Open in new window / Try shogun cloud
--- Log opened Mon Jun 11 00:00:41 2012
n4nd0yay!! multiclass example with SO is working fine :D01:37
n4nd0someone out there to share my happiness?01:38
-!- n4nd0_ [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun01:42
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Ping timeout: 245 seconds]01:42
-!- n4nd0_ is now known as n4nd001:44
emreceliktenn4nd0: Great02:18
emrecelikten:P02:18
n4nd0:)02:23
n4nd0good night!02:30
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving]02:30
-!- ckwidmer [~chris@HSI-KBW-046-005-237-106.hsi8.kabel-badenwuerttemberg.de] has quit [Remote host closed the connection]03:26
-!- emrecelikten [~Anubis@176.40.238.218] has quit [Read error: Connection reset by peer]05:47
-!- puffin444 [472e31fb@gateway/web/freenode/ip.71.46.49.251] has quit [Quit: Page closed]07:13
-!- CIA-9 [~CIA@cia.atheme.org] has quit []07:50
-!- CIA-18 [~CIA@cia.atheme.org] has joined #shogun07:54
-!- CIA-18 [~CIA@cia.atheme.org] has quit [Excess Flood]07:59
-!- CIA-18 [~CIA@cia.atheme.org] has joined #shogun08:01
-!- CIA-18 [~CIA@cia.atheme.org] has quit [Excess Flood]08:02
-!- CIA-18 [~CIA@cia.atheme.org] has joined #shogun08:07
-!- gsomix [~gsomix@80.234.31.247] has joined #shogun08:12
-!- uricamic [~uricamic@2001:718:2:1634:7816:2ae9:adc4:a6da] has joined #shogun08:56
-!- gsomix_ [~gsomix@95.67.167.251] has joined #shogun08:57
-!- gsomix [~gsomix@80.234.31.247] has quit [Ping timeout: 256 seconds]09:01
-!- heiko [~heiko@host86-177-178-116.range86-177.btcentralplus.com] has joined #shogun10:11
-!- eric_ [2e1fd566@gateway/web/freenode/ip.46.31.213.102] has joined #shogun10:35
eric_hi all10:35
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun10:38
n4nd0sonney2k: around?10:40
-!- eric_ [2e1fd566@gateway/web/freenode/ip.46.31.213.102] has quit [Ping timeout: 245 seconds]11:28
-!- heiko [~heiko@host86-177-178-116.range86-177.btcentralplus.com] has quit [Ping timeout: 245 seconds]11:41
-!- heiko [~heiko@host86-176-2-162.range86-176.btcentralplus.com] has joined #shogun11:42
-!- jps [5536538f@gateway/web/freenode/ip.85.54.83.143] has joined #shogun11:47
-!- romi_ [~mizobe@187.57.4.69] has joined #shogun12:01
wikingah report day!12:15
n4nd0yeah \o/12:17
wikingbtw: currently shogun will not compile at all w/o lapack support12:43
wikingi've just tried it on a simple machine to clone my repo and test some stuff there and i've realized this..12:43
-!- jps [5536538f@gateway/web/freenode/ip.85.54.83.143] has quit [Quit: Page closed]12:47
-!- gsomix_ [~gsomix@95.67.167.251] has quit [Ping timeout: 265 seconds]13:07
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving]13:12
-!- heiko [~heiko@host86-176-2-162.range86-176.btcentralplus.com] has quit [Ping timeout: 245 seconds]13:43
-!- gsomix [~gsomix@178.45.79.164] has joined #shogun13:49
-!- eric____ [2e1fd566@gateway/web/freenode/ip.46.31.213.102] has joined #shogun14:13
eric____hi again14:14
eric____I would like to test smthg: compute the spectrum kernel from strings (i have CStrginFeatures<uint16_t>) with CommWordString kernel14:23
eric____whats the function to use (preproc?) which compute the k-mer frequ feature vector ?14:24
eric____following the doc, CommWordString use SortWordString preproc14:25
eric____and it says : "For this to work the strings have to be mapped into  * a binary higher order representation first (cf. obtain_from_*() functions in  * CStringFeatures)"14:25
eric____what exaclty does obtain_from_features ? It send me an error about num_symbols ..14:26
-!- heiko [~heiko@host86-177-4-176.range86-177.btcentralplus.com] has joined #shogun14:35
