Open in new window / Try shogun cloud
--- Log opened Thu Mar 22 00:00:19 2012
-!- romi_ [~mizobe@187.57.1.50] has quit [Ping timeout: 246 seconds]00:33
-!- romi_ [~mizobe@187.57.1.50] has joined #shogun00:45
-!- romi_ [~mizobe@187.57.1.50] has quit [Ping timeout: 246 seconds]00:52
n4nd0sonney2k: I just sent a pull request for QDA but appears to me like none has been assigned to it, can you tell me if you have been notified?00:53
-!- vikram360 [~vikram360@117.192.164.158] has joined #shogun01:15
-!- PhilTillet [~Philippe@tillet-p42154.maisel.int-evry.fr] has quit [Remote host closed the connection]01:40
-!- mayanks43 [~Marcus@115.248.130.148] has joined #shogun01:57
-!- l0nr4n [~l0nr4n@unaffiliated/l0nr4n] has quit [Quit: This computer (or maybe me) has gone to sleep]02:06
-!- canonind [d2198538@gateway/web/freenode/ip.210.25.133.56] has joined #shogun02:15
-!- canonind [d2198538@gateway/web/freenode/ip.210.25.133.56] has quit [Quit: Page closed]02:33
-!- canonind [d2198538@gateway/web/freenode/ip.210.25.133.56] has joined #shogun02:33
-!- canonind [d2198538@gateway/web/freenode/ip.210.25.133.56] has quit [Client Quit]02:34
-!- Canonind [d2198538@gateway/web/freenode/ip.210.25.133.56] has joined #shogun02:35
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving]02:50
-!- vikram360 [~vikram360@117.192.164.158] has quit [Ping timeout: 252 seconds]03:20
-!- vikram360 [~vikram360@117.192.164.158] has joined #shogun03:23
-!- vikram360 [~vikram360@117.192.164.158] has quit [Ping timeout: 252 seconds]03:35
-!- yash [46ab0416@gateway/web/freenode/ip.70.171.4.22] has joined #shogun04:07
-!- romi_ [~mizobe@187.57.1.50] has joined #shogun04:43
-!- yash [46ab0416@gateway/web/freenode/ip.70.171.4.22] has quit [Quit: Page closed]04:43
-!- Miggy [~piggy@14.139.82.6] has quit [Ping timeout: 260 seconds]04:56
-!- mayanks43 [~Marcus@115.248.130.148] has left #shogun []05:26
-!- harshit_ [~harshit@182.68.142.173] has joined #shogun05:30
-!- in3xes [~in3xes@180.149.49.227] has joined #shogun06:10
-!- in3xes [~in3xes@180.149.49.227] has quit [Ping timeout: 244 seconds]06:32
-!- in3xes [~in3xes@210.212.58.111] has joined #shogun06:45
-!- harshit_ [~harshit@182.68.142.173] has quit [Ping timeout: 252 seconds]07:11
-!- gsomix [~gsomix@85.26.232.68] has joined #shogun07:23
gsomixhi07:23
-!- harshit_ [~harshit@182.68.142.173] has joined #shogun07:24
-!- in3xes_ [~in3xes@59.163.196.121] has joined #shogun07:27
-!- in3xes_ [~in3xes@59.163.196.121] has quit [Client Quit]07:27
-!- in3xes_ [~in3xes@59.163.196.121] has joined #shogun07:28
-!- in3xes__ [~in3xes@59.163.196.121] has joined #shogun07:37
-!- in3xes__ [~in3xes@59.163.196.121] has quit [Remote host closed the connection]07:40
-!- in3xes [~in3xes@210.212.58.111] has quit [Ping timeout: 276 seconds]07:41
-!- in3xes_ [~in3xes@59.163.196.121] has quit [Ping timeout: 265 seconds]07:41
-!- gsomix [~gsomix@85.26.232.68] has quit [Ping timeout: 246 seconds]07:47
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun07:49
harshit_hey @n4nd007:52
n4nd0harshit_: hi there!07:52
n4nd0harshit_: how is your stuff going?07:53
harshit_one more problem now07:53
harshit_my ./configure dont recognize lapack library07:53
harshit_i have installed it and also its c interface07:54
n4nd0ok07:54
n4nd0so how have you installed it?07:54
harshit_by making some changes in make.inc file07:55
harshit_and then installing fortran library07:55
harshit_and then installed the lapacke07:55
n4nd0mmm07:55
n4nd0what OS are you with?07:55
harshit_ubuntu 11.1007:55
n4nd0apt-get install liblapack-dev?07:56
n4nd0I normally install everything with apt-get07:56
n4nd0I think I did this also this way07:56
harshit_okay wait i am doing that07:57
harshit_Is lapack example in shogun's examples, Running on your machine07:58
harshit_?07:58
n4nd0which example in particular?07:59
harshit_mathematics_lapack.cpp07:59
n4nd0compiling ...08:00
harshit_hey thanks ./configure now recognized the lapack08:00
n4nd0yes, it looks like it runs without problems08:00
harshit_thanks apt-get is great08:01
n4nd0yes08:01
n4nd0if you are fond of more graphical stuff I recommend you synaptic08:01
n4nd0apt-get is good enough though08:01
harshit_also one more thing if say suppose i didht have any library installed at time i installed shogun08:02
n4nd0the only problem is that you need to know the name of the package to install / have intuition for the naming conventation08:02
n4nd0conventions*08:02
harshit_and then later i install that library08:02
harshit_then do i need to recompile whole shogun ?08:02
n4nd0yes, probably08:03
n4nd0I think there's no general answer for that (there may be some exceptions) but in general yes08:03
