| --- Log opened Fri May 18 00:00:40 2012 | ||
| -!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 00:08 | |
| n4nd0 | sonney2k, blackburn : I have seen in the logs you have been discussing about the labels hierarchy | 00:29 |
|---|---|---|
| blackburn | yeah | 00:29 |
| n4nd0 | so what is the conclusion? | 00:29 |
| n4nd0 | will we use a base class for labels? | 00:29 |
| n4nd0 | or? | 00:30 |
| blackburn | heh no idea lets see what is he will come with | 00:30 |
| n4nd0 | all right | 00:30 |
| n4nd0 | I'll keep an eye on github | 00:30 |
| -!- blackburn [~blackburn@188.122.250.167] has quit [Quit: Leaving.] | 00:48 | |
| -!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving] | 01:31 | |
| -!- av3ngr [av3ngr@nat/redhat/x-scprkyesygrtlzzt] has joined #shogun | 01:32 | |
| -!- av3ngr [av3ngr@nat/redhat/x-scprkyesygrtlzzt] has quit [Read error: Connection reset by peer] | 01:46 | |
| -!- av3ngr [av3ngr@nat/redhat/x-gckdfbduswynkgua] has joined #shogun | 01:46 | |
| -!- av3ngr [av3ngr@nat/redhat/x-gckdfbduswynkgua] has quit [Read error: Connection reset by peer] | 01:50 | |
| -!- abn_ [av3ngr@nat/redhat/x-siwxyjkjituyottb] has joined #shogun | 01:51 | |
| -!- av3ngr [av3ngr@nat/redhat/x-wgnhnksekfhutuhg] has joined #shogun | 02:16 | |
| -!- abn_ [av3ngr@nat/redhat/x-siwxyjkjituyottb] has quit [Ping timeout: 256 seconds] | 02:20 | |
| -!- abn_ [av3ngr@nat/redhat/x-edpkkfhhuuvrgzaj] has joined #shogun | 02:22 | |
| -!- av3ngr [av3ngr@nat/redhat/x-wgnhnksekfhutuhg] has quit [Ping timeout: 244 seconds] | 02:25 | |
| -!- Francis_Chan [~Adium@59.108.60.58] has joined #shogun | 02:40 | |
| -!- Francis_Chan [~Adium@59.108.60.58] has quit [Ping timeout: 256 seconds] | 03:31 | |
| -!- emrecelikten [~Anubis@82.222.148.100] has quit [Quit: Leaving.] | 03:50 | |
| -!- Francis_Chan [~Adium@59.108.60.58] has joined #shogun | 03:56 | |
| -!- abn__ [av3ngr@nat/redhat/x-ghucyoueatuuaupd] has joined #shogun | 04:05 | |
| -!- abn_ [av3ngr@nat/redhat/x-edpkkfhhuuvrgzaj] has quit [Ping timeout: 244 seconds] | 04:08 | |
| -!- abn__ [av3ngr@nat/redhat/x-ghucyoueatuuaupd] has quit [Ping timeout: 245 seconds] | 04:59 | |
| -!- wiking_ [~wiking@c-98-231-171-175.hsd1.md.comcast.net] has joined #shogun | 05:34 | |
| -!- wiking_ [~wiking@c-98-231-171-175.hsd1.md.comcast.net] has quit [Changing host] | 05:34 | |
| -!- wiking_ [~wiking@huwico/staff/wiking] has joined #shogun | 05:34 | |
| -!- wiking [~wiking@huwico/staff/wiking] has quit [Ping timeout: 272 seconds] | 05:36 | |
| -!- wiking_ is now known as wiking | 05:36 | |
| -!- wiking [~wiking@huwico/staff/wiking] has quit [Ping timeout: 245 seconds] | 05:46 | |
| -!- emrecelikten [~emrecelik@82.222.148.100] has joined #shogun | 07:04 | |
| -!- emrecelikten [~emrecelik@82.222.148.100] has quit [Client Quit] | 07:05 | |
| -!- vikram360 [~vikram360@117.192.186.253] has quit [Ping timeout: 272 seconds] | 07:57 | |
| -!- vikram360 [~vikram360@117.192.176.104] has joined #shogun | 08:21 | |
| -!- uricamic [~uricamic@2001:718:2:1634:1dd1:edc6:9e87:3324] has joined #shogun | 09:07 | |
| -!- blackburn [~blackburn@188.122.250.167] has joined #shogun | 09:21 | |
| CIA-113 | shogun: iglesias master * r2e96c0f / src/shogun/lib/SGMatrix.h : * minor indent fix in SGMatrix - http://git.io/UXuYag | 09:25 |
| -!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 09:31 | |
| n4nd0 | hey blackburn | 09:36 |
| n4nd0 | good morning | 09:36 |
| blackburn | hey | 09:37 |
| blackburn | morning | 09:37 |
| blackburn | how is it going? | 09:37 |
| n4nd0 | fine, it's starting to be nice weather regularly around here :) | 09:37 |
| n4nd0 | that makes me happy | 09:37 |
| n4nd0 | what about you? | 09:37 |
| blackburn | did you receive welcome package yet? | 09:37 |