-!- wiking [~wiking@huwico/staff/wiking] has quit [Ping timeout: 248 seconds]14:41
-!- wiking [~wiking@208.76.55.196] has joined #shogun14:42
-!- wiking [~wiking@208.76.55.196] has quit [Changing host]14:42
-!- wiking [~wiking@huwico/staff/wiking] has joined #shogun14:42
eric____someone here ?14:50
heikome :)14:50
-!- emrecelikten [~Anubis@176.40.238.218] has joined #shogun14:55
eric____heiko: :)14:57
eric____heiko: could you help me to run CommWordString kernel ?14:58
heikoeric, maybe whats the problem with it?14:58
eric____I would like to know how to compute the ?-mer spectrum kernel fro example14:58
eric____2-mer14:59
eric____(i have CStrginFeatures<uint16_t>14:59
eric____docs advise to run preprocess SortWordString which just sort the string14:59
eric____I am wondering where is the function which compute the 2-mer frequency feature vector ??15:00
heikoI am not really into that stuff, however, lets see15:01
heikoso you just want to use this common word kernel?15:01
heikohave you tried the example for it?15:01
heikowhy do you want to compute a feature vector?15:02
eric____lets explain what I have understand15:03
eric____for now my samples are stringfeatures: string list of uint16_t15:03
eric____I would like to compute a spectrum kernel of 2-mer15:04
heikook15:04
eric____and as the docs says : in this feature vector each entry denotes how often the k-mer appears in that string15:05
eric____so there is some preproc to do I think15:05
heikowhere is that written in the doc?15:05
heikocould you send a link so that I can see the context?15:05
eric____I have seen obtain_from_features method in CStringFeatures but I dont know how to use it .15:05
eric____http://www.shogun-toolbox.org/doc/en/current/classshogun_1_1CWeightedCommWordStringKernel.html15:06
eric____or http://www.shogun-toolbox.org/doc/en/current/classshogun_1_1CCommWordStringKernel.html for simple k fixed version15:06
heikook15:07
heikofirst: you usually dont compute fetaure vectors directly, this is just implicitly. If you *would* compute feature vectors, they had this described structure15:09
heikoimagine you use 10-mers over an alphabet with 10 letters15:09
heikothats 10^10 dimensions15:09
heikothis sorting just means, that in order to use these weights, you would have to sort the data that you attach to the kernel15:10
-!- eric____ [2e1fd566@gateway/web/freenode/ip.46.31.213.102] has quit [Ping timeout: 245 seconds]15:10
-!- K0stIa [~kostia@2001:718:2:1634:76f0:6dff:fe92:4013] has joined #shogun15:11
heikothe obtain_from_* methods are just to convert string features/char features to numerical representations15:13
heikoI would suggest to make the example run and then experiment a bit with it15:13
K0stIahi all, I have problem with installing shogun from github code. I'm doing ./configure and then make. And my make fails, it says15:14
K0stIastatistics/QuadraticTimeMMD.cpp:164:34: error: 'compute_eigenvectors' is not a member of 'shogun::SGMatrix<double>'15:14
K0stIacan anyone say smth about it ?15:14
heikoHi K0stIa15:14
heikoyes, I will have a look, just a second15:14
-!- vojtech [9320543b@gateway/web/freenode/ip.147.32.84.59] has joined #shogun15:15
heikoK0stIa, are you using the current git ?15:15
heikowhen did you clone/update last?15:16
heikoah I think I know whats wrong ...15:16
heikoyou can fix that by installing lapack libraries but I will also add a fix15:17
CIA-18shogun: Heiko Strathmann master * r27f5fd1 / (2 files): added lapack check for eigenspectrum based test to solve compile error - http://git.io/C5xsmQ15:24
