlovesdata: why? | 13:41 |
wiking | alexlovesdata: now i'm saying it ;) | 13:41 |
alexlovesdata | I won last years imageclef photoannotation | 13:41 |
alexlovesdata | I have good features | 13:41 |
wiking | alexlovesdata: :D then come and join us :) | 13:41 |
alexlovesdata | I only lack disk space | 13:41 |
blackburn | hah | 13:41 |
wiking | alexlovesdata: i have that :) | 13:41 |
wiking | alexlovesdata: 7 TB | 13:41 |
blackburn | alexlovesdata: wiking has a lot of | 13:41 |
wiking | if that's enough | 13:41 |
wiking | sonney2k: okey | 13:42 |
alexlovesdata | great .... we should talk after the meeting about pascal voc ;) | 13:42 |
ckwidmer | hehe, ok | 13:42 |
wiking | so please everybody stick around if u can for a max 10 mins | 13:42 |
ckwidmer | seems like sonney2k is back | 13:42 |
wiking | to join the conversation about unit testing | 13:42 |
blackburn | pluskid: cheng may be you next? | 13:42 |
ckwidmer | I would suggest finishing up the reports first | 13:42 |
pluskid | ok, shall I go? | 13:43 |
@sonney2k | alright cheng / pluskid want to contiue? | 13:43 |
@sonney2k | heh | 13:43 |
ckwidmer | but it's fine to discuss unit tests after that | 13:43 |
cheng | I volunteer pluskid to type... ;-) | 13:43 |
pluskid | I implemented the RelaxedTree algorithm. It tooks a bit longer than expected due to my under-estimate of the complexity of its colorization algorithm. But I finally get it hunted. Before the end of GSoC, I'm planning to add content to the shogun-tutorial of multiclass learning and implement an algorithm from "Gun". | 13:43 |
@sonney2k | (and after that puffin444 and gsomix) | 13:43 |
pluskid | and that's it :) | 13:43 |
@sonney2k | pluskid, considering the number of algorithms you contributed you will get lots of users | 13:44 |
cheng | Just a comment here that Gunnar's algorithm is basically cutting planes. Hopefully some of the other members can help. | 13:44 |
@sonney2k | pluskid, dont' forget about examples (graphical if possible) | 13:44 |
pluskid | sonney2k: OK, maybe we could include some nice figures in shogun-tutor | 13:45 |
@sonney2k | otherwise people have no idea how to use all that stuff | 13:45 |
blackburn | cheng: pluskid: I can actually prepare a script for road sign recognition stuff | 13:45 |
@sonney2k | and more importantly what you can do with all that | 13:45 |
blackburn | there you can measure performance | 13:45 |
cheng | blackburn: That's a cool app. | 13:45 |
pluskid | blackburn: cool, thanks! | 13:45 |
@sonney2k | pluskid, if you beat blackburn on that task he will cry for at least 1 hour :D | 13:45 |
blackburn | sonney2k: I won't :) | 13:46 |
blackburn | I don't care now | 13:46 |
cheng | blackburn: if we get it fast enough, I know some people here who want to do it real time. | 13:46 |
pluskid | lol | 13:46 |
@sonney2k | cheng, blackburn has it fast enough | 13:46 |
@sonney2k | (linear svm...) | 13:46 |
@sonney2k | with coffin | 13:46 |
@sonney2k | anyways more comments to pluskid/cheng? | 13:47 |
@sonney2k | if not then thanks pluskid - as usual excellent work | 13:47 |
@sonney2k | puffin444, want to continue? | 13:47 |
cheng | blackburn: http://www.nicta.com.au/research/projects/AutoMap/sign_geo-positioning_service | 13:47 |
blackburn | I am curious whether these cool algoritmhs would beat crammer-singer | 13:47 |
puffin444 | sure | 13:47 |
pluskid | btw: is cutting plane related to bundle method? | 13:47 |
pluskid | I'm totally ignorance of those | 13:47 |
blackburn | cheng: what's technology of recognition? | 13:47 |
@sonney2k | pluskid, yes - vojtech is the ueberexpert here | 13:48 |
pluskid | sonney2k: ok | 13:48 |
cheng | blackburn: No idea, coz I'm not involved. | 13:48 |
puffin444 | A lot has happened in the last month. I had to focus | 13:48 |
pluskid | but I'm afraid I don't even know what to ask now (before making some preliminary reading) | 13:48 |
blackburn | cheng: I used HOG + homogeneous kernel map + linear svm | 13:49 |
@sonney2k | puffin444, please continue | 13:49 |
blackburn | cheng: 97.82% on http://benchmark.ini.rub.de/?section=gtsrb&subsection=results | 13:49 |
blackburn | would be interesting to see better results with pluskid's new algos | 13:50 |
puffin444 | on alot of extra work in the model selection framework to support gradient selection. added also were some ARD kenels, sparse approximtion and a simple visual demo | 13:50 |
puffin444 | sorry about that | 13:50 |
cheng | blackburn: pluskid: Definitely a cool application demo for multiclass. | 13:50 |
pluskid | sure! | 13:51 |
puffin444 | As of now all but one feature in the schedule has bee merged. | 13:51 |
@sonney2k | puffin444, so what do you still want to do for shogun 2.0? | 13:52 |
puffin444 | In the coming days I plan to add laplacian aproimation for inference | 13:52 |
blackburn | and the main question puffin444 would you like to continue? ;) | 13:52 |
puffin444 | as wel as add documentation | 13:52 |
@sonney2k | heiko, see - blackburn is back :D | 13:52 |
heiko | sonney2k, yeah | 13:53 |
heiko | go blackburn, go! :) | 13:53 |
puffin444 | and somemore complex examples | 13:53 |
blackburn | heiko: sonney2k: where I am back to? | 13:53 |
@sonney2k | puffin444, graphical too if possible | 13:53 |
heiko | blackburn, in recruiting new people :D | 13:53 |
@sonney2k | I need nice figures somehow produced with the aid of shogun | 13:53 |
blackburn | it sounds like you had a secret conversation about my recruiting skills :D | 13:53 |
puffin444 | absolutely. sorry for the typing speed. it looslie talets are a horrible idea for irc | 13:53 |
@sonney2k | puffin444, ohh ok | 13:54 |
heiko | blackburn, these and other skills ;) | 13:54 |
@sonney2k | puffin444, great job! | 13:54 |
blackburn | puffin444: where did you get your 86 kg notebook? | 13:54 |