| blackburn | heh I have been happy with weather since late april I think | 09:38 |
| blackburn | there was virtually no spring at all | 09:38 |
| n4nd0 | wow, you're luckier in Russia in that sense that in Sweden | 09:38 |
| blackburn | just baam and summer | 09:38 |
| blackburn | :D | 09:38 |
| n4nd0 | I didn't receive anything yet | 09:38 |
| n4nd0 | you? | 09:38 |
| blackburn | no, but some guys did already according to gsoc ml | 09:39 |
| n4nd0 | neither at home nor on the mail, I guess the ACM comes by mail | 09:39 |
| n4nd0 | yes, it looks like that | 09:39 |
| n4nd0 | they said we shouldn't worry until May 25th right? | 09:39 |
| blackburn | yes | 09:39 |
| n4nd0 | ok | 09:39 |
| n4nd0 | can you believe that I have not gotten the grades for one of my exams yet? | 09:39 |
| n4nd0 | I did it on April man, 28th I think | 09:40 |
| shogun-buildbot | build #871 of cmdline_static is complete: Failure [failed test_1] Build details are at http://www.shogun-toolbox.org/buildbot/builders/cmdline_static/builds/871 blamelist: fernando.iglesiasg@gmail.com | 09:40 |
| blackburn | hahh that takes a while to rate your work damn | 09:40 |
| n4nd0 | haha no way | 09:40 |
| blackburn | why do they do it so sloow? | 09:40 |
| n4nd0 | this teacher is not very good | 09:40 |
| n4nd0 | she just appeared the 50% of the lectures | 09:41 |
| n4nd0 | and half an hour late, on average, the times she did it | 09:41 |
| n4nd0 | anyhow, I think tomorrow it's the last day to publish so they should be ready by then | 09:41 |
| blackburn | hah I had had one phd student who 'teached' us calculus | 09:42 |
| blackburn | just like your teacher - always late | 09:42 |
| n4nd0 | too bad | 09:43 |
| blackburn | we were so angry with him | 09:43 |
| n4nd0 | anyhow | 09:43 |
| blackburn | so we played kind of tennis | 09:43 |
| blackburn | right on lesson | 09:44 |
| n4nd0 | haha, in the class? | 09:44 |
| n4nd0 | wow | 09:44 |
| blackburn | yes with books and paper | 09:44 |
| blackburn | :D | 09:44 |
| shogun-buildbot | build #850 of r_static is complete: Failure [failed test_1] Build details are at http://www.shogun-toolbox.org/buildbot/builders/r_static/builds/850 blamelist: fernando.iglesiasg@gmail.com | 09:44 |
| blackburn | and only after two minutes of play he told us (me and one other guy) to leave the room :D | 09:45 |
| n4nd0 | shogun-buildbot: what did I do to you? | 09:45 |
| n4nd0 | blackburn: do you where I can find an example of use of libqp in shogun? | 09:48 |
| blackburn | n4nd0: it is used in libocas | 09:49 |
| blackburn | only I guess | 09:49 |
| n4nd0 | blackburn: yeah, it seems that it's used there, thanks ;) | 09:50 |
| -!- Francis_Chan [~Adium@59.108.60.58] has left #shogun [] | 10:12 | |
| -!- aphex [2e1fd566@gateway/web/freenode/ip.46.31.213.102] has joined #shogun | 10:43 | |
| aphex | hi all | 10:43 |
| blackburn | hi | 10:43 |
| aphex | I have a question about string kernel implemented in shogun | 10:44 |
| aphex | I would like to use CLocalAlignmentStringKernel on stringfeatures<char> with alphabet RAWBYTE. Is that compatible , | 10:45 |
| aphex | ? | 10:45 |
| blackburn | let me check | 10:46 |
| blackburn | aphex: ok looks like it is suited for amino-acids so no idea whether it would work for rawbyte | 10:49 |
| aphex | That's what I ve seen too. | 11:07 |
| aphex | Do you know what string kernel could be used for RAWBYTE alphabet ? | 11:07 |
| blackburn | iirc WeightedDegreeStringKernel should work on rawbyte | 11:09 |
| aphex | blackburn: tx. Is there an example with this kernel ? | 11:11 |
| blackburn | for which interface? | 11:11 |
| aphex | c++ | 11:12 |
| aphex | blackburn: note: I used string features of different sizes | 11:15 |