CIA-18shogun: Heiko Strathmann master * r1532e15 / (2 files): Merge pull request #579 from karlnapf/master - http://git.io/dli_Ng15:24
CIA-18shogun: Heiko Strathmann master * r0e9be8c / src/shogun/statistics/QuadraticTimeMMD.cpp : another check for LAPACK - http://git.io/ziupmw15:24
heiko K0stIa, try with current git code, I added a fix if LAPACK is not installed15:24
-!- eric____ [2e1fd566@gateway/web/freenode/ip.46.31.213.102] has joined #shogun15:59
eric____heiko: re15:59
-!- alexlovesdata [c25faea9@gateway/web/freenode/ip.194.95.174.169] has joined #shogun16:00
-!- eric____ [2e1fd566@gateway/web/freenode/ip.46.31.213.102] has quit [Ping timeout: 245 seconds]16:03
-!- K0stIa [~kostia@2001:718:2:1634:76f0:6dff:fe92:4013] has left #shogun []16:32
-!- K0stIa [~kostia@2001:718:2:1634:76f0:6dff:fe92:4013] has joined #shogun16:53
K0stIa@heiko, thnx, now it works :)16:53
-!- vojtech [9320543b@gateway/web/freenode/ip.147.32.84.59] has quit [Quit: Page closed]17:01
heikoK0stIa, nice :)17:18
heikoeric, your connection is even worse than mine :)17:18
-!- wiking_ [~wiking@78-23-189-112.access.telenet.be] has joined #shogun17:22
-!- wiking_ [~wiking@78-23-189-112.access.telenet.be] has quit [Changing host]17:22
-!- wiking_ [~wiking@huwico/staff/wiking] has joined #shogun17:22
-!- wiking [~wiking@huwico/staff/wiking] has quit [Ping timeout: 256 seconds]17:25
-!- wiking_ is now known as wiking17:25
-!- emrecelikten [~Anubis@176.40.238.218] has quit [Quit: Leaving.]17:27
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun17:30
-!- blackburn [d55781be@gateway/web/freenode/ip.213.87.129.190] has joined #shogun17:30
-!- K0stIa [~kostia@2001:718:2:1634:76f0:6dff:fe92:4013] has left #shogun []17:40
-!- blackburn [d55781be@gateway/web/freenode/ip.213.87.129.190] has quit [Quit: Page closed]17:40
-!- eric [2e1fd566@gateway/web/freenode/ip.46.31.213.102] has joined #shogun17:47
-!- eric is now known as Guest5786817:47
-!- blackburn [~blackburn@188.168.3.9] has joined #shogun17:50
-!- uricamic [~uricamic@2001:718:2:1634:7816:2ae9:adc4:a6da] has quit [Quit: Leaving.]17:52
-!- romi_ [~mizobe@187.57.4.69] has quit [Ping timeout: 260 seconds]18:03
-!- eric____ [2e1fd566@gateway/web/freenode/ip.46.31.213.102] has joined #shogun18:08
-!- Guest57868 [2e1fd566@gateway/web/freenode/ip.46.31.213.102] has quit [Ping timeout: 245 seconds]18:09
-!- emrecelikten [~emrecelik@176.40.238.218] has joined #shogun18:09
-!- eric____ [2e1fd566@gateway/web/freenode/ip.46.31.213.102] has quit [Ping timeout: 245 seconds]18:15
blackburnheiko: one trick - you may pull from fork w/o pull request18:22
heikoblackburn, ok, how? :)18:23
blackburnah you have no clone of shogun18:23
blackburnthen no way :)18:24
heikono :)18:25
heikoIf it annoys you I can do that18:25
blackburnnoo18:25
blackburnI just mean would be faster probably18:25
blackburnfor you18:25
blackburn3 more reports to go18:26
heikoreports?18:27
heikothe way I am doing it is also fast :) I always get confused if there is too much stuff18:28
heikoah weekly18:29
blackburnyeah18:31
n4nd0blackburn: I am doing my soon, I promise :)18:33
n4nd0blackburn: I am leaving my repo clean of new code first18:33
blackburnheh18:34
blackburnI have seen you get multiclass example to work alreadY?18:34
n4nd0blackburn: yeah!18:36
n4nd0blackburn: I have to use bigger datasets though18:38
n4nd0Nico told me that the evaluation license of MOSEK has an expiration date18:38
n4nd0so I could just solve very small problems18:38
blackburnbaad18:39
n4nd0but if I get a new one I will be able to solve big ones18:39
blackburnis there any chance to use that bmrm stuff?18:39
