--- Log opened Sat Apr 09 00:00:36 2011 | ||
blackburn1 | that's the reason why I have a different picture in doc | 00:00 |
---|---|---|
@sonney2k | give me a sec | 00:03 |
blackburn1 | okay, no problem :) | 00:03 |
* sonney2k is removing libshogun/classifier/boosting | 00:03 | |
-!- blackburn1 is now known as blackburn | 00:05 | |
blackburn | sonney2k: I removed it locally, it is working now | 00:05 |
CIA-31 | shogun: Soeren Sonnenburg master * r9bb3b2e / (229 files in 13 dirs): | 00:14 |
CIA-31 | shogun: remove boosting directory | 00:14 |
CIA-31 | shogun: If we merge again we have to merge with newer | 00:14 |
CIA-31 | shogun: multiboost algorithms anyway. The current ones show many warnings and | 00:14 |
CIA-31 | shogun: cause errors when generating the documentation. - http://bit.ly/hgDnvL | 00:14 |
CIA-31 | shogun: Soeren Sonnenburg master * rd7503c3 / src/modular/Classifier.i : remove also includes of boosting algos for now. - http://bit.ly/eoBd4C | 00:14 |
@sonney2k | blackburn, OK I've removed it for now | 00:16 |
@sonney2k | we have a branch for that merge anyways so does not matter yet | 00:16 |
@sonney2k | and should fix our problems | 00:16 |
-!- Mengyun [~Mengyun@S0106001a803d4363.vc.shawcable.net] has quit [Remote host closed the connection] | 00:20 | |
blackburn | sonney2k: I have another problem :) | 00:22 |
@bettyboo | ho | 00:22 |
@sonney2k | ? | 00:22 |
blackburn | sonney2k: these pictures of equations are enumerated | 00:22 |
@sonney2k | rm -rf html | 00:22 |
@sonney2k | and do again | 00:23 |
@sonney2k | make doc | 00:23 |
blackburn | should I send doc/html entirely | 00:23 |
blackburn | ? | 00:23 |
@sonney2k | no ignore doc completely | 00:23 |
blackburn | except doc for wave kernel? | 00:23 |
blackburn | I not pretty sure that understand you right | 00:24 |
blackburn | when sending a patch introducing that kernel, should I add to that commit .html and .png? | 00:25 |
@sonney2k | blackburn, no | 00:25 |
@sonney2k | we generate the docuemntation from the source code | 00:25 |
blackburn | I understand that | 00:26 |
@sonney2k | so doc/html is generated from the headers | 00:26 |
blackburn | not understand why I messed with documentation :) | 00:26 |
@sonney2k | so don't send anything in doc | 00:26 |
blackburn | I'm a bit stupid today :D | 00:26 |
blackburn | sonney2k: is it better to send a patch to ML? I could send it straightly to you | 00:29 |
@sonney2k | blackburn, it is always better to do things in public - other people can comment and also they see that you are active | 00:30 |
@sonney2k | don't be afraid / ashamed | 00:30 |
blackburn | ah, I see | 00:30 |
@sonney2k | you are doing a good job so there is no need for such feelings | 00:31 |
blackburn | patch is ready, going to test it last time | 00:32 |
blackburn | sonney2k: all the ML is filled by 'I have successfully installed shogun' :D | 00:33 |
@sonney2k | at least :) | 00:34 |
@bettyboo | yeah | 00:34 |
@sonney2k | so no excuses to not get it to work | 00:34 |
blackburn | sonney2k: I could translate installation guide or etc to russian, should I? | 00:35 |
@sonney2k | blackburn, but that is a lot of work - would be cool though | 00:36 |
* sonney2k wonders what hidden messages blackburn might include. | 00:36 | |
blackburn | what kind of? | 00:37 |
blackburn | http://www.shogun-toolbox.org/doc/ but there is no much information | 00:38 |
blackburn | I mean translating 'Installation Screenshots Tutorial Examples Implemented Methods Interfaces Frequently Asked Questions Authors License' | 00:38 |
blackburn | and home for doc | 00:38 |