n4nd0to build shogun faster I recommend you to do08:04
n4nd0./configure --disable-optimization [the other options you want to use]08:04
n4nd0and later run make for several threads08:04
n4nd0make -j2 or make -j408:04
harshit_that would be useful sometimes08:04
n4nd0yes08:05
-!- Canonind [d2198538@gateway/web/freenode/ip.210.25.133.56] has left #shogun []08:07
harshit_hey, my shogun compiled and is installed successfully but still when i run mathematics_lapack.cpp it doesn't show anything08:08
-!- Jiuding [d2198538@gateway/web/freenode/ip.210.25.133.56] has joined #shogun08:08
Jiudinghi feeling great to join you08:09
n4nd0harshit_: well, have you checked the code?08:10
harshit_yes .. in my case HAVE_LAPACK is not defined08:10
harshit_so ifdef evaluates to be false08:11
harshit_and doesnt show anything08:11
harshit_and HAVE_LAPACK is supposed to be defined in config.h08:11
n4nd0I think that there's no nothing to show in that example apart from error messages08:11
-!- gsomix [~gsomix@85.26.232.68] has joined #shogun08:12
n4nd0there's nothing* ... fuck, my fingers are clumsy in the morning08:12
harshit_but still i dont knw why my machine is doing that08:14
harshit_i think i should redo the entire procedure08:15
n4nd0how do you know that HAVE_LAPACK is evaluated to false?08:16
n4nd0have you tried it writing some printfs inside?08:16
harshit_becoz if i printf inside it doesnt print08:16
harshit_yes08:16
n4nd0ok08:16
n4nd0have you built shogun after installing lapack?08:17
harshit_yes08:17
n4nd0you should run configure, check that lapack is detected08:17
n4nd0run make, and finally make install08:17
n4nd0have you done all the steps?08:17
harshit_yes after you told me to install via apt-get08:17
harshit_it was detected by configure08:17
n4nd0so if you did later make and make install I guess it should work08:18
n4nd0however, it might be that there are other dependencies in order for lapack to work, I don't come up with other explanation08:18
harshit_yes i did08:18
harshit_i think only way out is to reinstall08:19
harshit_no ?08:19
harshit_reinstall everything08:19
harshit_also do you knw how to uninstall any library08:19
n4nd0mmm08:20
n4nd0I don't know if I recommend you to do that, one should be careful08:20
n4nd0yes, you can uninstall libraries with apt-get remove08:21
n4nd0easier with synaptic08:21
n4nd0I'd recommend you to take a look here first http://www.shogun-toolbox.org/doc/en/current/installation.html08:21
n4nd0in the requirements, it might be that you're still missing sth08:22
harshit_what is sth ?08:22
n4nd0it looks like that atlas library could be necessary08:22
n4nd0ups sorry, sth = something08:23
n4nd0:)08:23
harshit_i thought sth is some program08:24
harshit_also i tried to apt-get it :)08:24
n4nd0haha sorry08:24
harshit_okay thanks n4nd0 ,g2g will do it and08:25
harshit_will ask you some more questions .lol08:25
n4nd0apt-cache show is useful to read package descriptions before installation08:26
n4nd0try if you want for example08:26
n4nd0apt-cache show liblapack-dev08:26
n4nd0you'll see the description of the package you just installed08:26
harshit_yes it works. nice08:27
harshit_and also when we download documentation through synaptic, how can we access that08:28
-!- gsomix [~gsomix@85.26.232.68] has quit [Ping timeout: 246 seconds]08:31
n4nd0afaik the package documentation is usually man pages08:32
n4nd0e.g. now you should be able to do man lapack08:32
n4nd0if you want to know more info about a topic apropos08:32
harshit_but when i do man lapack it says :08:33
harshit_No manual entry for lapack08:33
harshit_same for every other documentation i download08:33
n4nd0probably you need the package liblapack-doc08:33
harshit_yeah get it..08:36
harshit_thanks08:36
harshit_bye08:36
-!- harshit_ [~harshit@182.68.142.173] has quit [Quit: Leaving]08:37
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving]08:55
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun09:37
-!- blackburn [5bdfb203@gateway/web/freenode/ip.91.223.178.3] has joined #shogun09:43
blackburnn4nd0: yes we were notified09:44
n4nd0blackburn: cool :)09:45
blackburnsonne|work: ping09:49
blackburnThis program is SG_FREE software09:56
blackburnoh lol auto replace09:56
n4nd0haha09:56
blackburnn4nd0: so did you fix this mess in your repo?10:05
n4nd0blackburn: yeah, it is looking way better now :)10:05
n4nd0blackburn: I have to ask you one thing about one vs. one training, not repo related10:05
blackburnn4nd0: just create new branches for each new pull10:06
blackburnand use rebase not merge10:06
blackburnn4nd0: ask10:06