@sonney2k | so gsomix - please continue | 13:54 |
gsomix | sonney2k, ok | 13:54 |
puffin444 | directly From hp 4 years go | 13:54 |
gsomix | Since last meeting I and sonney2k managed to add python protocol's stuff in our swig code. | 13:55 |
blackburn | puffin444: sorry I mean where it is now :) | 13:55 |
@sonney2k | puffin444, I guess oliver has some nice examples (graphical too) - so please ask him - maybe you can easily port things over | 13:55 |
gsomix | sorry, copy-past :] | 13:55 |
gsomix | We have implemented many cool features for DenseFeatures: buffer protocol for really fast memory transactions with numpy objects and for numeric operators (a=mycoolfeatures+numpy.array); mapping and sequence protocol for numpy like slicing (e.g. mycoolfeatures[ : , some_Idx ] = some_const); numeric protocol with using buffer protocol for fast numeric inplace operators (mycoolfeatures+=numpy.array). | 13:55 |
gsomix | At now I'm working on methods for zero-memory-copy working with numpy.array and protocols for Labels, SGVecotr, SGMatrix (last two for 'director classes' stuff). Next I plan work on typemaps for modelselection (hi, heiko) and nice error messages. Btw I need clean up my code and add python3 support. | 13:55 |
gsomix | After GSoC I plan further work with modular interfaces: more director classes, 'protocoled' classes, typemaps and examples... | 13:55 |
@sonney2k | not me - all gsomix's work | 13:55 |
puffin444 | yes he has given some great exampes | 13:55 |
gsomix | and I want work in shogun because there is blackburn (mimimi, =***) | 13:56 |
gsomix | that's all | 13:56 |
heiko | hi gsomix :) | 13:56 |
@sonney2k | cheng, that is sth you would like | 13:56 |
@sonney2k | ^ | 13:56 |
blackburn | no optics though in shogun | 13:56 |
blackburn | and NEVER will be | 13:56 |
blackburn | :D | 13:56 |
@sonney2k | you can do x + y now (x shogun RealFeatures, y numpy array) | 13:56 |
blackburn | ckwidmer must like it too | 13:56 |
@sonney2k | slicing all that works and modifiactaions too | 13:56 |
@sonney2k | ckwidmer, too yes | 13:56 |
blackburn | real pythonista | 13:57 |
ckwidmer | :) | 13:57 |
cheng | sonney2k: excellent! I don't know how to program in any real language anymore... | 13:57 |
@sonney2k | and the no longer copying memory around for python too | 13:57 |
gsomix | btw, I need to go. heiko, when will I be able to discuss with your about typemaps and modelselection? I'll available after 5 August. | 13:57 |
@sonney2k | heiko - you should give gsomix a lecture how to create the parameter / model seection tree | 13:58 |
pluskid | can I just use numpy array in place of where RealFeatures are expected? | 13:58 |
gsomix | *will be available | 13:58 |
blackburn | pluskid: no not yet | 13:58 |
blackburn | :D | 13:58 |
@sonney2k | pluskid, but the other way round! | 13:58 |
gsomix | *when you will be able to | 13:58 |
gsomix | awww :) | 13:58 |
heiko | gsomix, ok | 13:58 |
@sonney2k | pluskid, so you could use e.g. PyML or mlpy or whatever python toolbox and interact ! | 13:58 |
pluskid | ok | 13:58 |
heiko | gsomix, anytime that fits you | 13:58 |
gsomix | heiko, thanks! :) | 13:59 |
@sonney2k | alright thanks gsomix | 13:59 |
gsomix | sonney2k, heiko ok, should to go, soryy | 13:59 |
blackburn | sonney2k: I actually think we could add apply(SGMatrix) | 13:59 |
blackburn | and train(SGMatrix) | 13:59 |
heiko | gsomix, ok bye, just ping me | 13:59 |
@sonney2k | now please some homework - send me nice figures produced with the aid of shogun | 13:59 |
@sonney2k | for the new homepage | 13:59 |
@sonney2k | (next week please) | 13:59 |
@sonney2k | and then finally when shall we have our last celebration - August 16 11 UTC? | 14:00 |
@sonney2k | we will then see how many warnings are left on the buildbot | 14:00 |
heiko | sonney2k, any interest on whats happening with the statistics-framework? | 14:00 |
@sonney2k | heiko, oops | 14:00 |
@sonney2k | sorry | 14:00 |
@sonney2k | heiko, please continue | 14:00 |
heiko | no worries :) | 14:00 |
blackburn | lol | 14:01 |
heiko | wont take long since copy/paste | 14:01 |
blackburn | nobody likes statistics | 14:01 |
heiko | Since last time, I implemented dependence tests using the HSIC. As for the two-sample tests these come in multiple flavours (null-distribution approximations). | 14:01 |
heiko | I did some refactoring/making things more effective for bootstrapping using precomputed kernel matrices (very important for some tests, since these are computed ~250 times) | 14:01 |
@sonney2k | blackburn, no no | 14:01 |
heiko | HSIC tests are done -- with examples (graphical ones to come) | 14:01 |
* sonney2k *loves* statistics | 14:01 | |
heiko | blackburn, but everybody like statistics with kernels :D | 14:01 |
heiko | For synthetical example data, I added a data generator class (please use it everybody!), that makes artificial data available from all interfaces. | 14:01 |
@sonney2k | (blackburn - I had to say that to keep heiko as core dev) | 14:01 |
heiko | A cool thing that I added for linear time two-sample tests (based on MMD) is automatic kernel weight selection in an MKL style, using libqp (thanks for that vojtech) (see NIPS paper to come, currently working on review-reply experiments with Arthur) | 14:01 |
heiko | Apart from that, I wrote the MMD part of the tutorial. This has to be completed with HSIC and some practical stuff. | 14:01 |
heiko | Plans for next weeks: adding streaming feature class to linear time MMD, and then polish: tutorial/documentation/examples, and tests (hopefully in the new framework by wiking) | 14:02 |
heiko | I have some more ideas what could be done, but I think I focus on getting things stable/nicely looking now. | 14:02 |
heiko | [end] | 14:02 |
blackburn | I actually like some things from statistics | 14:02 |