| blackburn | well check libshogun examples | 11:16 |
| blackburn | all kernels are pretty similar | 11:16 |
| aphex | blackburn: ok thx | 11:17 |
| -!- aphex [2e1fd566@gateway/web/freenode/ip.46.31.213.102] has quit [Quit: Page closed] | 11:26 | |
| -!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Read error: Operation timed out] | 12:16 | |
| -!- pluskid [~pluskid@111.120.55.156] has joined #shogun | 14:35 | |
| -!- eric_ [2e1fd566@gateway/web/freenode/ip.46.31.213.102] has joined #shogun | 14:36 | |
| eric_ | hi again | 14:36 |
| eric_ | blackburn: There is definitly somthg wrong when using Xvalidation with multiclass svm other than "CMulticlassLibLinear" | 14:37 |
| eric_ | blackburn: CMulticlassLibSVM or CGMNPSVM with same linear kernel config give always XVal: [0.100000,0.100000] with alpha=0.050000, mean=0.100000 | 14:38 |
| -!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 14:53 | |
| n4nd0 | hey pluskid, how is it going? | 15:44 |
| pluskid | n4nd0: hi | 15:45 |
| pluskid | I was busy preparing docs for a VISA interview the last two days | 15:45 |
| pluskid | almost done with ecoclib | 15:46 |
| n4nd0 | nice | 15:46 |
| pluskid | will move to some papers Cheng mentioned to me | 15:46 |
| pluskid | how's it going with SO? | 15:46 |
| n4nd0 | I hope you visa interview goes good | 15:46 |
| n4nd0 | it's going good, I find it fun | 15:46 |
| pluskid | thanks! it will be on May 29 | 15:46 |
| n4nd0 | you are going to US, right? | 15:46 |
| pluskid | n4nd0: cool | 15:47 |
| pluskid | yes, I'm going to the US | 15:47 |
| pluskid | in mid August | 15:47 |
| n4nd0 | awesome | 15:47 |
| n4nd0 | I would really love to go there to do my phd | 15:47 |
| pluskid | have you decided where to get a PhD yet? | 15:48 |
| n4nd0 | no | 15:48 |
| n4nd0 | I expect to finish during June 2013, in about a year | 15:48 |
| pluskid | hmm, I see | 15:48 |
| n4nd0 | when do you think I should start getting in touch with people there and prepare it seriously? | 15:49 |
| pluskid | so you will apply schools this Autumn | 15:49 |
| pluskid | n4nd0: I suggest you start to collect information about schools/profs. in the summer holiday | 15:49 |
| pluskid | that would be a *huge* survey :D | 15:50 |
| pluskid | to get to know who and where you would be interested in | 15:50 |
| n4nd0 | yeah | 15:50 |
| n4nd0 | for the moment I know that I would like to go to a sunny state | 15:50 |
| n4nd0 | :D | 15:50 |
| n4nd0 | I think I would really enjoy living in California | 15:50 |
| pluskid | Ah, California is awesome, everybody wants to go there :D | 15:51 |
| pluskid | and there are many good universities | 15:51 |
| pluskid | with strong machine learning groups | 15:51 |
| n4nd0 | nice | 15:51 |
| pluskid | n4nd0: what do you want to do for your PhD? | 15:52 |
| pluskid | I mean, what subarea of ml? | 15:52 |
| n4nd0 | pluskid: I am interested in robotics, in particular in localization, planning, SLAM and the like | 15:54 |
| pluskid | hmm, interesting topic | 15:54 |
| pluskid | Andrew Ng has a project of a robot dog, which looks pretty cool | 15:55 |
| pluskid | very robust on walking in various difficulty conditions | 15:55 |
| n4nd0 | pluskid: btw, I was going to ask you about the labels hierarchy you suggested in a PR | 15:56 |
| n4nd0 | did you get some feedback from sonney2k on that? | 15:56 |
| -!- uricamic [~uricamic@2001:718:2:1634:1dd1:edc6:9e87:3324] has quit [Quit: Leaving.] | 16:02 | |
| pluskid | in a PR? | 16:02 |
| pluskid | what is PR? | 16:03 |
| pluskid | anyway, the CLabels hierarchy I suggested was rejected by sonney2k | 16:03 |
| n4nd0 | pull request | 16:03 |
| n4nd0 | ok | 16:03 |
| n4nd0 | we'll have to wait and see how he wants to have it then :) | 16:04 |