n4nd0blackburn: I didn't manage to get to work the trick using the same vector all the time, I don't know if you remember10:07
n4nd0blackburn: the thing was to allocate an array big enough and pass to CSubset on each iteration the adequate lenght of this array10:07
n4nd0blackburn: probably not the complete one10:07
n4nd0blackburn: the thing is that later, when one does remove subset both in feature and labels, the vector is also de-allocated so in the next second iteration it explodes10:08
n4nd0blackburn: let me show you in code10:08
n4nd0https://github.com/iglesias/shogun/blob/multiclass_1vs1/src/shogun/machine/MulticlassMachine.cpp10:10
n4nd0the nice idea would be to have lines 147 and 148 out of the for loop10:10
blackburnn4nd0: let it stay for now10:12
blackburnnot really high demanding thing10:12
n4nd0blackburn: ok :)10:12
n4nd0blackburn: about the repo, I read about rebase yesterday and I think I got the difference wrt merge10:13
n4nd0blackburn: but why do you think it is better to do it that way?10:13
blackburnn4nd0: yeah you just put your commits on top of it10:13
blackburnno 'merged blabla from blabla' commits10:13
n4nd0ok10:14
n4nd0but in any case I should not worry about merging my branches into master when doing a pull request10:14
blackburnn4nd0: the approach is to keep master updated and rebase your branches on top of it10:15
blackburnin this way everything looks clearer10:16
blackburnhttp://book.git-scm.com/assets/images/figure/rebase5.png10:16
blackburnhttp://book.git-scm.com/assets/images/figure/rebase5.png10:16
blackburnehm10:16
blackburnone picture heh10:16
blackburnn4nd0: commit graph would not look like some social graph :D10:17
n4nd0hehe ok :)10:17
blackburnn4nd0: train 1-v-1 looks ready10:21
blackburnwhat u'd need now is insert classify 1-v-1 from multiclasssvm10:21
-!- muddo [~muddo@gateway/tor-sasl/muddo] has quit [Ping timeout: 276 seconds]10:25
n4nd0blackburn: yes10:26
n4nd0blackburn: the simple example I created with LibLinear was working at least10:27
-!- muddo [~muddo@gateway/tor-sasl/muddo] has joined #shogun10:27
blackburnn4nd0: you mean it trained successfully?10:27
n4nd0blackburn: yes10:27
blackburnnicer10:28
blackburnnice*10:28
n4nd0blackburn: https://github.com/iglesias/shogun/commit/61c4eeb055ac17bbbb3c4bed6f41e863064d94e110:28
n4nd0blackburn: that program breks because apply is not implemented yet, what is a good signal this far :)10:28
n4nd0brb10:28
blackburnok10:28
blackburnn4nd0: I'd suggest you to add method add_subset to kernel10:33
blackburnn4nd0: for custom kernel it just delegates to set_row and set_col subsets10:34
blackburnfor others just checks lhs==rhs and sets subsets for lhs and rhs10:34
n4nd0do you mean add_subset in KernelMachine or just in Kernel?10:38
blackburnin kernel would be enough I think10:41
blackburnthisway you may simply delegate it in add_machine_subset and remove_machine_subset10:41
n4nd0aha I see what you mean10:42
n4nd0and the method in kernel should be virtual so it can be specialized in CustomKernel10:43
blackburnyeah10:44
blackburnLUNCH TIME10:44
blackburn:D10:44
n4nd0enjoy your meal10:45
-!- wiking [~wiking@huwico/staff/wiking] has quit [Quit: wiking]10:59
-!- wiking [~wiking@huwico/staff/wiking] has joined #shogun10:59
-!- wiking [~wiking@huwico/staff/wiking] has quit [Ping timeout: 264 seconds]11:04
-!- wiking [~wiking@huwico/staff/wiking] has joined #shogun11:10
blackburnre11:22
-!- Jiuding [d2198538@gateway/web/freenode/ip.210.25.133.56] has quit [Ping timeout: 245 seconds]11:23
-!- blackburn_ [5bdfb203@gateway/web/freenode/ip.91.223.178.3] has joined #shogun12:30
-!- blackburn [5bdfb203@gateway/web/freenode/ip.91.223.178.3] has quit [Ping timeout: 245 seconds]12:32
-!- l0nr4n [~l0nr4n@hugogascon.ml.tu-berlin.de] has joined #shogun12:45
-!- l0nr4n [~l0nr4n@hugogascon.ml.tu-berlin.de] has quit [Changing host]12:45
-!- l0nr4n [~l0nr4n@unaffiliated/l0nr4n] has joined #shogun12:45
-!- vikram360 [~vikram360@117.192.163.50] has joined #shogun13:04
-!- harshit_ [~harshit@182.68.142.173] has joined #shogun13:09
-!- vikram360 [~vikram360@117.192.163.50] has quit [Ping timeout: 246 seconds]13:10
-!- vikram360 [~vikram360@117.192.164.215] has joined #shogun13:10
-!- harshit_ [~harshit@182.68.142.173] has quit [Ping timeout: 252 seconds]13:46
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: Lost terminal]13:54
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun13:55
-!- gsomix [~gsomix@85.26.165.233] has joined #shogun13:55
gsomixhi13:56
gsomixphilosophy is too deep for me :)13:56
-!- gsomix [~gsomix@85.26.165.233] has quit [Client Quit]14:00
-!- gsomix [~gsomix@85.26.165.233] has joined #shogun14:00