blackburn | but I don't get much | 14:02 |
heiko | sonney2k, blackburn, you can do some really cool things with it | 14:02 |
wiking | heiko: i'm really looking forward to test mmd! | 14:02 |
blackburn | heiko: what can I do with these cool tests? | 14:02 |
wiking | sounds awesome! | 14:03 |
heiko | current experiment here is to place music data on a sine carrier wave via amplitude modulatilon, and then try to dinstinguish two songs from the same interpret | 14:03 |
wiking | heiko: actually have a dataset for that and i'd like to see the weights it finds ... as i have some other methods that generated some 'optimal' weights... | 14:03 |
heiko | balckburn, tell two probability distributions apart with all the gains of "kernel methods" | 14:03 |
heiko | wiking, what kind of data is it and what kind of problem? | 14:04 |
-!- vikram360 [~vikram360@117.192.189.69] has joined #shogun | 14:04 | |
@sonney2k | heiko, what kernel do you use? gaussian or some spline? | 14:04 |
wiking | heiko: i have features from different domains... thus i'm using combined kernels | 14:04 |
heiko | gaussian currently | 14:04 |
heiko | kernels need to be characteristic in order to be usefull | 14:04 |
wiking | heiko: and i'd like to see an optimal weight distrib for the kernels | 14:04 |
heiko | (= their range covers the complete R in the fourier domain) | 14:05 |
heiko | wiking, cool lets try that soon! | 14:05 |
-!- gsomix [~gsomix@95.153.188.241] has quit [Ping timeout: 246 seconds] | 14:05 | |
@sonney2k | ok thanks heiko! | 14:05 |
@sonney2k | one question about the data generators | 14:05 |
@sonney2k | what can they produce? | 14:05 |
wiking | heiko: yep i'll update shogun on that machine where the features are and i'll give it a go... it's available in modular_* interface... or? | 14:05 |
@sonney2k | is there already an example for using them for e.g. python_modular? | 14:05 |
heiko | sonney2k, currently its really loose, just a class where you can add (possibly static) methods that poduce data | 14:06 |
heiko | I use them from my examples | 14:06 |
heiko | my methods return a matrix | 14:06 |
@sonney2k | heiko, and do you properly seed the rnd number generator to get things reproducible? | 14:06 |
heiko | sonney2k, yeah I have started paying attention to that :) | 14:06 |
blackburn | hmm actually I find the idea interesting | 14:06 |
@sonney2k | k | 14:06 |
blackburn | one could say | 14:06 |
heiko | but unfortunately its kind of hard to get it in synch with matlab | 14:06 |
blackburn | just try 4324 seed and you will see a bug | 14:06 |
heiko | but for example | 14:07 |
heiko | I do a seed to compare results of x-validation on normal and precomputed kernel | 14:07 |
heiko | however, thats not data related | 14:07 |
@sonney2k | alright so thanks again heiko | 14:07 |
blackburn | heiko: btw | 14:08 |
@sonney2k | now final meeting on August 16, 11 UTC - is that ok with everyone? | 14:08 |
blackburn | if we had *distributed* model selection it could be a killer feature | 14:08 |
blackburn | do you think so too? | 14:08 |
@sonney2k | blackburn, post shogun 2.0 I would say :D | 14:08 |
@sonney2k | so blackburn note it down as wishlist in the bugtracker | 14:08 |
heiko | sonney2k, 16 is fine more me | 14:08 |
wiking | fine by me | 14:08 |
uricamic | ok for me | 14:08 |
vojtech | ok for me | 14:09 |
blackburn | fine | 14:09 |
n4nd0 | ok | 14:09 |
cheng | good for me. | 14:09 |
heiko | blackburn, I would say this is great, but has to be separate for differnt kinds of model selection | 14:09 |
@sonney2k | puffin444, alexlovesdata ? | 14:09 |
heiko | but grid-search: for sure! | 14:09 |
pluskid | not sure whether I can be there | 14:09 |
blackburn | heiko: yes, for grid (for now) | 14:09 |
pluskid | I will be a student volunteer for KDD, which runs through Aug. 12~16 | 14:09 |
puffin444 | There's a chance that I will be moving to Pittsburgh on that date | 14:09 |
pluskid | If it finishes early, then I guess I can attend | 14:09 |
pluskid | since it is about 19:00 in local time | 14:10 |
heiko | blackburn, actually thats not too hard using threads (only non-locked machines) | 14:10 |
heiko | blackburn, ah, subsets | 14:10 |
heiko | not thread safe | 14:10 |
blackburn | heiko: not parallel | 14:10 |
blackburn | distributed! | 14:10 |
heiko | but one could start seperate processes | 14:10 |
@sonney2k | blackburn, heiko - later! | 14:10 |
heiko | I see | 14:10 |
ckwidmer | 16th ok with me | 14:10 |
@sonney2k | pluskid, puffin444 well then your mentors can hopefully join and we do some mini celebration with only you later | 14:11 |
@sonney2k | alexlovesdata, ^? | 14:11 |
alexlovesdata | ye | 14:11 |
@sonney2k | ok | 14:12 |
@sonney2k | alright then - thanks everyone for attending again | 14:12 |
@sonney2k | continue the nice work | 14:12 |
@sonney2k | and send me *pictures* | 14:12 |
@sonney2k | nice ones for the new website | 14:12 |
@sonney2k | n4nd0, and send bern4rd kudos for his work! | 14:12 |
n4nd0 | sonney2k: sure | 14:12 |
@sonney2k | (on the new website) | 14:12 |
-!- cronor [~cronor@141.23.76.80] has quit [Remote host closed the connection] | 14:12 | |
@sonney2k | case dismissed! | 14:12 |
wiking | shall we start a short discussion on unit testing now for max 10 mins? | 14:13 |
@sonney2k | wiking, go | 14:13 |
-!- cronor [~cronor@fb.ml.tu-berlin.de] has joined #shogun | 14:13 | |
wiking | ok so i've said everyting what i think about how we should include unit testing into shogun in the discussion of the PR | 14:13 |
-!- puffin444 [62e3926e@gateway/web/freenode/ip.98.227.146.110] has quit [Quit: Page closed] | 14:13 | |
heiko | give me 5 mins | 14:13 |
heiko | phone ringing | 14:13 |
wiking | but basically i'm not supporting the idea of using submodules | 14:13 |
wiking | i rather go with a separate branch and when we have enough code | 14:14 |