blackburnI am not into SO actually you know :)18:39
n4nd0blackburn: to use it for what do you mean?18:40
blackburninstead of mosek18:40
n4nd0blackburn: ah ok18:41
n4nd0they are different things18:41
n4nd0so we use mosek for the optimization algorithm18:41
n4nd0bmrm is the training algorithm18:41
n4nd0it relies in a optimizer as well18:41
n4nd0but it is a different one and for bmrm libqp is enough18:42
n4nd0at least I think it is like this, I'm not 100% sure18:42
blackburnaham18:42
-!- emrecelikten [~emrecelik@176.40.238.218] has left #shogun []18:58
@sonney2kn4nd0, so what was the issue? what did you have to change to get it to work?19:08
@sonney2kheiko, thanks for the fix19:08
@sonney2kheiko, I guess you have seen that I reworked CStatistics/CMath/SGVector/SGMatrix...19:08
@sonney2kI think it is much better now19:09
heikoyes, my examples broke :)19:09
heikoI agree19:09
n4nd0sonney2k: the issue I wanted to tell you about is related to modular interface, I cannot make work my SO there19:09
heikomuch nicer19:09
n4nd0sonney2k: but I have not fixed it yet19:09
@sonney2kheiko, ahh and please keep the scheme w/ doing pull requests and then merges19:09
@sonney2kI think I should do the same19:09
n4nd0sonney2k: I did the example in libshogun interface19:09
heikosonney2k, well, ok :) it feels more comfortable for me anyway19:09
heikoand this way I dont break the main repo19:09
heiko(which I sometimes manage to do with my fork :)19:10
@sonney2kn4nd0, so waht is the issue w/ modulear here?19:10
blackburnis BFGS any cool for svms?19:10
blackburnany evidences?19:10
@sonney2kthe buildbot is back to live19:10
@sonney2kshogun-buildbot, you rock!19:10
shogun-buildbotWhat you say!19:10
@sonney2kn4nd0, I got positive feedback wrt mosek btw19:11
n4nd0sonney2k: what feedback?19:12
n4nd0sonney2k: about modular19:14
n4nd0sonney2k: TypeError: in method 'StructuredOutputMachine_set_model', argument 2 of type 'CStructuredModel *'19:14
@sonney2kn4nd0, that we will get licenses19:14
n4nd0sonney2k: I get errors like this19:14
blackburnn4nd0: didn't you fix that?19:14
n4nd0sonney2k: it seems that it doesn't recognize the inheritance properly19:14
n4nd0blackburn: no :O19:14
blackburnokay lets try to get rid of it19:15
@sonney2kn4nd0, that is easily fixed - just put a class CStructuredModel; in the header19:15
n4nd0blackburn: the last we said about it was ... we'll ask sonney2k19:15
blackburnyes I recall19:15
blackburniirc it ^ didn't helped?19:15
n4nd0sonney2k: I tried with that, blackburn told me about it19:15
@sonney2kn4nd0, that will help19:15
n4nd0but it didn't make the trick19:15
n4nd0I'll try again then19:15
@sonney2kit will19:15
@sonney2kif not show me the changed header19:16
n4nd0sonney2k: ok, give me some minutes, I have just rebased and I have to adapt the changes done in CMath19:16
@sonney2kfor some reason swig forgets that we are in shogun:: namespace if we don't add that everywhere :(19:16
blackburninheritance problems emerge essentially when you have wrong order of includes19:16
@sonney2kn4nd0, k19:16
@sonney2kblackburn, it is not an inheritance problem here19:17
blackburnsome kind of19:17
blackburnhowever yes rather polymorphism :D19:17
@sonney2kblackburn, today I got the galaxy s319:23
@sonney2konly to realize that it needs a micro sim card19:24
blackburncool19:24
@sonney2kso I had to cut my sim card with a scissors19:24
blackburnhowever I still prefer my robust phone19:25
n4nd0seriously?19:25
n4nd0the scissors thing19:25
@sonney2kyes19:25
@sonney2kit works19:25
n4nd0hahaha19:25
blackburnwhy not19:25
n4nd0c'mon, it sounds funny19:26
@sonney2kn4nd0, easier than transitioning CLabels :D19:26