blackburn | sonney2k: I could add a comment to my proposal with that offer :) | 00:40 |
@bettyboo | lol ;> | 00:40 |
@sonney2k | http://www.shogun-toolbox.org/doc_cn/python_examples.html | 00:42 |
@sonney2k | it looks so ... just great | 00:42 |
@sonney2k | blackburn, it would definitely be cool! | 00:42 |
@sonney2k | I cannot proof-read it though :) | 00:43 |
blackburn | sonney2k: If I'll go to GSoC I hope to stay within team, so I have no need to write some trash :) | 00:44 |
blackburn | like I sent today :D | 00:44 |
@sonney2k | heh | 00:44 |
@sonney2k | don't worry... | 00:44 |
blackburn | btw, new patch is here | 00:45 |
blackburn | sonney2k: is it okay to make some addition to my proposal? | 00:47 |
blackburn | in comments I mean | 00:47 |
@sonney2k | yes sure | 00:48 |
CIA-31 | shogun: Sergey Lisitsyn master * r9d96f57 / (4 files in 2 dirs): Introduced new kernel: WaveKernel - http://bit.ly/dThciv | 00:50 |
blackburn | sonney2k: what's with another one? :) | 00:53 |
CIA-31 | shogun: Sergey Lisitsyn master * r33d7b29 / (2 files): Made HistogramIntersectionKernel generalized with introducing parameter beta - http://bit.ly/fsmaI3 | 00:56 |
@sonney2k | blackburn, thanks for the patches :) | 00:56 |
blackburn | ah, sorry :) | 00:56 |
@bettyboo | rahaha ;> | 00:57 |
blackburn | sonney2k: there is a small typo in wave kernel description | 00:57 |
blackburn | K(x,x') = \frac{\theta}{\| x-y \|} \sin \frac{\| x-y \|}{\theta} | 00:57 |
blackburn | my fault | 00:57 |
blackburn | but sending commit with it could be a smallest commit ever seen :) | 00:57 |
@sonney2k | blackburn, I suggest you start using github... fork shogun there and fix it then send a pull request ... | 00:58 |
@sonney2k | or send a patch again | 00:58 |
blackburn | sonney2k: I will send you last patch in a 5 minute, okay? and tomorrow will fork shogun, etc | 00:59 |
@sonney2k | no rush | 00:59 |
@sonney2k | thanks for your contributions! | 00:59 |
blackburn | it is a cauchy kernel, very similar to wave | 00:59 |
blackburn | a bit copypaste | 01:00 |
blackburn | sonney2k: I don't worth thanks :) | 01:01 |
@bettyboo | hihi ;D | 01:01 |
serialhex | sonney2k: are you still awake?? (it's like 1am there) | 01:07 |
@sonney2k | serialhex, I have to do a review... | 01:08 |
serialhex | sonney2k: ahh... ok. i was just wondering if you have the exponential or log kernels hiding somewhere not on the methods page. otherwise i'm going to implement them (and possibly more) | 01:09 |
@sonney2k | serialhex, the exponential / gaussian kernel - in fact kernel 3-5 are too similar to justify different names | 01:11 |
@sonney2k | so yes they are there | 01:11 |
serialhex | ahh, ok | 01:11 |
@sonney2k | serialhex, actually not 100% true | 01:11 |
@sonney2k | the exponential kernel would make sense ... | 01:12 |
serialhex | ok, can do... i'll do that and the bayesian one too probably | 01:12 |
* serialhex gets to werk | 01:12 | |
@sonney2k | one could potentially modify the gaussian kernel to be 'exponential' | 01:13 |
blackburn | sonney2k: I'll send a last patch using mail list in a two minutes, okay? | 01:14 |
serialhex | ok, i'll take a look at that | 01:14 |
@sonney2k | blackburn, thanks - but no rush. I really have to do that review now. | 01:14 |
blackburn | sonney2k: ah, sorry | 01:14 |
blackburn | just trying get things done and go sleep :) | 01:15 |
serialhex | yeah, the exp looks REALLY easy, i'll get that one & probably a few more sonney2k | 01:15 |
@sonney2k | blackburn, ok | 01:16 |
blackburn | and at least have to fed my bear, drink vodka and go to KGB :D | 01:17 |