wiking | we just merge | 14:14 |
wiking | this way i can 'commit' all the stuff we'll need for unit testing | 14:14 |
wiking | and then we'll just have one big PR :P | 14:14 |
wiking | but then again i guess | 14:14 |
wiking | there's no discussion anymore | 14:14 |
wiking | about whether we shoudl use google's c++ unit testing and mocking framework for this, or? | 14:15 |
wiking | but anyhow we go | 14:15 |
wiking | i think what would be the best that if possible everybody who's now working on a gsoc project | 14:15 |
wiking | at least add unit testing to their own part | 14:15 |
@sonney2k | wiking, I am thinking of splitting up shogun even a bit more into shogun / interfaces / data / tutorial / examples&tests | 14:15 |
wiking | since you know the best what should happen with your code | 14:16 |
wiking | creating unit tests and mocking tests are rather simple | 14:16 |
@sonney2k | wiking, the biggest advantage of having a submodule would be that you can directly work on that stuff and merge at your own will | 14:16 |
wiking | and i've added a very simple example into the PR | 14:16 |
wiking | sonney2k: yep i know :) | 14:16 |
wiking | sonney2k: that's very tempting | 14:16 |
wiking | sonney2k: but my concern is more about how a PR would go | 14:16 |
wiking | let's say after we get unit testing into a good state | 14:17 |
wiking | and we could require that each new module/class | 14:17 |
wiking | must have unit testing added as well | 14:17 |
wiking | then it'll be very complicated | 14:17 |
wiking | to do a simple PR | 14:17 |
heiko | re | 14:17 |
wiking | sonney2k: because it'd be great to have one of those pre-commit hooks working in shogun repo for developers... of course we cannot enforce developers to use that, but still we could ask them nicely in a README.developer | 14:18 |
wiking | or something | 14:18 |
@sonney2k | wiking, I cannot follow... what do you want to discuss? | 14:18 |
@sonney2k | wiking, yes that makes sense | 14:18 |
@sonney2k | but that is just extra candy | 14:18 |
wiking | sonney2k: the submodule vs directly into shogun repo | 14:19 |
blackburn | I do not think hooks are so useful | 14:19 |
n4nd0 | all right, time to get on board | 14:19 |
n4nd0 | see you guys, thank you for the meeting! | 14:19 |
blackburn | n4nd0: have a good flight | 14:19 |
n4nd0 | blackburn: thank you :) | 14:19 |
@sonney2k | n4nd0, have a nice flight | 14:19 |
n4nd0 | thanks | 14:19 |
-!- n4nd0 [~nando@192.36.80.8] has quit [Quit: leaving] | 14:19 | |
wiking | sonney2k: and that to require a little bit of unit tests from all the gsoc developers | 14:19 |
wiking | if possible | 14:19 |
heiko | n4nd0 have a good flight! | 14:19 |
wiking | n4nd0 safe tripe! | 14:19 |
wiking | *trip | 14:19 |
wiking | sonney2k: so for example could we agree to ask all the gsoc-ers in shogun to create unit test, except gsomix? :) | 14:20 |
wiking | sonney2k: nothing so fancy... just something really basic | 14:20 |
@sonney2k | wiking, I think we are a little late for that now (requiring contributions from everyone - maybe one example or so from each gsocer) | 14:20 |
heiko | wiking, I would agree on that, at least some | 14:20 |
@sonney2k | ... is the only thing we can possibly manage | 14:21 |
wiking | sonney2k: i mean really i dont want any full blown mocking stuff | 14:21 |
wiking | just a very basic unit testing | 14:21 |
wiking | for at least one class then :) | 14:21 |
@sonney2k | wiking, for some part of the code - ok | 14:21 |
wiking | sonney2k: yeah that'd be more than enough i think | 14:21 |
wiking | sonney2k: just to keep it a little flowing | 14:21 |
wiking | so maybe somebody could pick up the thread later... | 14:21 |
@sonney2k | yeah | 14:22 |
@sonney2k | ok so back to submodule / not submodule | 14:22 |
wiking | ok | 14:22 |
wiking | i'm really | 14:22 |
@sonney2k | what is the advantage of having it not as submodule? | 14:22 |
wiking | up for getting examples out to a submodule | 14:22 |
wiking | :)) | 14:22 |
@sonney2k | 1 commit, 1 PR right? | 14:22 |
wiking | sonney2k: yes | 14:22 |
@sonney2k | anything else? | 14:22 |
wiking | sonney2k: this way you can really see that unit testing is at least handled somehow... | 14:23 |
wiking | sonney2k developers must use it | 14:23 |
wiking | sonney2k: as git submodule --init is optional ;) | 14:23 |
@sonney2k | anyone else have an opinion on that? | 14:24 |
wiking | sonney2k: i know that currently almost everybody is using the git HEAD of the master when using shoung | 14:24 |
@sonney2k | wiking, most people just need the source | 14:24 |
wiking | but idealy only developers would do that | 14:24 |
heiko | I like this approach, this way, only developers see it | 14:24 |
@sonney2k | they don't care about tests :) | 14:24 |
wiking | sonney2k: yes... but what i'm saying is that actually most people wouldn't need the source at all :))) | 14:24 |
wiking | sonney2k: just the binary release :P | 14:25 |
@sonney2k | wiking, we are still not there yet... | 14:25 |
wiking | sonney2k: i know of course... | 14:25 |
wiking | i'm just saying that as an end goal | 14:25 |
wiking | that would be the ideal | 14:25 |
@sonney2k | blackburn, pluskid, uricamic opinions? | 14:25 |
wiking | yeah but let's just do a quick vote on this, or? :) | 14:26 |
wiking | with sonney2k having a veto on it ;) | 14:26 |
@sonney2k | wiking, I don't have strong feelings on that | 14:26 |
blackburn | sorry I am ab it lost | 14:26 |
pluskid | no preference here :-/ | 14:26 |
wiking | blackburn: submodule or no submodule for unit testing | 14:26 |
wiking | blackburn: or dontcare ;) | 14:27 |
@sonney2k | just thought that it is more clean and convenient to have it all in one tests/ dir and let you take control over it | 14:27 |