n4nd0sonney2k: :D so the CStructuredModel I should include it in the child class right?19:26
@sonney2k?19:27
@sonney2kn4nd0, in the header of StructuredOutputMachine19:27
n4nd0sonney2k: ah ok19:27
@sonney2kyou put class CStructuredModel; directly after the namespace shogun {19:27
n4nd0sonney2k: yeah, I just thought that I had to do that in the classes that extend CStructuredModel19:28
@sonney2kn4nd0, in every class that needs a CStructuredModel as input19:28
n4nd0sonney2k: ok got it, thanks! I am going to try it19:28
n4nd0sonney2k: about the mosek licenses19:29
n4nd0sonney2k: nico told me that the evaluation ones expire after three months19:29
n4nd0sonney2k: but that I should be able to get a new evaluation one19:30
n4nd0sonney2k: that has all the capabilities for yet three more months19:30
wiking  n4nd0 yes it does19:42
wikingn4nd0: if u want i can send u a new license19:42
wikingi needed a new one as well19:42
n4nd0sonney2k: all right, you fixed my problem with modular interface :)19:42
CIA-18shogun: Sergey Lisitsyn master * rf5163e8 / (7 files in 2 dirs): Merge branch 'multiclass' of git://github.com/pluskid/shogun (+5 more commits...) - http://git.io/hn3mxQ19:45
@sonney2kn4nd0, dont' ask how long it took me to find the workaround/solution19:46
n4nd0:O19:47
CIA-18shogun: Sergey Lisitsyn master * r90641c6 / src/shogun/lib/slep/SpInvCoVa/invCov.cpp : Made sparse inverse covariance not that verbose - http://git.io/5RFMOw19:48
CIA-18shogun: Sergey Lisitsyn master * r7cd8964 / examples/undocumented/python_modular/mathematics_sparseinversecovariance_modular.py : Added Sparse Inverse Covariance example - http://git.io/amEXRg20:06
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Ping timeout: 248 seconds]20:08
CIA-18shogun: Sergey Lisitsyn master * rdf58c63 / examples/undocumented/python_modular/mkl_linadd_regression.py : Dropped linadd mkl test example - http://git.io/6NxhqQ20:09
-!- vikram360 [~vikram360@117.192.187.238] has joined #shogun20:55
-!- vikram360 [~vikram360@117.192.187.238] has quit [Ping timeout: 245 seconds]21:00
-!- alexlovesdata [c25faea9@gateway/web/freenode/ip.194.95.174.169] has quit [Quit: Page closed]21:02
-!- vikram360 [~vikram360@117.192.170.22] has joined #shogun21:04
-!- vikram360 [~vikram360@117.192.170.22] has quit [Ping timeout: 245 seconds]21:11
shogun-buildbotbuild #605 of csharp_modular is complete: Success [build successful]  Build details are at http://www.shogun-toolbox.org/buildbot/builders/csharp_modular/builds/60521:15
-!- gsomix [~gsomix@178.45.79.164] has quit [Quit: Ex-Chat]21:52
@sonney2kok sth killed the buildbot again hmmhh22:01
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun22:02
n4nd0wiking: hey! do you have a moment now?22:03
wikingyeah sure22:03
wikingshoot22:03
wikingu need a new license?22:03
n4nd0wiking: it's about SO framework22:03
wikingok22:03
n4nd0wiking: I think I can manage to download a license22:03
wikingwell i can just fwd u again as last time22:04
wikingso that you don't have to ask for a new one22:04
n4nd0wiking: ok then22:04
n4nd0wiking: did you get a new one?22:04
wikingyeps22:04
wikingwhat was the email again22:04
wiking?22:04
wikingdone22:05
wikingshould be good for 2 more months22:05
n4nd0wiking: so Nico found some flaws about having the StructuredModel separated into different classes22:06
wikingaaah22:06
wikinghit me22:06
n4nd0one for the argmax, another for the psi, etc22:06
wikingyep the one we've talked about last week22:06
n4nd0the problem is that they normally rely on each other and it's difficult to make them fully separated22:06