| pluskid | currently MC works fine with the current labels | 16:04 |
| pluskid | so that hierarchy might makes thing complicated | 16:04 |
| pluskid | not necessary | 16:04 |
| pluskid | so currently we will not refactor the labels system for MC | 16:04 |
| pluskid | but I'm afraid something is needed for SO | 16:05 |
| pluskid | I think you know the best what is needed in SO | 16:05 |
| pluskid | maybe we might consider refactoring after your SO labels are created | 16:05 |
| n4nd0 | I have already created them | 16:07 |
| pluskid | oh, cool | 16:08 |
| pluskid | where are they? | 16:08 |
| n4nd0 | but they are kind of out of hierarchy | 16:08 |
| n4nd0 | take a look to my branch so in github | 16:08 |
| n4nd0 | or in the pull request here, https://github.com/shogun-toolbox/shogun/pull/529 | 16:08 |
| pluskid | we might be able to incorporate them into the hierarchy | 16:08 |
| n4nd0 | for the idea I have in mind what would be best is to have them under CLabels, something like CLabels <--- CSOLabels | 16:09 |
| n4nd0 | but the current CLabels should change | 16:10 |
| n4nd0 | since I am not interested to inherit in the SGVector< float64_t > for non-SO labels for example | 16:10 |
| pluskid | I didn't find it, what's the filename for your SOLabels? | 16:11 |
| -!- wiking [~wiking@c-98-231-171-175.hsd1.md.comcast.net] has joined #shogun | 16:14 | |
| -!- wiking [~wiking@c-98-231-171-175.hsd1.md.comcast.net] has quit [Changing host] | 16:14 | |
| -!- wiking [~wiking@huwico/staff/wiking] has joined #shogun | 16:14 | |
| n4nd0 | pluskid: https://github.com/iglesias/shogun/blob/so/src/shogun/features/StructuredLabels.h | 16:16 |
| -!- emrecelikten [~emrecelik@82.222.148.100] has joined #shogun | 16:16 | |
| n4nd0 | but there is almost nothing there yet, quite basic so far | 16:16 |
| n4nd0 | I am waiting for the decision of the hierarchy | 16:16 |
| pluskid | yeah, it's difficulty to work with if CStructuredLabels is not inherited from CLabels | 16:18 |
| pluskid | because CSOMachine is interited from CMachine, right? | 16:18 |
| pluskid | so you are actually accepting CLabels for set_labels() method | 16:18 |
| pluskid | but what you are expecting should be CStructuredLabels | 16:19 |
| pluskid | the refactoring is non-trivial, since the implementations are different (SGVector vs. CDynamicObjectArray) | 16:19 |
| pluskid | I think you'd better discuss with sonney2k first | 16:20 |
| n4nd0 | yes | 16:21 |
| n4nd0 | yesterday blackburn and sonney2k were discussing about it | 16:22 |
| n4nd0 | if it made sense to have a base class CLabels | 16:22 |
| pluskid | so do they have any conclusion? | 16:22 |
| pluskid | a base CLabels that is very empty? | 16:22 |
| n4nd0 | that's is what I support | 16:23 |
| n4nd0 | but there is no conclusion yet | 16:23 |
| pluskid | the problem is this will need a lot of work :p | 16:25 |
| pluskid | to modify the existing code to perform explicit casting | 16:25 |
| n4nd0 | yes, labels are used everywhere | 16:26 |
| -!- eric_ [2e1fd566@gateway/web/freenode/ip.46.31.213.102] has quit [Quit: Page closed] | 16:45 | |
| -!- pluskid [~pluskid@111.120.55.156] has quit [Quit: Leaving] | 17:02 | |
| blackburn | no not a lot of work | 17:03 |
| blackburn | pretty easy to do I think | 17:03 |
| -!- emrecelikten [~emrecelik@82.222.148.100] has quit [Quit: Leaving.] | 18:13 | |
| -!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving] | 18:17 | |
| -!- vikram360 [~vikram360@117.192.176.104] has quit [Ping timeout: 256 seconds] | 18:23 | |
| -!- puffin444 [62e3926e@gateway/web/freenode/ip.98.227.146.110] has joined #shogun | 18:28 | |
| -!- puffin444 [62e3926e@gateway/web/freenode/ip.98.227.146.110] has quit [Ping timeout: 245 seconds] | 18:45 | |