-!- harshit_ [~harshit@182.68.142.173] has joined #shogun14:18
n4nd0blackburn_: how important is to do the m_parameters->add(...) for each of the members of the class in the constructor?14:22
n4nd0blackburn_: does it give problems sometimes?14:22
blackburn_n4nd0: it is required for proper serialization14:22
n4nd0wihout clear reason it has started to fail for a boolean member14:22
blackburn_n4nd0: fail to compile?14:23
blackburn_what is the error/member?14:24
n4nd0seg fault14:24
n4nd0is a float sorry not a boolean14:24
n4nd0float64_t14:24
blackburn_wait, segfault related to float64_t??14:24
n4nd0yes14:24
blackburn_show me the code14:24
blackburn_i.e. it fails with m_parameters->add and float64_t??14:25
blackburn_that shouldn't be possible14:25
n4nd0http://snipt.org/ugDh014:26
n4nd0the line between the rpints and seg faults14:26
n4nd0info from gdb14:26
n4nd0Program received signal SIGSEGV, Segmentation fault.14:26
n4nd00xb4dd2170 in shogun::Parameter::add_type (this=0x917ec10, type=0xbfffecec, param=0x917d50c, name=0xb4e732bf "m_tolerance", description=0xb4e732ad "Tolerance member.") at base/Parameter.cpp:236914:26
n4nd02369if (strcmp(m_params.get_element(i)->m_name, name) == 0)14:26
blackburn_whoaa14:27
n4nd0??14:27
blackburn_no idea how can it be possible14:28
n4nd0but it started to happen when I added this m_slog member14:29
n4nd0a SGVector< int32_t >14:29
n4nd0before that it was working good14:29
n4nd0I cannot see any connection though14:29
blackburn_try to comment it out then14:32
n4nd0shit, now it seg faults in the next m_parameters->add14:34
n4nd0something must have got messed while adding a new member14:35
n4nd0I don't see why though14:35
sonne|workn4nd0: stack overflow?14:35
blackburn_sonne|work: how can it cause it14:35
n4nd0sonne|work: do you mean because there may be too many attributes?14:36
sonne|workmaybe too much allocated on stack...14:37
sonne|workwhat does valgrind say?14:37
n4nd0in the moment that method is executed almost nothing has been allocated14:37
blackburn_sonne|work: how can I reach you to get answer? :D14:37
sonne|workI am just reading your email14:37
blackburn_I asked you in mail yesterday14:38
sonne|workI think what you want to do is a lot of work14:38
blackburn_in irc14:38
blackburn_and in mail again!14:38
blackburn_no, I already did a half of it14:38
sonne|workand I don't think that one should do this14:38
-!- menonnik [~chatzilla@59.178.164.96] has joined #shogun14:38
sonne|workjust for the purpose of getting it published14:38
blackburn_I like this idea pretty much14:38
blackburn_no,  I like idea itself14:38
sonne|workcheng seemed fine without this so there is no need14:38
blackburn_sonne|work: it would be better in this way14:39
sonne|workwell of course you can do whatever you want but you need e.g. kernels/distances some interface to some other language or cmdline interface, some minimal build system14:39
blackburn_no!14:39
sonne|workthat is >1 week of work14:39
menonniksir, i wanted some clarifications regarding implementation if C5.0. Are we training the data to form a tree or a rule set?14:40
blackburn_sonne|work: it can even improve further development14:40
n4nd0sonne|work: it executed with valrgind without seg fault14:40
blackburn_I liked callbacks of OCAS14:40
sonne|workbut ocas has all this extra - extra matlab interface, extra implementation of dot features for sparse & dense, extra parallelizaed code, extra implementation for double and float...14:41
blackburn_sonne|work: I don't want to create my own lib!14:41
blackburn_I just want to extract code to some lib everyone can borrow as easy as libsvm14:42
blackburn_or liblinear14:42
blackburn_my point is just extract deep internals to libedrt.cpp14:42
blackburn_with callback interface14:42
blackburn_everything else should stay14:42
sonne|workliblinear also has a cmdline interface & file format reader/writer14:43
blackburn_well if they won't like it again I'll create some simple interface14:44
blackburn_but this time I just want to extract its guts to make it borrowable14:44
blackburn_I hope 3rd submission is possible?14:44
sonne|workblackburn_: to publish you need to be able to show that your system works somehow - so you need to provide some example code - some interface14:44
sonne|workthey will need an interface14:44
blackburn_it works with shogun..14:45
blackburn_you suggest the same14:45
sonne|workyou should aim at getting it in this time for sure14:45
blackburn_I do not understand your point :(14:45
sonne|workand I don't see how a separation helps here14:45
blackburn_because it would be hard to re-use that code anywhere14:46