blackburnoh 8/8, thanks guys!22:07
wikinghehehe blackburn :)22:07
wikingn4nd0: ok i get it that they are relying on each other22:07
wikingbut still why would it be difficult to separate them?22:07
n4nd0wiking: I am not sure exactly about the reasons22:08
n4nd0wiking: but what I found is that for example22:08
n4nd0if the argmax needs the loss and joint feature function22:08
n4nd0then you have with the argmax a class that contains the same logic as the whole model22:08
n4nd0and it loses a bit of sense22:09
wikingoh22:09
wikingi see what you mean22:09
n4nd0:)22:09
wikingthat we might end up22:09
wikinghaving a class that contains a class that contains the the parent class22:09
wikingright?22:09
wikingA has B has A22:09
wikingor?22:10
n4nd0mmm I don't think that problem really arises22:10
n4nd0but it would be more like22:10
n4nd0you have the Model that contains the argmax, the loss and psi22:11
n4nd0but argmax needs loss and psi too22:11
wikingblackburn: btw: russia will win this eurocup22:11
wikingn4nd0: yep that's more or less what i'm saying22:11
n4nd0wiking: ok then22:11
wikingthat you have class A that has B, C,D22:11
wikingand B will have C, D as well22:11
n4nd0exactly22:11
wikingc = loss, d = psi22:11
n4nd0that was the problem I found, but I think that Nico had other reasons too22:12
wikingwell just pass the reference to the parent?22:12
wikingn4nd0: was it discussed here?22:12
wikingi.e. can i find it in irc logs?22:12
n4nd0wiking: yes, I'll send you the link now22:12
wikingok cool22:13
n4nd0wiking: but even if you pass the reference to the parent22:13
n4nd0wiking: the thing would be that the model or the argmax loses its meaning22:13
wikingwell not necessarily22:13
wikingi mean if u need psi and loss for argmax22:13
wikingthen u need it22:14
wikingthus it's still argmax22:14
wikingthe only question is how do u have the reference on these22:14
n4nd0wiking: http://www.shogun-toolbox.org/irclogs/%23shogun.2012-06-07.log.html the converation about it starts around 15:4522:15
wikingok22:15
wikingreading/parsing22:15
wiking"latent guys"22:17
wiking:D22:17
n4nd0:)22:17
n4nd0people cannot see you22:18
wiking"nicococoexample: for latent svms i think we also need a delta(y_i,y_bar,h_latent)"22:18
wikingthat's not true22:18
n4nd0it's just the joint feature right?22:18
wikingno i mean22:18
wikingi don't need that delta22:18
wikingonly delta(y, y_bar)22:18
n4nd0yeah ok, I meant like it's the joint feature the only one that accounts for h22:19
wikingmmm22:21
wikingi've read it22:21
wikingtilld 16:5722:21
wiking*till22:21
n4nd0so? would you still like more to have the division into classes?22:21
n4nd0CArgMax, etc22:21
wikingwell22:21
wikingfor me that makes my life easier22:21
wikingbut then again22:21
wikingif u are strongly against it22:22
n4nd0I think that to have just a CStructuredModel and extend that one is good22:22
wikingthen i guess we'll just have to live with that22:22
wikingand then yeah22:22
wikingjust derive my own stuff from CStructuredModel22:22
n4nd0why do you think it makes your life easier for you?22:22
wikingbecause i don't have to care about anything else in CStructuredModel22:23
wikingjust about psi and argmax ;)22:23
n4nd0wiking: https://github.com/iglesias/shogun/blob/so/src/shogun/so/MulticlassModel.h take a look here22:23
wikingexactly22:23
n4nd0so the plan would be to add an additional get_joint_feature_vector22:23
n4nd0for latent22:23
wikingthis is complicated ;)22:23
n4nd0that gets int32_t, StructuredData and another int32_t22:23
wikingi mean if u think about that i only want to define a psi and an argmax function22:23