@sonney2k | like all Russians do presumably ;-) | 01:18 |
@bettyboo | ;D | 01:18 |
@sonney2k | well done betty! | 01:18 |
@bettyboo | sonney2k: well I have done for matlab interface only | 01:18 |
serialhex | she's good! | 01:18 |
@sonney2k | I hope knrrrd will tell me about her super secret algorithms | 01:18 |
brocolli | started sifting through apps yet sonney2k? ;) | 01:25 |
@sonney2k | brocolli, no doing a review | 01:25 |
brocolli | nice, journal? | 01:25 |
brocolli | or AAAI | 01:26 |
blackburn | oh, it's working right. have to drink vodka now | 01:31 |
blackburn | sonney2k: I answered you in ML with another one patch. it's the last :) won't disturb you no more | 01:31 |
@bettyboo | ;> | 01:31 |
blackburn | sorry for being annoying | 01:32 |
@sonney2k | brocolli, ICML still | 01:33 |
brocolli | Noice! | 01:33 |
-!- aifargonos [~aifargono@46.18.27.35] has left #shogun [] | 01:34 | |
@sonney2k | blackburn, next time please split the patch into two commits | 01:35 |
@sonney2k | one fixing the typo and one introducing the new kernel | 01:35 |
blackburn | sonney2k: sorry, it was so small :) I may do it now | 01:35 |
@bettyboo | yep | 01:35 |
@sonney2k | blackburn, yes sure but this is just wrong with git (ask yoh about this - he will lecture you :) | 01:36 |
blackburn | sonney2k: okay :) will do it next time | 01:37 |
@bettyboo | funny ;D | 01:37 |
CIA-31 | shogun: Sergey Lisitsyn master * r4fed492 / (5 files in 2 dirs): Introduced new kernel: CauchyKernel, corrected misprint at WaveKernel description - http://bit.ly/eUUGN7 | 01:37 |
blackburn | sonney2k: thank you | 01:38 |
@sonney2k | blackburn, there is one more thing - serialization support for your kernels - have a look at gaussian kernel: there is a m_paramters->add line ... you also need this for your new kernels to save parameters | 01:38 |
@sonney2k | blackburn, this way you can pickle the kernel from python | 01:38 |
blackburn | sonney2k: okay, I will do | 01:38 |
@sonney2k | don't do it now but at some point OK | 01:38 |
blackburn | i have 3.39 AM here, so will not do now :) | 01:39 |
@sonney2k | well thanks and have a good sleep then! | 01:40 |
blackburn | I think not going sleep: have to wake up at 7-40 :)) | 01:40 |
blackburn | sonney2k: later I'll modify CKNN and will do that serialization support | 01:40 |
blackburn | hope there will not take place a 'commit race' :) | 01:41 |
@bettyboo | HA ;> | 01:41 |
-!- ameerkat [~ameerkat@184-98-140-155.phnx.qwest.net] has joined #shogun | 01:44 | |
* sonney2k has finished his review | 01:45 | |
@sonney2k | and hurray off to bed | 01:45 |
@bettyboo | nite | 01:45 |
blackburn | good night, Sören | 01:46 |
blackburn | good night, everyone :) | 01:46 |
@bettyboo | funny :> | 01:46 |
blackburn | bettyboo: you too | 01:47 |
@bettyboo | blackburn: Then there's Learning Kernel Classifiers by the great Ralf Herbrich, but it's too rigorous for some. | 01:47 |
-!- blackburn [~qdrgsm@188.168.4.21] has quit [Quit: Leaving.] | 01:47 | |
-!- alesis-novik [~alesis@188.74.87.84] has joined #shogun | 02:02 | |
serialhex | brocolli: you awake? | 02:48 |
serialhex | guess not... | 02:50 |
-!- Ziyuan [~Ziyuan@116.22.39.134] has quit [Ping timeout: 260 seconds] | 03:14 | |
-!- Ziyuan [~Ziyuan@116.22.50.70] has joined #shogun | 03:15 | |
-!- Ziyuan [~Ziyuan@116.22.50.70] has quit [] | 06:11 | |
-!- hiteshk [~hitesh_na@117.211.88.150] has joined #shogun | 06:31 | |
-!- siddharth [~siddharth@117.211.88.150] has quit [Ping timeout: 240 seconds] | 06:34 | |
-!- ameerkat [~ameerkat@184-98-140-155.phnx.qwest.net] has quit [Read error: Connection reset by peer] | 06:45 | |