sonne|workwhat I mean is reviewers were rather positive. this and an improved selling of EDRT alone should suffice to get it accepted14:46
blackburn_but with libedrt.cpp and callback it would be reusable14:46
blackburn_all the converter stuff will just delegate its apply to libedrt14:47
blackburn_sonne|work: so do you prefer to let code stay as it is?14:47
sonne|workyou didn't describe exactly what you want to do. to me it seems one can just use libshogun and do all the re-use one wants14:47
sonne|workblackburn_: well yes but maybe you need to specify what you want to do before I can give a definitive answer14:48
blackburn_sonne|work: I suggest to extract LLE/LTSA/all other stuff internals14:48
blackburn_with callback distances/feature vector access/kernels14:48
n4nd0sonne|work, blackburn_ : how can it be that it executes with valgrind but not alone?14:49
blackburn_n4nd0: valgrind does some sandboxing..14:49
sonne|workblackburn_: print functions, exceptions too - right?14:50
blackburn_sonne|work: in that case one could borrow libedrt code and provide custom distances14:50
blackburn_with no need to extend shogun somehow14:50
blackburn_it is more custom than shogun14:50
blackburn_sonne|work: why would I need any exceptions?14:51
blackburn_but probably yes14:51
blackburn_I don't think there should be any printing though14:51
sonne|workerror  messages / progress14:51
blackburn_yeah, you are right, it would be needed14:52
blackburn_sonne|work: so shogun's LLE would stay but it would delegate m_distance->distance(i,j) as callback14:52
sonne|workand then you need implementations for every callback14:53
sonne|workas illustration at least14:53
blackburn_I need only distances, feature vector access and kernels14:53
blackburn_nothing more14:53
sonne|work(which is a lot already)14:53
blackburn_&m_distance->distance14:53
sonne|workso say gaussian kernel, euclidian dist, and dense realvalued features14:53
blackburn_&m_kernel->kernel14:54
blackburn_sonne|work: why should I write this?14:54
sonne|workto publish a separate libedrt - you would need to show that it is sth useful so you need the functionality + some illustration how it can be used in the most simple case14:55
blackburn_got it14:56
blackburn_well not that much work14:56
blackburn_sonne|work: again, would I need to do separate libedrt??14:56
-!- gsomix [~gsomix@85.26.165.233] has quit [Quit: ????? ? ?? ??? (xchat 2.4.5 ??? ??????)]14:56
blackburn_I see no difference between not 'touching it at all' and 'extract some code'14:57
blackburn_in both ways it is dependent on shogun14:57
blackburn_my intention is not about getting published14:59
blackburn_but about getting it more useful14:59
sonne|workblackburn_: in which way is it more useful? what can you do what you couldn't?15:00
blackburn_sonne|work: one can borrow the code and use it w/o much changing it15:01
blackburn_w/o changing at alll15:01
blackburn_as for shogun - it makes me able to provide pretty nice custom options15:01
blackburn_like computing full distance matrix or not15:01
blackburn_same for kernels15:02
sonne|workwhat do you mean by borrow the code?15:02
blackburn_sonne|work: e.g. waffles can integrate this code in a day15:02
blackburn_it is really easy to provide euclidean distance and feature vector access15:04
blackburn_nothing more is required15:04
blackburn_sonne|work: i.e. just like ocas but developed inside shogun15:05
sonne|workso what you mean is: one only needs to integrate libedrt not whole libshogun?15:05
blackburn_yeah15:05
blackburn_libedrt is shogun independent in that means15:05
blackburn_sonne|work: looks like you don't much like it?15:06
-!- harshit_ [~harshit@182.68.142.173] has quit [Ping timeout: 260 seconds]15:07
blackburn_sonne|work: point is while libedrt provides such custom interfaces15:10
blackburn_i.e. distance/kernel callbacks15:10
sonne|worknot so much no... to use libedrt in shogun one then needs to create (like for libocas) all kinds of wrapping functions15:10
blackburn_but in shogun it is possible to use it everywhere15:10
blackburn_and use file features/etc15:11
blackburn_sonne|work: all wrapping functions are already done in features, distances and kernels15:11
sonne|workno not all - you need one more layer of wrapping functions, like compute_output in CSVMOcas15:11
blackburn_sonne|work: ??15:12
sonne|workthat calls dense_dot_range15:12
sonne|workso you would have e.g. KernelPCA and need a mean to get k(i,j)15:12
blackburn_no, I'm not going to extract kernel pca15:12
sonne|workso you need some kernel_function()15:12
blackburn_it is a different kind of preprocessor15:12
blackburn_only converters15:13