blackburn | wiking: so what is the question? :) | 14:27 |
wiking | blackburn: :DDDDDDDDDDDDDDDDDDDDDDDDDD | 14:27 |
wiking | blackburn: VODKA?: ) | 14:27 |
blackburn | wrong | 14:27 |
* wiking failed big time :) | 14:27 | |
@sonney2k | wiking, vodka is the answer! | 14:27 |
@sonney2k | not the question | 14:27 |
blackburn | exactly | 14:28 |
wiking | blackburn: more vodka? | 14:28 |
wiking | ok | 14:28 |
blackburn | it is the answer again | 14:28 |
wiking | ok anyhow :>> as i'm failing here | 14:28 |
wiking | submodule then? | 14:28 |
wiking | i mean it'll be a little bit of a hassle | 14:28 |
@sonney2k | blackburn, heiko, pluskid - submodule y/n/<don't care> | 14:28 |
blackburn | n | 14:29 |
blackburn | hmm | 14:29 |
alexlovesdata | I do not understand the problem really, I hope none is angry about it | 14:29 |
blackburn | n | 14:29 |
pluskid | don't care | 14:29 |
wiking | heiko: ? | 14:29 |
blackburn | alexlovesdata: I am crazy angry about that | 14:29 |
blackburn | :D | 14:29 |
wiking | blackburn: :D | 14:29 |
* wiking just migrating to llvm 4.0... yey :) | 14:29 | |
blackburn | wiking: will you be available 2 hrs later? | 14:29 |
@sonney2k | alexlovesdata, russians always are (when they lack the right level of vodka in blood) | 14:29 |
wiking | blackburn: i hoped not to be | 14:29 |
blackburn | argh | 14:29 |
-!- ckwidmer [45c9b18c@gateway/web/freenode/ip.69.201.177.140] has quit [Quit: Page closed] | 14:29 | |
wiking | blackburn: maybe 5-6 hours later? | 14:29 |
blackburn | 5-6 would work too | 14:30 |
wiking | heiko: pinguuu | 14:30 |
blackburn | the more the better | 14:30 |
wiking | blackburn: that's ok by me i guess ;P | 14:30 |
blackburn | wiking: not sure if alexlovesdata would accept it though | 14:30 |
blackburn | so just let me know what you've decide | 14:30 |
heiko | wiking, I also dont really have a preference as both approaches have their downsides | 14:30 |
alexlovesdata | accept what? | 14:30 |
wiking | ok | 14:30 |
blackburn | alexlovesdata: 5-6 hrs later :) | 14:30 |
wiking | sonney2k: your call :) | 14:30 |
alexlovesdata | you mean to talk in 5-6 hours? | 14:30 |
blackburn | yeah | 14:30 |
wiking | alexlovesdata: yes | 14:30 |
@sonney2k | wiking, I prefer submodule | 14:31 |
wiking | sonney2k: ok then we go with that | 14:31 |
alexlovesdata | then lets chat at 9 PM ? | 14:31 |
@sonney2k | wiking and you in code? | 14:31 |
blackburn | what's timezone? | 14:31 |
@sonney2k | no idea what the vote is now | 14:31 |
wiking | sonney2k: well i'm for n | 14:31 |
alexlovesdata | middle european ... too late for you (+1 I think) ... | 14:31 |
wiking | sonney2k: so 2 non-submodule, 3 dontcare | 14:31 |
alexlovesdata | ot are you living in Tshukshcia? | 14:31 |
blackburn | not too late I live in UTC-5 | 14:31 |
blackburn | virtually | 14:31 |
blackburn | :D | 14:31 |
@sonney2k | wiking, so then no submodule | 14:31 |
wiking | sonney2k: but u have the veto :)))) | 14:31 |
blackburn | alexlovesdata: tshukshcia :D | 14:32 |
wiking | sonney2k: u shall use it wisely ;P | 14:32 |
@sonney2k | as I said I don't have strong feelings about it | 14:32 |
wiking | sonney2k: ok... then no submoodule | 14:32 |
wiking | sonney2k: then can we merge? | 14:32 |
heiko | wiking, how to proceed now? | 14:32 |
@sonney2k | and one can at any time convert a dir into a submodule | 14:32 |
alexlovesdata | yeee. wiking got it ;) | 14:32 |
blackburn | alexlovesdata: it is a name of people so you should rather ask 'are you tshukcia?' | 14:32 |
blackburn | :D | 14:32 |
wiking | sonney2k: or is there something you want it to be change still in that PR? | 14:32 |
@sonney2k | argh puffin left already | 14:33 |
alexlovesdata | I did not know how the landscape would be named ;) | 14:33 |
wiking | heiko: i think now it's time to ask nicely gsocers to try to add some simple unit testing for their code, after the PR has been merged | 14:33 |
blackburn | alexlovesdata: chukotka | 14:33 |
alexlovesdata | ok!! | 14:33 |
heiko | wiking, Ill merge it and fly over it, see how its works etc | 14:33 |
wiking | heiko: and for you and me to crazy produce unit testing :))) | 14:33 |
wiking | heiko: GREAT! | 14:33 |
alexlovesdata | wiking, blackburn ... so at what time we like to talk about VOC? | 14:34 |
wiking | ok so i think then unit testing is settled | 14:34 |
heiko | I will write a mail later this week, nicely asking people to add tests for the most iportant methods, and how to design these etc | 14:34 |
alexlovesdata | 9 PM local time ok for wiking? | 14:34 |
wiking | heiko: sounds like a plan! | 14:34 |
wiking | alexlovesdata: for sure not | 14:34 |
wiking | alexlovesdata: i've just realised | 14:34 |
heiko | ok Ill merge .. | 14:34 |
wiking | alexlovesdata: i have family errands to attend to :( | 14:34 |
blackburn | alexlovesdata: I live in place where lada crap is produced http://autoteile-blog.daparto.de/wp-content/themes/Lada-Niva.jpg :D | 14:34 |
CIA-18 | shogun: Heiko Strathmann master * r5847502 / (272 files in 36 dirs): Merge pull request #667 from vigsterkr/utest (+5 more commits...) - https://github.com/shogun-toolbox/shogun/commit/5847502840dc0b3c67c55360d20612e8460ba716 | 14:34 |
alexlovesdata | I will also work this weekend | 14:34 |
wiking | can we have a little chate | 14:34 |
wiking | heiko: WOOOHOOOO!!! | 14:34 |
wiking | alexlovesdata: me too | 14:35 |
wiking | blackburn: weekend? | 14:35 |
heiko | wiking, that was a long one :) | 14:35 |
blackburn | alexlovesdata: wiking: the time is ok for me | 14:35 |
blackburn | anytime | 14:35 |
blackburn | :D | 14:35 |
wiking | so anybody interested in pascal VOC this year? | 14:35 |
wiking | blackburn: ok cool | 14:35 |
heiko | I might be, but not before september | 14:35 |
alexlovesdata | lada niva is produced in UTC -5 ??? | 14:35 |
heiko | And I gotta go now, buy guys! | 14:35 |