| -!- puffin444 [62e3926e@gateway/web/freenode/ip.98.227.146.110] has joined #shogun | 19:17 | |
| puffin444 | hey | 19:17 |
| blackburn | hey | 19:18 |
| puffin444 | Do you know when karlnapf is typically on irc? He wanted to chat with me about the Model Selection framework. | 19:20 |
| blackburn | puffin444: he tends to appear rather randomly last days - he has exams - so no idea | 19:23 |
| puffin444 | Oh okay. | 19:23 |
| blackburn | but IIRC I have seen him there late (for me) evenings | 19:24 |
| puffin444 | He had some ideas about extending the framework to suit hyperparameter learning. | 19:24 |
| puffin444 | You | 19:24 |
| blackburn | me? :D | 19:25 |
| puffin444 | What timezone are you in? | 19:25 |
| blackburn | ah | 19:25 |
| blackburn | utc+4 I guess | 19:25 |
| puffin444 | sry typo :) | 19:25 |
| puffin444 | hmm.... so early afternoon for me. | 19:28 |
| puffin444 | We have another GSoC IRC meeting the 23rd, yes? | 19:28 |
| blackburn | let me check in logs | 19:30 |
| blackburn | puffin444: 25th | 19:32 |
| blackburn | 11 UTC | 19:32 |
| blackburn | puffin444: did that presentation of you already take place? | 19:36 |
| puffin444 | Oh not yet. Not until next week. I am still waiting to hear back for more information. | 19:38 |
| blackburn | I see | 19:38 |
| puffin444 | I think I am going to go ahead and start making a presentation anyway because time is running out. | 19:38 |
| CIA-113 | shogun: Soeren Sonnenburg master * r88f1152 / (3 files in 3 dirs): remove vec_index - http://git.io/46JbEQ | 19:40 |
| blackburn | another one bloody commit from soeren | 19:41 |
| puffin444 | bloody? | 19:46 |
| blackburn | yeah he cuts off everything :D | 19:48 |
| puffin444 | I see :) | 19:53 |
| -!- vikram360 [~vikram360@117.192.176.104] has joined #shogun | 20:03 | |
| -!- shogun-buildbot [~shogun-bu@7nn.de] has quit [Quit: buildmaster reconfigured: bot disconnecting] | 20:19 | |
| -!- shogun-buildbot [~shogun-bu@7nn.de] has joined #shogun | 20:19 | |
| -!- puffin444 [62e3926e@gateway/web/freenode/ip.98.227.146.110] has quit [Quit: Page closed] | 20:25 | |
| @sonney2k | blackburn, the solution we had yesterday doesn't work | 20:50 |
| blackburn | sonney2k: which one? | 20:50 |
| @sonney2k | blackburn, removal of CMachine apply() | 20:51 |
| blackburn | ah | 20:51 |
| blackburn | why? | 20:51 |
| @sonney2k | and have customized apply() stuff in sub-machines | 20:51 |
| @sonney2k | blackburn, because multiclass machine wants to call applyu | 20:51 |
| @sonney2k | from general machines... | 20:51 |
| blackburn | sonney2k: we can assume it is a binary machine | 20:52 |
| @sonney2k | but machine has no apply -> dead | 20:52 |
| @sonney2k | we don't have 'binary machines' | 20:52 |
| blackburn | we should have binary machine | 20:52 |
| @sonney2k | no | 20:52 |
| @sonney2k | we discussed that yesterday kernelmachines | 20:52 |
| @sonney2k | are binary,regression, ... | 20:52 |
| @sonney2k | everything | 20:52 |
| blackburn | then I see no way | 20:52 |
| blackburn | well wait | 20:53 |
| @sonney2k | I think we should have apply() return RealLabels | 20:53 |
| @sonney2k | it won't work for SO though | 20:53 |
| blackburn | sonney2k: there is a way | 20:54 |
| blackburn | pretty easy | 20:54 |
| blackburn | sonney2k: for domain adaptation multiclass svm I extracted outputs of sub-machines to separate functions | 20:54 |
| blackburn | they are virtual and could be overrided in linear and kernel multiclass machine | 20:54 |
| blackburn | with proper casting | 20:55 |
| blackburn | general multiclass is abstract at all so no problem there | 20:55 |
| blackburn | it could have no impl at all | 20:56 |