-!- ameerkat [~ameerkat@184-98-140-155.phnx.qwest.net] has joined #shogun | 06:46 | |
-!- hiteshk [~hitesh_na@117.211.88.150] has left #shogun [] | 07:02 | |
-!- siddharth [~siddharth@117.211.88.150] has joined #shogun | 07:43 | |
-!- siddharth [~siddharth@117.211.88.150] has quit [Read error: Connection reset by peer] | 07:50 | |
-!- siddharth [~siddharth@117.211.88.150] has joined #shogun | 07:52 | |
serialhex | bettyboo: hey, how are you? | 08:14 |
@bettyboo | serialhex: so it is hello, how are you? | 08:14 |
serialhex | bettyboo: i am good... kind of tired though, yourself? | 08:14 |
@bettyboo | serialhex: that is, a basis vector is only associated with ONE example of the training data | 08:14 |
serialhex | bettyboo: that is very true, but why? | 08:15 |
@bettyboo | serialhex: it is generated automatically ... build first | 08:16 |
@bettyboo | nothing | 08:16 |
serialhex | nothing? really bettyboo | 08:16 |
@bettyboo | serialhex: Vapnik is pretty old-school research, the fundamentals of SVMs can be found even in his 1970's books | 08:16 |
serialhex | i know that bettyboo, tell me something i dont know | 08:17 |
@bettyboo | serialhex: I don't know what that bot is telling all the time.. :) | 08:17 |
@bettyboo | serialhex, never mind ;> | 08:17 |
serialhex | now THAT i agree with :P | 08:17 |
@bettyboo | ;D serialhex | 08:17 |
serialhex | :D bettyboo | 08:17 |
@bettyboo | serialhex: speedups for the special case where kernel is linear or a string kernel | 08:17 |
serialhex | bettyboo: nice | 08:18 |
@bettyboo | serialhex: Don't forget nice people like Chris M. Bishop | 08:18 |
serialhex | yeah, he's alright even though he works at microsoft | 08:19 |
serialhex | do you like microsoft bettyboo | 08:19 |
@bettyboo | serialhex: I don't worth thanks :) | 08:19 |
serialhex | i dont work either bettyboo | 08:21 |
@bettyboo | serialhex: it works well in my machine now | 08:22 |
serialhex | what do you do for money?? bettyboo?? | 08:22 |
@bettyboo | serialhex: i think features can not handle this case. | 08:22 |
serialhex | no, maybe not bettyboo | 08:22 |
@bettyboo | serialhex: i would imagine so... if you have 100 dimensions & not all of them are needed, then you'll have to sift through a bunch of 100 dimensional arrays all over the place... way slow!! | 08:22 |
-!- aifargonos [~aifargono@46.18.27.35] has joined #shogun | 09:29 | |
-!- hiteshk [~hitesh_na@117.211.88.150] has joined #shogun | 09:36 | |
-!- hiteshk [~hitesh_na@117.211.88.150] has left #shogun [] | 09:38 | |
-!- epps [~epps@unaffiliated/epps] has joined #shogun | 09:56 | |
-!- ameerkat [~ameerkat@184-98-140-155.phnx.qwest.net] has quit [Ping timeout: 258 seconds] | 12:09 | |
-!- dvevre [b49531e3@gateway/web/freenode/ip.180.149.49.227] has joined #shogun | 13:17 | |
-!- Tanmoy [75d35896@gateway/web/freenode/ip.117.211.88.150] has quit [Ping timeout: 252 seconds] | 13:20 | |
-!- epps [~epps@unaffiliated/epps] has quit [Ping timeout: 252 seconds] | 13:29 | |
-!- oiwah [~oiwah@i114-181-177-143.s04.a013.ap.plala.or.jp] has joined #shogun | 13:31 | |
-!- oiwah [~oiwah@i114-181-177-143.s04.a013.ap.plala.or.jp] has quit [Client Quit] | 13:34 | |
siddharth | sonney2k, can GPL copyright header be embedded automatically? | 14:29 |
-!- epps [~epps@84.18.157.200] has joined #shogun | 14:33 | |
-!- epps [~epps@84.18.157.200] has quit [Changing host] | 14:33 | |
-!- epps [~epps@unaffiliated/epps] has joined #shogun | 14:33 | |
-!- siddharth [~siddharth@117.211.88.150] has quit [Remote host closed the connection] | 14:55 | |