blackburn | alexlovesdata: you skipped 'virtually' word | 14:35 |
wiking | heiko: cya! and thanks | 14:35 |
wiking | ok so anyhow | 14:35 |
blackburn | alexlovesdata: I live in UTC+4 but virtually live like in UTC-5 :D | 14:35 |
wiking | just to keep alexlovesdata updated | 14:35 |
wiking | and then i'll have to leave as well | 14:36 |
alexlovesdata | so half-way chukotka | 14:36 |
-!- heiko [~heiko@host86-181-156-213.range86-181.btcentralplus.com] has left #shogun [] | 14:36 | |
wiking | alexlovesdata: we are currently trying to participate in this actually | 14:36 |
alexlovesdata | in what of the pascal VOC categories? | 14:36 |
blackburn | exactly | 14:36 |
wiking | alexlovesdata: http://www.image-net.org/challenges/LSVRC/2012/ | 14:36 |
blackburn | some dogs IIRC | 14:36 |
alexlovesdata | so the large scale stuff | 14:36 |
wiking | alexlovesdata: mainly task1 and task2 | 14:37 |
wiking | alexlovesdata: but of course VOC woudl be great as well | 14:37 |
wiking | i have a cluster of 9 nodes | 14:37 |
wiking | running hadoop | 14:37 |
wiking | and have 7tb space on a hdfs | 14:37 |
alexlovesdata | how many CPUS? | 14:37 |
alexlovesdata | node = ? | 14:37 |
wiking | alexlovesdata: 9*4 | 14:37 |
wiking | or sorry!! | 14:37 |
alexlovesdata | good | 14:37 |
wiking | only 8 nodes | 14:37 |
blackburn | guys just discuss it now if you have time now and let me know what can I help with | 14:37 |
wiking | so 8*4 cpu | 14:38 |
alexlovesdata | for imageclef2011 I had only 32 32 bit cpus for feature comp and that was enough | 14:38 |
wiking | alexlovesdata: we thought about starting with this paper's ideas: | 14:38 |
wiking | just a sec | 14:38 |
wiking | i have to find it | 14:38 |
wiking | Large-scale Image Classification: | 14:39 |
wiking | Fast Feature Extraction and SVM Training | 14:39 |
wiking | https://www.linyq.com/~linyqcom/0694.pdf | 14:39 |
alexlovesdata | how many train and test data precisely ? | 14:39 |
-!- vikram360 [~vikram360@117.192.189.69] has quit [Ping timeout: 244 seconds] | 14:39 | |
wiking | alexlovesdata: There are a total of 1281167 images for training. The number of images for each synset ranges from 732 to 1300. | 14:40 |
wiking | There are 50,000 validation images, with 50 images per synset. | 14:40 |
alexlovesdata | you know the CVPR 2012 paper by vedaldi and zisserman on how to gett tht coding right? | 14:40 |
wiking | alexlovesdata: exactluy | 14:40 |
alexlovesdata | the problem in this coding is to get the parameters ... and it is very slow | 14:40 |
alexlovesdata | you will not have time to recompute the coding 10 times | 14:41 |
wiking | anyhow | 14:41 |
wiking | the thing is | 14:41 |
wiking | that i have to go now :))) | 14:41 |
wiking | so i just wanted to give an update what we thought about doing | 14:41 |
alexlovesdata | so when will we talk about it? Saturday afternoon? | 14:41 |
wiking | and to spread the word | 14:41 |
wiking | that we are trying to do something here with blackburn | 14:41 |
wiking | :))) | 14:41 |
-!- cheng [~cheng@115-64-111-17.tpgi.com.au] has quit [Quit: Leaving.] | 14:41 | |
wiking | and that we need ideas and manpower :) | 14:42 |
wiking | but yes | 14:42 |
shogun-buildbot_ | build #144 of bsd1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/144 | 14:42 |
wiking | saturday would work for me in the afternoon | 14:42 |
wiking | so let's try that | 14:42 |
wiking | alexlovesdata: ok? | 14:42 |
wiking | alexlovesdata: and send me hog please when u can :))) | 14:42 |
alexlovesdata | ok, then I will send around a mail to you ... when I have finished the first draft of my diss until Sat afternoon/evening | 14:42 |
alexlovesdata | yes, you will get hte hog | 14:42 |
alexlovesdata | I have ideas how to get a reasonable result on that | 14:42 |
wiking | okeeey | 14:43 |
blackburn | that's pretty cool | 14:43 |
wiking | but now i really gotta go | 14:43 |
alexlovesdata | pls go | 14:43 |
wiking | family errands started 30 mins ago already :))) | 14:43 |
alexlovesdata | lets talk another time | 14:43 |
wiking | so tomorrow | 14:43 |
alexlovesdata | tomorrow | 14:43 |
blackburn | alexlovesdata: did you try FREAK already? | 14:43 |
wiking | and thnx | 14:43 |
alexlovesdata | what is FREAK? | 14:43 |
blackburn | alexlovesdata: some cool descriptor | 14:44 |
@sonney2k | blackburn, is that from opencv? | 14:44 |
blackburn | sonney2k: yes already in | 14:44 |
blackburn | http://infoscience.epfl.ch/record/175537/files/2069.pdf | 14:44 |
blackburn | here is the paper | 14:44 |
blackburn | sonney2k: there is a standalone code at github too as well | 14:45 |
shogun-buildbot_ | build #145 of bsd1 - libshogun is complete: Failure [failed test] Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/145 blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>, Viktor Gal <viktor.gal@maeth.com> | 14:45 |
CIA-18 | shogun: Soeren Sonnenburg master * rfe8dddc / examples/undocumented/libshogun/regression_gaussian_process_fitc.cpp : ifdev have_eigen3 regression_gaussian_process_fitc.cpp example - https://github.com/shogun-toolbox/shogun/commit/fe8dddc971f98eb0bf8200ef9de50f1276262b3c | 14:49 |
@sonney2k | blackburn, can we directly use opencv for such stuff? | 14:55 |
blackburn | sonney2k: what do you mean? | 14:55 |
blackburn | use for what exactly? | 14:55 |
CIA-18 | shogun: Soeren Sonnenburg master * rd02abd4 / (2 files): ifdef examples requiring NLOPT with HAVE_NLOPT - https://github.com/shogun-toolbox/shogun/commit/d02abd49cc050596dd9e1bcc09e14236d7984275 | 14:59 |
@sonney2k | for any kind of coffin based features if required | 15:00 |
blackburn | sonney2k: well in case of only one keypoint I think wrapper can be easily done | 15:00 |
blackburn | I don't really know what is better to wrap | 15:01 |