sonne|workwhich underneath is calling kernel->kernel(i,j)15:13
blackburn_sonne|work: yes, my interface prototype requires15:13
blackburn_double (*kernel)(int, int)15:13
blackburn_&m_kernel->kernel15:13
blackburn_that's all15:13
blackburn_same with distance15:14
blackburn_and double* (*feature_vector)(int)15:14
sonne|workno that won't work15:14
blackburn_why?15:14
sonne|workbecause kernel is not a static method15:15
sonne|workso you need a wrapper function ' kernel_function(i,j)' within your class that knows about a kernel object15:15
blackburn_ah I see15:15
blackburn_it is not a problem at all15:15
sonne|workwithin that you can then do kernel->kernel(i,j)15:15
blackburn_I can just provide pointer there15:15
blackburn_it is not a problem still15:15
blackburn_sonne|work: I have to go now, thank you15:17
blackburn_I hope you got my idea15:17
sonne|workyeah I got your idea but I still don't like it :)15:17
blackburn_sonne|work: damn15:18
blackburn_sonne|work: I'll show code next week and then we decide15:18
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: Lost terminal]15:18
blackburn_I still think it would be useful even inside shogun (I can change things easier)15:18
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun15:19
blackburn_see you15:19
-!- blackburn_ [5bdfb203@gateway/web/freenode/ip.91.223.178.3] has quit [Quit: Page closed]15:19
-!- menonnik [~chatzilla@59.178.164.96] has quit [Read error: Connection timed out]15:20
-!- menonnik [~chatzilla@59.178.164.96] has joined #shogun15:21
-!- harshit_ [~harshit@182.68.142.173] has joined #shogun15:25
n4nd0sonne|work: so it seems that getting rid of a couple of class members in QDA that were not used any more solved the problem15:32
n4nd0sonne|work: but I still don't understand why it did happen15:32
sonne|workn4nd0: did you by chance allocate some variables on stack?15:33
sonne|worki mean like15:33
sonne|workdouble bla[10000]15:33
sonne|work?15:33
sonne|workthat would explain it15:33
sonne|work(stack overflow)15:33
n4nd0no15:33
sonne|workSGVector<float64_t> x(1000) doesn't create on stack btw15:33
sonne|workI mean the object yes but not the vector15:34
n4nd0there was an SGVector, a float64_t, a boolean and some pointers15:34
sonne|workhmmhh15:34
n4nd0one of the members was SGVector15:34
n4nd0maybe I should have used SGVector* even if I jus wanted one vector?15:35
n4nd0just*15:35
sonne|workno SGVector is merely a pointer and a boolean underneath15:36
n4nd0then I don't understand why15:36
sonne|workyeah me neither15:38
-!- menonnik [~chatzilla@59.178.164.96] has quit [Quit: ChatZilla 0.9.88 [Firefox 10.0.2/20120215223356]]15:41
n4nd0sonne|work: about valgrind, I get a huge quantity of errors when I run my python examples15:49
n4nd0I am using a suprresions file, but still15:49
n4nd0I have read sth about compiling python, do you know if it improves the results doing it that way?15:50
sonne|workn4nd0: yeah one needs more suppressions for python...15:52
n4nd0sonne|work: I will check how to do that then ...15:53
sonne|workrecompiling with certain settings would help - but these errors are there on purpose (faster ...)15:53
n4nd0the problem is that I find it quite difficult to understand whether the code is buggy or not with all those messages15:54
sonne|workwhat you can do is debug either just libshogun based c++ example or from octave modular or grep for shogun::15:57
sonne|workit is tough for me too btw...15:57
n4nd0libshogun based c++ example looks like the best option from my point of view16:02
n4nd0it also increases the development time though16:02
sonne|workn4nd0: well octave modular is also a good option16:03
sonne|work(octave doesn't have valgrind errors)16:03
n4nd0I will take a look with octave16:04
n4nd0valgrind + octave right?16:05
sonne|workyes16:06
sonne|workjust valgrind octave :)16:06
n4nd0thank you16:06
-!- harshit_ [~harshit@182.68.142.173] has quit [Ping timeout: 244 seconds]16:10
n4nd0sonne|work: do you happen to know why some methods seem to be not accesible from octave?16:21
n4nd0for example apply is not found from the example classifier_lda_modular.m16:21
n4nd0for the class LDA16:21
n4nd0does this have sth to do with swig?16:22
-!- pythonroar [d2198538@gateway/web/freenode/ip.210.25.133.56] has joined #shogun16:24
pythonroarhow to hand up my student application?16:25
-!- muddo [~muddo@gateway/tor-sasl/muddo] has quit [Quit: Leaving]16:27
n4nd0pythonroar: the period is still not open, wait for next week16:27
n4nd0pythonroar: check more info at http://news.gmane.org/gmane.comp.ai.machine-learning.shogun16:27
-!- muddo [~muddo@gateway/tor-sasl/muddo] has joined #shogun16:36