| blackburn | sonney2k: is there any other issue? | 20:56 |
| @sonney2k | blackburn, I have no idea what you are saying | 20:58 |
| blackburn | cool | 20:58 |
| blackburn | sonney2k: from the very beginning? | 20:58 |
| blackburn | sonney2k: all casts before apply in multiclass machine | 20:59 |
| blackburn | can go to linear and kernel multiclass machine | 20:59 |
| blackburn | with casts to linear and kernel machines | 20:59 |
| -!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 20:59 | |
| blackburn | so if we add apply in both kernel and linear machine | 20:59 |
| blackburn | your issue is solved | 20:59 |
| blackburn | sonney2k: got? | 20:59 |
| @sonney2k | no | 21:01 |
| @sonney2k | I don't know what you are talking about | 21:01 |
| @sonney2k | the only solution I see now is: | 21:01 |
| @sonney2k | - keep apply() in CMachine | 21:01 |
| blackburn | sonney2k: again.. | 21:01 |
| @sonney2k | - check the label type of the returned label | 21:01 |
| blackburn | can we move apply to | 21:01 |
| blackburn | kernel | 21:01 |
| blackburn | and linear | 21:01 |
| blackburn | machines? | 21:01 |
| @sonney2k | and cast appropriately | 21:01 |
| @sonney2k | no | 21:02 |
| blackburn | what is the issue you want to solve?? | 21:02 |
| @sonney2k | because some machines work with any other machine that have an apply() method | 21:02 |
| -!- vikram360 [~vikram360@117.192.176.104] has quit [Ping timeout: 246 seconds] | 21:03 | |
| blackburn | sonney2k: where you want to put apply? | 21:03 |
| @sonney2k | in CMachine if possible | 21:04 |
| blackburn | if not? | 21:04 |
| @sonney2k | in submachines | 21:04 |
| blackburn | which ones? | 21:04 |
| @sonney2k | linear, kernel, etc | 21:05 |
| @sonney2k | but it is possible | 21:05 |
| @sonney2k | only problem is that interfaces that cannot cast just get some Label object | 21:05 |
| blackburn | why did you say we can't put apply to linearmachine? | 21:05 |
| @sonney2k | which sucks... | 21:05 |
| @sonney2k | blackburn, we have to have it in CMachine | 21:06 |
| @sonney2k | because e.g. multiclass machine calls apply() on general machines | 21:06 |
| blackburn | sonney2k: I said two times we can use apply on linear/kernel machines in multiclass machines | 21:07 |
| @sonney2k | how? | 21:07 |
| blackburn | sonney2k: put apply part to {linear,kernel}multiclassmachine | 21:07 |
| @sonney2k | not nice either | 21:08 |
| blackburn | ?? | 21:08 |
| blackburn | solves your issue | 21:08 |
| @sonney2k | but you would need to determine the machine type in mc machine | 21:09 |
| @sonney2k | then cast to linear/kernel machine and call apply there | 21:09 |
| blackburn | in linearmulticlassmachine all machines are linearmachines | 21:09 |
| blackburn | implemented virtual method of {linear,kernel}multiclassmachine does the job of casting and applying | 21:10 |
| blackburn | now we anyway cast it to CMachine | 21:10 |
| blackburn | anyway I need to made that apply part virtual | 21:11 |
| blackburn | because DA MC SVM needs to combine outputs of target and source domain submachines | 21:11 |
| @sonney2k | ok | 21:11 |
| blackburn | I will commit da mc svm soon | 21:12 |
| @sonney2k | but I would try to get apply() to return just the baseclass | 21:12 |
| blackburn | base class? | 21:12 |
| blackburn | CLabels? | 21:12 |
| @sonney2k | CLabels | 21:12 |
| blackburn | didn't you want to remove base class there? | 21:12 |
| @sonney2k | well I thought I had no other option | 21:12 |
| blackburn | is there a way now? | 21:13 |
| @sonney2k | but I think we have an option | 21:13 |
| @sonney2k | yeah we always just return CLabel* on apply() | 21:13 |
| blackburn | is there any other issue except that multiclass? | 21:13 |