blackburn | shogun machine to opencv or opencv stuff to shogun's features | 15:02 |
blackburn | okay will be back in ~3 hrs | 15:02 |
shogun-buildbot_ | build #248 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/248 | 15:17 |
shogun-buildbot_ | build #206 of deb2 - static_interfaces is complete: Failure [failed test libshogun] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/206 blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>, Viktor Gal <viktor.gal@maeth.com> | 15:19 |
-!- uricamic [~uricamic@2001:718:2:1634:39be:d411:577:692b] has quit [Quit: Leaving.] | 15:27 | |
@sonney2k | blackburn, license wise we can use / link to opencv's stuff but not the other way round | 15:36 |
shogun-buildbot_ | build #249 of deb3 - modular_interfaces is complete: Failure [failed test libshogun] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/249 blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>, Viktor Gal <viktor.gal@maeth.com> | 15:43 |
shogun-buildbot_ | build #146 of bsd1 - libshogun is complete: Failure [failed test] Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/146 blamelist: Soeren Sonnenburg <sonne@debian.org> | 15:48 |
-!- vikram360 [~vikram360@117.216.149.224] has joined #shogun | 15:58 | |
shogun-buildbot_ | build #250 of deb3 - modular_interfaces is complete: Failure [failed test libshogun] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/250 blamelist: Soeren Sonnenburg <sonne@debian.org> | 16:07 |
shogun-buildbot_ | build #207 of deb2 - static_interfaces is complete: Failure [failed test libshogun] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/207 blamelist: Soeren Sonnenburg <sonne@debian.org> | 16:09 |
-!- vikram360 [~vikram360@117.216.149.224] has quit [Ping timeout: 244 seconds] | 16:12 | |
-!- vikram360 [~vikram360@117.192.174.5] has joined #shogun | 16:12 | |
-!- vikram360 [~vikram360@117.192.174.5] has quit [Ping timeout: 255 seconds] | 16:19 | |
-!- vikram360 [~vikram360@117.192.166.45] has joined #shogun | 16:19 | |
-!- tiger_eye [~tiger_eye@rrcs-24-106-116-254.central.biz.rr.com] has joined #shogun | 16:22 | |
-!- vikram360 [~vikram360@117.192.166.45] has quit [Ping timeout: 240 seconds] | 16:32 | |
CIA-18 | shogun: Soeren Sonnenburg master * r61ff3da / (2 files): fix include path: base/config.h -> lib/config.h - https://github.com/shogun-toolbox/shogun/commit/61ff3daa6366c65fa45426f19486789764e6bde8 | 16:44 |
-!- vojtech [9320543b@gateway/web/freenode/ip.147.32.84.59] has quit [Quit: Page closed] | 16:49 | |
shogun-buildbot_ | build #147 of bsd1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/147 | 16:50 |
shogun-buildbot_ | build #251 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/251 blamelist: Soeren Sonnenburg <sonne@debian.org> | 17:10 |
shogun-buildbot_ | build #208 of deb2 - static_interfaces is complete: Failure [failed test cmdline_static] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/208 blamelist: Soeren Sonnenburg <sonne@debian.org> | 17:12 |
-!- cronor [~cronor@fb.ml.tu-berlin.de] has quit [Quit: cronor] | 17:13 | |
-!- pluskid [~pluskid@108.171.196.83] has quit [Quit: Leaving] | 18:21 | |
blackburn | alexlovesdata: thanks for sharing some thoughts | 19:05 |
alexlovesdata | :) | 19:05 |
alexlovesdata | funny I am about to go now | 19:05 |
alexlovesdata | instead of PCA we could use sift with 3x3 | 19:05 |
alexlovesdata | is probably more sane for reducing dimension load eg for fisher vectors if you like to go for them | 19:06 |
-!- alexlovesdata [~binder@goldenezahl.ml.tu-berlin.de] has left #shogun [] | 19:06 | |
@sonney2k | wiking, seen the buildbot breakage? | 20:10 |
@sonney2k | due to test move stuff | 20:10 |
-!- sonney2k_ [~shogun@7nn.de] has joined #shogun | 20:18 | |
-!- zxtx [~zv@173-228-88-191.dsl.dynamic.sonic.net] has joined #shogun | 20:18 | |
-!- sonney2k [~shogun@7nn.de] has quit [Ping timeout: 265 seconds] | 20:19 | |
-!- puffin444 [62e3926e@gateway/web/freenode/ip.98.227.146.110] has joined #shogun | 20:21 | |
-!- tiger_eye [~tiger_eye@rrcs-24-106-116-254.central.biz.rr.com] has left #shogun ["Leaving"] | 20:27 | |
-!- Netsplit *.net <-> *.split quits: sonney2k_ | 20:36 | |
-!- Netsplit *.net <-> *.split quits: shogun-buildbot_, octopine, blackburn | 20:36 | |
-!- Netsplit *.net <-> *.split quits: zxtx | 20:36 | |
-!- Netsplit *.net <-> *.split quits: naywhayare | 20:36 | |
-!- Netsplit over, joins: zxtx, blackburn, shogun-buildbot_, naywhayare, octopine | 20:37 | |
-!- puffin444 [62e3926e@gateway/web/freenode/ip.98.227.146.110] has quit [Ping timeout: 245 seconds] | 20:40 | |
-!- puffin444 [62e3926e@gateway/web/freenode/ip.98.227.146.110] has joined #shogun | 21:02 | |
-!- puffin444 [62e3926e@gateway/web/freenode/ip.98.227.146.110] has quit [Ping timeout: 245 seconds] | 21:39 | |
-!- puffin444 [62e3926e@gateway/web/freenode/ip.98.227.146.110] has joined #shogun | 21:42 | |
-!- KMcQuisten [d8338942@gateway/web/freenode/ip.216.51.137.66] has joined #shogun | 22:04 | |
KMcQuisten | Hello again! | 22:04 |
blackburn | hey | 22:04 |
blackburn | what is problem today? ;) | 22:05 |
KMcQuisten | Thanks for the help yesterday. I think I have another small problem. I think something is getting freed twice when I use DirectorKernels in MKL. The process runs fine and I get my results, but the script segfaults upon completion. Once again I have a script to send you to reproduce. | 22:05 |
blackburn | sure send me | 22:06 |
blackburn | thank you for improving user experience :) | 22:06 |
KMcQuisten | No problem. I use this package every day, so i'm glad to help make it better | 22:09 |
blackburn | KMcQuisten: what is you are applying it for btw? | 22:09 |
KMcQuisten | Modeling sequence and target data to predict the efficacy of RNAi gene knockdown constructs. | 22:10 |
blackburn | I think we should track it down somehow | 22:11 |