-!- pythonroar [d2198538@gateway/web/freenode/ip.210.25.133.56] has quit [Quit: Page closed]16:39
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving]16:43
-!- vikram360 [~vikram360@117.192.164.215] has quit [Ping timeout: 246 seconds]16:54
-!- vikram360 [~vikram360@117.192.188.225] has joined #shogun16:55
-!- wiking [~wiking@huwico/staff/wiking] has quit [Read error: No route to host]17:02
-!- wiking [~wiking@huwico/staff/wiking] has joined #shogun17:02
-!- muddo [~muddo@gateway/tor-sasl/muddo] has quit [Ping timeout: 276 seconds]17:13
-!- muddo [~muddo@gateway/tor-sasl/muddo] has joined #shogun17:13
-!- muddo [~muddo@gateway/tor-sasl/muddo] has quit [Remote host closed the connection]17:22
-!- PhilTillet [~Philippe@npasserelle10.minet.net] has joined #shogun17:27
PhilTillethey hey hey!17:27
PhilTilletHello -ChanServ-17:28
@sonney2kPhilTillet, hi17:41
PhilTilletsonney2k, hi !17:41
@sonney2kPhilTillet, did you find some time to continue the benchmarking?17:42
PhilTilletthere's some progress :) I also have a conference to prepare :p I have done the gaussian kernel for what we talked about yesterday17:43
PhilTilletbased on the code of a matrix-matrix product17:43
@sonney2kwell you cannot really assume that the whole matrix is in memory but yeah sometimes it is17:44
PhilTilleti thought it worked to split the training set ?17:44
@sonney2kahh you mean the set of x_i ?17:46
@sonney2kyes that one is in there17:46
@sonney2kbut not the set of x (might be computed on demand)17:46
PhilTilletbut didn't you link a paper from ntu where it was explained?17:48
PhilTillethow to train when data cannot fit in memory17:48
-!- muddo [~muddo@gateway/tor-sasl/muddo] has joined #shogun17:50
@sonney2kPhilTillet, wait this is a totally different problem17:50
@sonney2kone thing is to apply a trained classifier (this is what you are parallelizing) the other is about training17:52
PhilTilletI see17:53
PhilTilletSo what do you do anyway on a CPU when training data cannot fit memory?17:54
@sonney2kstream the data from file and use anonline algorithm or sub-sample or ...17:55
@sonney2k(have to leave train)17:59
@sonney2kcu17:59
-!- blackburn [~qdrgsm@83.234.54.68] has joined #shogun18:09
-!- gsomix [~gsomix@188.168.2.59] has joined #shogun18:36
-!- harshit_ [~harshit@182.68.142.173] has joined #shogun19:19
-!- blackburn [~qdrgsm@83.234.54.68] has quit [Ping timeout: 265 seconds]19:40
-!- blackburn [5bde8018@gateway/web/freenode/ip.91.222.128.24] has joined #shogun19:40
-!- PhilTillet [~Philippe@npasserelle10.minet.net] has quit [Ping timeout: 252 seconds]19:53
-!- l0nr4n [~l0nr4n@unaffiliated/l0nr4n] has quit [Quit: This computer (or maybe me) has gone to sleep]20:05
harshit_hello20:41
-!- blackburn_ [5bde8018@gateway/web/freenode/ip.91.222.128.24] has joined #shogun20:46
-!- blackburn [5bde8018@gateway/web/freenode/ip.91.222.128.24] has quit [Ping timeout: 245 seconds]20:49
gsomixharshit_, hi :)20:55
harshit_Hey gsomix, hav a few questions, do you hav time ?20:56
gsomixharshit_, unfortunately I do not have time. before I had prepared for laboratory work in physics and is now going to bed.21:02
gsomixhowever, I wish you good night :)21:03
gsomixbb21:03
harshit_no worries .. hav a nice sleep21:03
blackburn_harshit_:21:04
-!- romi_ [~mizobe@187.57.1.50] has quit [Ping timeout: 246 seconds]21:05
harshit_hey blackburn_ : do you hav time ?21:05
blackburn_yes21:05
harshit_i am porting newton svm to shogun for a while and now its almost done21:05
harshit_Do i need to port the non linear version of newton svm too ?21:06
blackburn_it is up to you21:06
harshit_okay, and i am making use of  gmres function in lapack ,21:07
blackburn_looks like this gsoc idea should be removed once you finish it21:08
harshit_so can i make its wrapper in mathematics/lapack.h21:08
shogun-buildbotbuild #186 of nightly_none is complete: Success [build successful]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/nightly_none/builds/18621:08
harshit_i mean, am i allowed to push updates to mathematics/lapack.h21:10
harshit_Also wanted to discuss my deep learning idea .21:10
harshit_btw are you in the team of shogun ?21:11
-!- blackburn_ [5bde8018@gateway/web/freenode/ip.91.222.128.24] has quit [Ping timeout: 245 seconds]21:13
-!- PhilTillet [~Philippe@tillet-p42154.maisel.int-evry.fr] has joined #shogun21:16
-!- romi_ [~mizobe@187.101.162.14] has joined #shogun21:18
shogun-buildbotbuild #185 of nightly_all is complete: Success [build successful]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/nightly_all/builds/18521:26
-!- blackburn [5bde8018@gateway/web/freenode/ip.91.222.128.24] has joined #shogun21:33