-!- siddharth [~siddharth@117.211.88.150] has joined #shogun | 14:57 | |
@sonney2k | siddharth, hmmhh I always copy & paste it | 15:01 |
siddharth | sonney2k, ok..Also I forked shogun today...so after some code changes I should commit and then pull request right? | 15:03 |
siddharth | instead of sending patch | 15:04 |
@sonney2k | siddharth, yes | 15:05 |
CIA-31 | shogun: asia master * r385ae97 / (2 files): Introduced new kernel: Multiquadric Kernel (+8 more commits...) - http://bit.ly/gwqxOK | 15:25 |
-!- blackburn [~qdrgsm@188.122.237.148] has joined #shogun | 15:58 | |
blackburn | hello there | 15:59 |
@bettyboo | blackburn, ho | 15:59 |
-!- siddharth [~siddharth@117.211.88.150] has quit [Read error: Connection reset by peer] | 16:05 | |
-!- siddharth [~siddharth@117.211.88.150] has joined #shogun | 16:22 | |
-!- Ziyuan [~Ziyuan@116.22.50.70] has joined #shogun | 16:33 | |
Ziyuan | How to assign EKernelType, | 16:34 |
Ziyuan | If we implement new kernels independently. | 16:34 |
josip | Hello, I have a "structured" data set. In the sense that, I have several subjects and each subject has multiple patterns (actually mamography images). I want to do cross-validation, but teh split has to be done on patients, not on the data points. Is this possible in Shogun? I have this done with some scripts, but I'd love to get rid off thhem | 16:35 |
@bettyboo | cu | 16:35 |
josip | not images, but features extracted from candidate regions of the mamography images | 16:35 |
serialhex | sonney2k: hey, i'm working on getting that patch fixed up... it's being a PITA though :P | 16:47 |
siddharth | Ziyuan, change in the Kernel.h file | 16:54 |
Ziyuan | I know | 16:55 |
Ziyuan | But I mean, how to avoid collision | 16:55 |
Ziyuan | or to organize reasonably | 16:56 |
siddharth | Ziyuan, do git show...this will allow you to see current patches applied...then accordingly make the changes | 17:01 |
Ziyuan | OKOK... | 17:20 |
-!- epps [~epps@unaffiliated/epps] has quit [Read error: Connection reset by peer] | 18:03 | |
-!- epps [~epps@unaffiliated/epps] has joined #shogun | 18:03 | |
-!- argetlahmsource [~argetlahm@adsl-99-160-142-89.dsl.euclwi.sbcglobal.net] has joined #shogun | 18:36 | |
-!- blackburn1 [~qdrgsm@188.122.237.148] has joined #shogun | 19:16 | |
-!- blackburn [~qdrgsm@188.122.237.148] has quit [Read error: No route to host] | 19:16 | |
-!- ameerkat [~ameerkat@184-98-140-155.phnx.qwest.net] has joined #shogun | 19:17 | |
-!- akhil_ [75d35896@gateway/web/freenode/ip.117.211.88.150] has joined #shogun | 19:19 | |
-!- akhil_ [75d35896@gateway/web/freenode/ip.117.211.88.150] has quit [Quit: Page closed] | 19:28 | |
-!- siddharth [~siddharth@117.211.88.150] has quit [Ping timeout: 252 seconds] | 19:33 | |
-!- siddharth [~siddharth@117.211.88.150] has joined #shogun | 19:36 | |
-!- Ryaether [~Ryaether@50-80-170-245.client.mchsi.com] has joined #shogun | 20:26 | |
Ziyuan | Hmm...I installed libatlas-dev and libatlas-base-dev but fail to get Atlas support. | 20:43 |
Ziyuan | And atlas3-headers/atlas3-base-dev cannot be found. | 20:43 |
-!- yayo3 [542a925d@gateway/web/freenode/ip.84.42.146.93] has joined #shogun | 20:48 | |
yayo3 | hi | 20:48 |
Ziyuan | I am a newbie of developing under linux so please don't mind my asking unwise questions... | 20:48 |
Ziyuan | hi | 20:48 |
siddharth | sonney2k, how to correct whitespace errors while applying patch? | 20:58 |
siddharth | should we ignore them? | 20:58 |
siddharth | git config core.whitespace=nowarn | 20:58 |
@sonney2k | siddharth, I prefer fixing them | 21:04 |
@sonney2k | I use vim for editing so I search for ' $' and erase the spaces | 21:04 |