blackburn | so you are using string features widely, right? | 22:11 |
KMcQuisten | Yes. I use StringCharFeatures, StringWordFeatures to implement Spectrum Kernel, and RealFeatures to model thermodynamic properties of nucleotide sequences | 22:13 |
blackburn | sounds cool | 22:13 |
KMcQuisten | Reproduction script sent | 22:13 |
blackburn | I am asking because I am not really into that stuff | 22:14 |
KMcQuisten | It's quite fun. We're going to use it to predict RNAi sequences that will effectively knockdown genes critical to cancer growth | 22:14 |
blackburn | I am collaborating with c. widmer on some cancer as well | 22:14 |
KMcQuisten | Let the computer figure it out instead of spending millions in a lab | 22:14 |
blackburn | but for now I only have worked on microarray data | 22:15 |
KMcQuisten | Right on | 22:15 |
blackburn | okay thanks for the script let me check now | 22:15 |
KMcQuisten | That's great for the large scale ON/OFF switching to see what genes are on in tumors, then we do specific screens against those genes to build datasets for modelling knockdown of those genes. | 22:16 |
blackburn | I feel stupid now :D | 22:16 |
KMcQuisten | Ha! Don't feel bad. I knew nothing about molecular genetics when I started work on this project. Just a lowly mathematician among the bio nerds :) | 22:17 |
blackburn | I wish I could take some course on that but that country is just totally outdated | 22:18 |
KMcQuisten | ? | 22:19 |
blackburn | I mean nobody in my university and probably even city knows anything about molecular biology :) | 22:20 |
-!- zxtx [~zv@173-228-88-191.dsl.dynamic.sonic.net] has quit [Ping timeout: 248 seconds] | 22:20 | |
KMcQuisten | Ah, I see | 22:20 |
blackburn | and no optional courses in educational system at all | 22:21 |
blackburn | KMcQuisten: where editdist comes from? | 22:25 |
KMcQuisten | Gah! It's another package. I forgot that it was in there | 22:25 |
KMcQuisten | Let me edit the definition | 22:25 |
KMcQuisten | and resend | 22:25 |
blackburn | KMcQuisten: what can I use instead? | 22:26 |
KMcQuisten | sum([1 for x,y in zip(s1,s2) if x!=y]) | 22:27 |
KMcQuisten | that's hamming distance, a bit simpler | 22:27 |
KMcQuisten | but it should work without needing a reference to the editdist package | 22:28 |
blackburn | I used sequence matcher but it fails to converge | 22:28 |
blackburn | let me try hamming-like | 22:28 |
puffin444 | blackburn, is the build still broken? | 22:29 |
blackburn | okay I reproduced it | 22:29 |
blackburn | I don't know :D | 22:29 |
KMcQuisten | Nope, I rebuilt about 20 min ago and it's fine | 22:29 |
blackburn | puffin444: well check that http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/251/steps/test%20python_modular/logs/stdio | 22:29 |
blackburn | gaussian_process_modular seems to die or so | 22:30 |
KMcQuisten | I always do a fresh pull and rebuild to make sure my bugs are still bugs before I come in here for help :) | 22:30 |
blackburn | KMcQuisten: btw you can use apply now | 22:30 |
blackburn | with libsvr | 22:30 |
KMcQuisten | Yes I noticed :) got it all sqaured away in my code | 22:30 |
KMcQuisten | thanks blackburn | 22:30 |
blackburn | not the best way we did | 22:31 |
blackburn | so one need to track properly these renames | 22:31 |
KMcQuisten | Yeah, making swig magic happen is easily broken | 22:31 |
puffin444 | uh oh, the buildbot doesn't have pylab? | 22:32 |
blackburn | puffin444: is it a graphicla examplE? | 22:32 |
blackburn | put it into the graphical then ;) | 22:32 |
puffin444 | yes that is the one | 22:32 |
blackburn | KMcQuisten: interesting bug | 22:33 |
blackburn | KMcQuisten: did director kernel work for you before? | 22:33 |
KMcQuisten | Yes, we did. I think some of the recent changes with ref counting may have broken this. | 22:34 |
KMcQuisten | It had worked, and it's still working, save for not properly deallocating at the end. | 22:35 |
KMcQuisten | The machine trains fine, I can extract alphas and support vectors, and make predictions on unseen functions, it just crashes when the object is destroyed. | 22:35 |
blackburn | yeah I see | 22:36 |
blackburn | actually reference counting works bad here | 22:37 |
KMcQuisten | oh? | 22:39 |
blackburn | KMcQuisten: each get_lhs and get_rhs increases reference counter but it doesn't get down hmmmm | 22:40 |
KMcQuisten | odd | 22:42 |
blackburn | KMcQuisten: in your example it goes to 124747 :D | 22:43 |
KMcQuisten | Gah! | 22:43 |
KMcQuisten | That's no good | 22:43 |
KMcQuisten | Ok, I have to run to a research group meeting. Feel free to email me if you discover anything | 22:48 |
blackburn | sure | 22:49 |
-!- KMcQuisten [d8338942@gateway/web/freenode/ip.216.51.137.66] has quit [Quit: Page closed] | 22:49 | |
-!- puffin444 [62e3926e@gateway/web/freenode/ip.98.227.146.110] has quit [Quit: Page closed] | 23:11 | |
CIA-18 | shogun: puffin444 master * rafc77dc / (2 files in 2 dirs): Put gaussian process regression demo in proper folder - https://github.com/shogun-toolbox/shogun/commit/afc77dc3c6c54ca8f210b6546117c89675040a64 | 23:27 |
CIA-18 | shogun: Sergey Lisitsyn master * r13785f4 / (2 files in 2 dirs): Merge pull request #689 from puffin444/master - https://github.com/shogun-toolbox/shogun/commit/13785f408b3d1d4cfe3e37a5f269703617d96aab | 23:27 |
shogun-buildbot_ | build #252 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/252 blamelist: puffin444 <walke434@msu.edu> | 23:56 |
shogun-buildbot_ | build #209 of deb2 - static_interfaces is complete: Failure [failed test cmdline_static] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/209 blamelist: puffin444 <walke434@msu.edu>, Sergey Lisitsyn <lisitsyn.s.o@gmail.com> | 23:58 |
--- Log closed Sat Aug 04 00:00:17 2012 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!