-!- blackburn [5bde8018@gateway/web/freenode/ip.91.222.128.24] has quit [Ping timeout: 245 seconds]21:52
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun22:03
@sonney2kn4nd0, I meant the octave_modular interface22:05
@sonney2knot the octave_static one22:05
n4nd0sonney2k: I tried the example in that interface22:05
n4nd0sonney2k: maybe I wrote static, sorry, but I tried in modular interface22:06
@sonney2kn4nd0, what did you try?22:06
n4nd0sonney2k: octave_modular/classifier_lda_modular.m22:07
@sonney2kand what happened?22:07
n4nd0error: member not found22:07
n4nd0error: called from:22:07
n4nd0error:   classifier_lda_modular.m at line 25, column 122:07
-!- l0nr4n [~l0nr4n@unaffiliated/l0nr4n] has joined #shogun22:09
n4nd0sonney2k: any clue about that?22:11
@sonney2kI am compiling for octave_modular...22:12
n4nd0ok22:12
n4nd0what happens is that apply is not found in lda but I've no idea why22:12
@sonney2kIIRC it worked yesterday22:13
n4nd0maybe there is something wrong with my environment22:13
@sonney2kn4nd0, it just works here22:15
@sonney2kwhich version of swig and octave?22:15
n4nd0GNU Octave, version 3.2.422:16
@sonney2kand are you sure that you did make install as root?22:16
@sonney2klapack/atlas got also found?22:16
n4nd0SWIG Version 2.0.422:16
@sonney2ksame here22:16
@sonney2kn4nd0, does classifier_libsvm_modular work?22:17
n4nd0aham22:17
n4nd0look here22:17
n4nd0Checking for Atlas support ... no22:17
@sonney2kmaybe that is the culprit22:17
n4nd0facepalm for me ...22:17
n4nd0I also note this22:17
n4nd0Checking for Atlas support ... no22:17
n4nd0ups sorry22:18
n4nd0this one I meant22:18
@sonney2kinstall libatlas-dev  / libatlas-base-dev22:18
n4nd0Checking for octave-config ... no22:18
n4nd0is that an issue?22:18
@sonney2kn4nd0, octave-config is in octave3.2-headers pkg here22:19
@sonney2kI use this to determine link flags22:19
n4nd0Atlas is still not detected by configure, I have installed the packages you told me22:23
n4nd0maybe with libatlas-cpp-0.6-dev?22:23
-!- harshit_ [~harshit@182.68.142.173] has quit [Quit: Leaving]22:26
@sonney2kn4nd0, hmmhh check configure.log22:29
@sonney2kI have libatlas3gf-base installed too22:29
n4nd0sonney2k: oh, didn't know about this configure.log, thanks22:31
n4nd0sonney2k: yeah I had that package too, must be another thing, I have to check it22:31
-!- romi_ [~mizobe@187.101.162.14] has quit [Ping timeout: 246 seconds]22:31
@sonney2kthough that shouldn't be the issue - if you have lapack/blas it should work nevertheless22:32
n4nd0from the log I am reading clapack.h: No such file or directory22:34
n4nd0didn't expect that since I have been using blas from C++22:34
@sonney2kn4nd0, ohh looks like clapack.h moved from /usr/include to /usr/include/atlas22:43
@sonney2kso try ./configure --includes=/usr/includes/atlas22:44
-!- romi_ [~mizobe@187.57.11.40] has joined #shogun22:44
n4nd0sonney2k: Atlat support ... yes :)22:46
n4nd0let's see if this fixes the problem in octave22:46
@sonney2kn4nd0, don't forget to do make install as root (necessary for octave...)22:53
@sonney2kanyways I am off to bed. cu22:53
n4nd0sonney2k: it is working now :)22:53
@sonney2kgreat22:54
n4nd0sonney2k: thank you I'll check tom valgrind with octave then to check qda22:54
n4nd0sonney2k: it is time to sleep for me too22:54
n4nd0good night!22:54
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving]22:57
CIA-64shogun: Soeren Sonnenburg master * r1336edf / src/configure : add /usr/include/atlas detection upon configure - http://git.io/CVEklA22:59
shogun-buildbotbuild #552 of r_static is complete: Failure [failed test_1]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/r_static/builds/552  blamelist: sonne@debian.org23:09
shogun-buildbotbuild #574 of cmdline_static is complete: Failure [failed test_1]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/cmdline_static/builds/574  blamelist: sonne@debian.org23:12
shogun-buildbotbuild #553 of octave_static is complete: Failure [failed test_1]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/octave_static/builds/553  blamelist: sonne@debian.org23:16
-!- muddo [~muddo@gateway/tor-sasl/muddo] has quit [Ping timeout: 276 seconds]23:17
shogun-buildbotbuild #541 of python_static is complete: Failure [failed test_1]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/python_static/builds/541  blamelist: sonne@debian.org23:20
PhilTilletSo many bots !!!23:30
PhilTillet:o23:30
-!- muddo [~muddo@gateway/tor-sasl/muddo] has joined #shogun23:30
-!- romi_ [~mizobe@187.57.11.40] has quit [Ping timeout: 246 seconds]23:42
--- Log closed Fri Mar 23 00:00:19 2012