@sonney2k | I guess there are scripts doing that though | 21:05 |
@sonney2k | Ziyuan, please force enable lapack in shogun configure (--enable-lapack IIRC) | 21:05 |
@sonney2k | then you should get some more meaningful error msg | 21:05 |
blackburn1 | sonney2k: if you have much time please take a look at pull requests, it seems to be ready | 21:06 |
-!- blackburn1 is now known as blackburn | 21:10 | |
Ziyuan | OK, thank you sonney | 21:10 |
Ziyuan | More: When I make the stuff in examples/documented/libshogun, I've got many "undefined reference" errors. | 21:11 |
@sonney2k | Ziyuan, that shouldn't happen | 21:18 |
@sonney2k | is the include path / link path perhaps wrong | 21:18 |
CIA-31 | shogun: Justin Patera master * redb5c2c / src/libshogun/kernel/ExponentialKernel.cpp : Fixing serializattion stuff in ExponentialKernel (and learning git the hard way :P ) (+7 more commits...) - http://bit.ly/i1d3QK | 21:21 |
serialhex | YAY!!! i've been commited!! :P | 21:23 |
@bettyboo | ;> serialhex | 21:23 |
serialhex | we should celebrate bettyboo!! what do you want?? vodka? yager? beer?? | 21:24 |
@bettyboo | serialhex: make doc | 21:24 |
serialhex | i dont know how to make a doc bettyboo, tell me how to make one | 21:24 |
@bettyboo | serialhex: to show changes we have to do make install right? | 21:24 |
serialhex | there you go again... always talking about your ML stuff... i dont think you ever really listen to me bettyboo! | 21:25 |
@bettyboo | serialhex: is enjoying the Berlin sun - dont you think that is fair? | 21:25 |
serialhex | :P i'm enjoying the sun, just not in berlin | 21:26 |
Ziyuan | I am checking | 21:26 |
blackburn | (11:24:34 PM) bettyboo: serialhex: make doc | 21:27 |
@bettyboo | blackburn: I don't know what implies this way | 21:27 |
blackburn | oh, it made my day :D | 21:27 |
blackburn | vodka? beer? make doc! | 21:27 |
serialhex | :P | 21:27 |
serialhex | yeah... she's cool! | 21:27 |
-!- Tanmoy [75d35896@gateway/web/freenode/ip.117.211.88.150] has joined #shogun | 21:30 | |
-!- Ziyuan [~Ziyuan@116.22.50.70] has quit [Remote host closed the connection] | 21:30 | |
-!- _Ziyuan [~Ziyuan@116.22.50.70] has joined #shogun | 21:31 | |
@sonney2k | blackburn, I sent you a comment to your patch via github | 21:32 |
blackburn | sonney2k: I see, will do a fix in a minute | 21:32 |
@sonney2k | blackburn, I haven't used this so far - did you get an email reply | 21:33 |
blackburn | "shogun-toolbox кому: мне", yeap | 21:33 |
blackburn | I received an email from github | 21:34 |
@sonney2k | very good | 21:36 |
blackburn | sonney2k: should getters and setters will be 'set_q_param' just like 'q_param'? | 21:36 |
@sonney2k | blackburn, I would just name the member varialbe m_q | 21:36 |
@sonney2k | and that's it | 21:36 |
blackburn | okay | 21:36 |
blackburn | set_q will stay set_q, right? | 21:36 |
@sonney2k | so we know that this is from the class | 21:36 |
@sonney2k | yes | 21:36 |
@sonney2k | and you can still use q | 21:36 |
blackburn | may be do the same with k? | 21:37 |
-!- epps [~epps@unaffiliated/epps] has quit [Ping timeout: 248 seconds] | 21:37 | |
@sonney2k | yes then | 21:37 |
blackburn | sonney2k: done | 21:42 |
@sonney2k | blackburn, one last one line patch | 21:46 |
blackburn | okay | 21:46 |
@sonney2k | k_ -> k :) | 21:46 |
@sonney2k | then I cannot find anything further :) | 21:46 |
@bettyboo | sonney2k, ;D | 21:46 |
blackburn | the greatest one | 21:46 |
blackburn | will be a milestone for shogun toolbox | 21:46 |
@sonney2k | blackburn, thanks for all your work | 21:47 |