| @sonney2k | I mean if we construct a MulticlassLabel or so it will be cast to CLabel and can be cast back | 21:14 |
| @sonney2k | so no problem | 21:14 |
| @sonney2k | and it will work with StructuredOutputLabel too | 21:14 |
| @sonney2k | so CMachine has a nice interface | 21:14 |
| blackburn | yes I am ok with it too | 21:14 |
| @sonney2k | I think for swig one can hack sth up to return the real label type | 21:15 |
| @sonney2k | by using %ignore and %rename | 21:15 |
| @sonney2k | and potential helper functions | 21:15 |
| @sonney2k | that do the casting to CRealLabel etc | 21:15 |
| shogun-buildbot | build #233 of nightly_none is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/nightly_none/builds/233 | 21:16 |
| blackburn | http://cs11441.userapi.com/v11441184/20b/B6B0ffVO2Qk.jpg | 21:17 |
| blackburn | 2140 warnings wat?! | 21:17 |
| blackburn | ah | 21:17 |
| blackburn | dynamicarray warning for each type | 21:17 |
| @sonney2k | I think gsomix removed some warning ignore | 21:19 |
| @sonney2k | alright | 21:19 |
| @sonney2k | I think I will change everything back to CMachine -> CLabels* apply | 21:19 |
| blackburn | hah it seems you do some great changes in your .git | 21:20 |
| n4nd0 | sonney2k, blackburn: hi guys | 21:20 |
| blackburn | n4nd0: hey | 21:20 |
| n4nd0 | I don't think I really understand the plans you have for CLabels | 21:21 |
| blackburn | do you think you are the only one? | 21:21 |
| n4nd0 | can you tell me about it more thoroughly? you know I am quite interested in it | 21:21 |
| blackburn | :D | 21:21 |
| blackburn | hmm I need some workforce | 21:22 |
| shogun-buildbot | build #241 of nightly_all is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/nightly_all/builds/241 | 21:42 |
| CIA-113 | shogun: Sergey Lisitsyn master * r33fc1e4 / (10 files in 5 dirs): Added DomainAdaptationMulticlassLibLinear - http://git.io/sAjyZw | 21:48 |
| blackburn | n4nd0: okay actually we need clabels or not | 21:56 |
| blackburn | I am not sure we really need it but it looks better with it | 21:56 |
| blackburn | :D | 21:56 |
| n4nd0 | :D | 21:59 |
| shogun-buildbot | build #153 of nightly_default is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/nightly_default/builds/153 | 22:21 |
| -!- gsomix [~gsomix@83.234.54.165] has joined #shogun | 22:31 | |
| gsomix | hi all | 22:31 |
| gsomix | I'm alive. | 22:31 |
| blackburn | I sent you a cool pic | 22:32 |
| gsomix | I had a hard week. tomorrow I'm going with the news. | 22:32 |
| gsomix | blackburn, yeah | 22:32 |
| blackburn | where have you been? | 22:32 |
| blackburn | optics trip? :D | 22:33 |
| gsomix | blackburn, in botan-land. | 22:33 |
| gsomix | sorry for that | 22:34 |
| gsomix | blackburn, about warnings. I'll remove it. | 22:34 |
| shogun-buildbot | build #873 of cmdline_static is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/cmdline_static/builds/873 | 22:35 |
| blackburn | yes that could be cool to get rid of them | 22:35 |
| gsomix | aha, and about directors. I have read docs. tomorrow I'm going to experiment. | 22:36 |
| gsomix | good night | 22:40 |
| gsomix | sonney2k, (^___^)/ | 22:40 |
| -!- gsomix [~gsomix@83.234.54.165] has quit [Quit: Ex-Chat] | 22:43 | |
| shogun-buildbot | build #852 of r_static is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/r_static/builds/852 | 22:46 |
| shogun-buildbot | build #762 of octave_static is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/octave_static/builds/762 | 22:56 |
| shogun-buildbot | build #839 of python_static is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/python_static/builds/839 | 23:02 |
| --- Log closed Sat May 19 00:00:40 2012 | ||
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!