blackburn | sonney2k: no problem, thank you :) | 21:48 |
@bettyboo | ;D | 21:48 |
blackburn | sonney2k: and another fix in a minute. m_q(1.0) should be before train_labels | 21:52 |
@sonney2k | ok | 21:52 |
blackburn | ./classifier/KNN.h:191: warning: ‘shogun::CKNN::train_labels’ will be initialized after | 21:52 |
blackburn | ./classifier/KNN.h:179: warning: ‘float64_t shogun::CKNN::m_q’ | 21:52 |
blackburn | classifier/KNN.cpp:20: warning: when initialized here | 21:52 |
blackburn | if I understand it right | 21:52 |
dvevre | hey all. i've a basic question.. is there any reason for not adding 'width' through m_parameters->add in DistanceKernel? | 21:54 |
@bettyboo | moin dvevre | 21:54 |
blackburn | dvevre: I think it should be there | 21:55 |
blackburn | dvevre: I did it in WaveKernel and CauchyKernel which are very similar to DistanceKernel | 21:55 |
dvevre | blackburn: yeah. | 21:55 |
dvevre | i checked, but it hasn't been added that way | 21:55 |
dvevre | so i was wondering if there was some particular reason why | 21:55 |
blackburn | sonney2k: now no warnings, seems that it is ready | 21:56 |
dvevre | bettyboo: hello! i had to wiki what 'moin' meant :) | 21:58 |
@bettyboo | dvevre: it would really make sense to do the octave / and older matlab binary formats too... for that look also at the source of octave (it's linked to at the gsoc page) | 21:58 |
dvevre | bettyboo: right. i'll take a look right away | 21:59 |
@bettyboo | dvevre: if you have much time please take a look at pull requests, it seems to be ready | 21:59 |
@bettyboo | *oink* ;> | 21:59 |
blackburn | oh, my sentence! | 21:59 |
CIA-31 | shogun: Sergey Lisitsyn master * rde4b2cc / (2 files): Fixed some naming - http://bit.ly/fDO8lM | 22:01 |
CIA-31 | shogun: Sergey Lisitsyn master * r02dab1c / src/libshogun/classifier/KNN.cpp : Changed initialization order to avoid warnings. - http://bit.ly/fX553y | 22:01 |
CIA-31 | shogun: Sergey Lisitsyn master * ra9264fb / src/libshogun/classifier/KNN.cpp : Removed _ from k_. (Commit the Great) - http://bit.ly/hd87eK | 22:01 |
CIA-31 | shogun: Soeren Sonnenburg master * r6c3339d / (6 files in 2 dirs): Merge branch 'master' of https://github.com/SergeyLisitsyn/shogun into SergeyLisitsyn-master - http://bit.ly/gq7tbR | 22:01 |
blackburn | oh no | 22:01 |
blackburn | *thought that all 10 will be here* | 22:01 |
blackburn | I really like Commit the Great | 22:02 |
@sonney2k | blackburn, thanks again - yes CIA-31 is cool | 22:02 |
blackburn | sonney2k: thank you for watching mistakes :) | 22:03 |
@sonney2k | dvevre, it is a bug | 22:03 |
dvevre | sonney2k: should i modify it? | 22:04 |
@sonney2k | dvevre, serialization is rather new so please send a patch | 22:04 |
@sonney2k | very welcome | 22:04 |
dvevre | sonney2k: sure :) | 22:04 |
@bettyboo | he | 22:04 |
@sonney2k | blackburn, I am overlooking many small mistakes - oh well thanks for being patient wit me | 22:04 |
blackburn | sonney2k: it's better for both of us :) | 22:05 |
@bettyboo | ;> | 22:05 |
-!- siddharth [~siddharth@117.211.88.150] has quit [Remote host closed the connection] | 22:12 | |
-!- epps [~epps@unaffiliated/epps] has joined #shogun | 22:12 | |
-!- siddharth [~siddharth@117.211.88.150] has joined #shogun | 22:16 | |
* sonney2k loses track of all the patches | 22:32 | |
@sonney2k | I guess doing it all via github would drastically simplify things for me | 22:32 |
blackburn | sonney2k: you could say it in ml :) | 22:33 |
@bettyboo | ;> | 22:33 |
-!- dvevre [b49531e3@gateway/web/freenode/ip.180.149.49.227] has quit [Ping timeout: 252 seconds] | 22:33 | |
@sonney2k | blackburn, how do you do a pull request? | 22:33 |
blackburn | hm.. how detailly should I describe it? | 22:33 |
blackburn | ah! http://help.github.com/fork-a-repo/ | 22:34 |
blackburn | here it is | 22:34 |
-!- epps [~epps@unaffiliated/epps] has quit [Ping timeout: 248 seconds] | 22:35 | |
blackburn | sonney2k: it is sufficient to fork repo and configure remote for upstream | 22:36 |
@sonney2k | blackburn, and then how do you do a pull request such that I can see it? | 22:36 |
-!- dvevre [b49531e3@gateway/web/freenode/ip.180.149.49.227] has joined #shogun | 22:37 | |
blackburn | sonney2k: it is on personal fork's page, there is a 'pull request' button | 22:37 |
-!- epps [~epps@unaffiliated/epps] has joined #shogun | 22:37 | |
blackburn | as I understand, there is no way to do it via git, so only via github.com | 22:37 |
blackburn | sonney2k: you can find about it here: http://help.github.com/pull-requests/ | 22:38 |
@sonney2k | blackburn, ahh great - I will write sth. to the ml | 22:39 |
-!- _Ziyuan is now known as Ziyuan | 22:40 | |
blackburn | sonney2k: mention there about 'git fetch upstream' and 'git merge upstream' | 22:41 |
blackburn | sonney2k: with it there won't be a problem with ~20 patches in pull request :) | 22:42 |
@bettyboo | ;> | 22:42 |
blackburn | like it was | 22:42 |
blackburn | sonney2k: seems that to demarcate different pulls we should use branches | 22:48 |
-!- dvevre [b49531e3@gateway/web/freenode/ip.180.149.49.227] has quit [Ping timeout: 251 seconds] | 22:48 | |
blackburn | i.e. one pull request for some kernel and another for some classifier | 22:49 |
-!- dvevre [b49531e3@gateway/web/freenode/ip.180.149.49.227] has joined #shogun | 22:51 | |
@sonney2k | blackburn, yes - but most people just do one change - so it is ok | 22:56 |
blackburn | sonney2k: I'll answer you in ML about 'git remote add upstream' | 23:03 |
blackburn | sonney2k: answered in ML with some addition | 23:11 |
-!- yayo3 [542a925d@gateway/web/freenode/ip.84.42.146.93] has quit [Quit: Page closed] | 23:25 | |
@sonney2k | thanks | 23:25 |
-!- yayo3 [~jakub@ip-78-45-113-245.net.upcbroadband.cz] has joined #shogun | 23:37 | |
CIA-31 | shogun: Soeren Sonnenburg master * r15b28f1 / (42 files in 7 dirs): | 23:43 |
CIA-31 | shogun: get_name needs to return the class name (withough C prefix) | 23:43 |
CIA-31 | shogun: This patch goes through all *.h files changing get_name to return the | 23:43 |
CIA-31 | shogun: class name of the of the class withouth the C prefix. Since Array*.h | 23:43 |
CIA-31 | shogun: contained functions to set the array name, the array set_name and | 23:43 |
CIA-31 | shogun: get_name functions have been renamed to set_array_name / get_array_name | 23:43 |
CIA-31 | shogun: to fix the naming clash. - http://bit.ly/ikr2zl | 23:43 |
CIA-31 | shogun: Marcin Copik master * ra1b95d5 / src/libshogun/lib/File.cpp : Added implementation of functions for I/O of bool matrices and string lists - http://bit.ly/gRWLxu | 23:43 |
CIA-31 | shogun: Soeren Sonnenburg master * r461e05d / (2 files): Merge branch 'master' of https://github.com/frx/shogun - http://bit.ly/dUD1es | 23:45 |
-!- yoh [~yoh@washoe.dartmouth.edu] has quit [Ping timeout: 248 seconds] | 23:54 | |
yayo3 | so sonney2k, you want us to resend the already sent patches through github? | 23:54 |
@sonney2k | yayo3, errm not the ones that are already applied of course - only future patches / the ones that have not yet been applied | 23:55 |
yayo3 | ok | 23:56 |
--- Log closed Sun Apr 10 00:00:36 2011 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!