wikingthis is way too complicated22:23
n4nd0but you will need to do just that22:24
wikingand then each time22:24
wikingactually if u wanna be fair22:24
wikingyou'll need to override const char* get_name() const  function as well22:24
wikingsince the given class is a new one... ;)22:24
n4nd0aned get_dim22:24
wikingso it's just weird in my head22:25
wikingthat for a simple 2 things22:25
wikingactually 2 functions22:25
wikingi have to override a whole class22:25
wikingand ok i get why shouldn't we use function pointers... i.e. swig22:25
n4nd0yes, I mean the reasons you are telling me are just saved using function pointers22:26
n4nd0and we cannot do that22:26
wikinghehe yeah22:26
wikingbut we could do still structs or classes ;)22:26
wikingbut then again22:26
n4nd0so we can either use a class for each thing (for argmax, for delta, for psi)22:27
n4nd0or a class where we can put everything together22:27
wikingif u strongly feel about this and as far as i can see nico is quite assertive about this22:27
wikingthen go with a one class option22:27
n4nd0exactly22:27
n4nd0this is what we are going for22:27
wikingbut i do not agree ;)22:27
n4nd0haha22:27
wikingbut it's your code and your stuff22:27
wikingso i do not want to interfere too much :)22:28
n4nd0I do not agree there :)22:28
n4nd0I mean, it is my code ok22:28
wikingbut then again i'm gonna use it22:28
wikingand i want to make my life easier :P22:28
n4nd0but it is the "only" option we have22:28
n4nd0because we cannot use function pointers22:28
wikingfriend classes :)22:28
n4nd0and discarding them, the option I choose is the one class option22:28
n4nd0I suggested that quite at the beginning, but blackburn and sonney2k disagreed about them :S22:29
wikinghahahhaha22:29
wikingu see22:29
blackburnwhat did I diagree?22:29
wikingblackburn: baaastard!22:29
n4nd0using friend classes22:29
blackburn:D22:29
blackburnhmm sorry I am lazy to read your discussion - is there any problem related with not having friends?22:30
blackburnnot for real but in your classes I mean22:30
blackburn:D22:30
wikingblackburn: :>22:30
wikinghahahahah https://fbcdn-sphotos-a.akamaihd.net/hphotos-ak-ash3/547749_472264696132937_557005192_n.jpg22:30
n4nd0we don't even get to install the 10% in Sweden ...22:31
wikingblackburn: well using friend for classes would make our lives easier :)22:31
blackburnhaha22:31
wikingn4nd0: :>22:31
blackburnI think being flexible is cool so if you really think friends would make it better do it22:32
n4nd0wiking: how would you like to use them in any case?22:32
wikingn4nd0: well have argmax as a friend class in structuredmodel22:34
wikingso anything u have in structuredmodel you'll be able to call from argmax22:34
n4nd0wiking: but still, you have to make a class for argmax22:34
wikingand why's that a problem in that case?22:34
n4nd0and you would do this in order to avoid doing a class child of structuredmodel22:34
wikingyes because argmax class would just do the calculation of argmax22:35
wikingthus if you inherit from it22:35
wikingu want to change the argmax calculation22:35
wikingand nothing else22:35
wikingnot like in case of structuredmodel22:35
wikingif u inherit from that22:35
n4nd0you can inherit from structuredmodel and just do that too22:35
wikingu either change argmax or psi22:35
wikingand here or is not exclusive22:35
wikingand u might not want to change anything22:36
wikingjust argmax22:36
n4nd0you can do that as it is right now22:36
wikingyes22:36
n4nd0then?22:36
wikingby inheriting from structuredmodel22:36
wikingright?22:36
n4nd0yes22:36
blackburnn4nd0: bahh 2-122:36
n4nd0blackburn: :(22:36
n4nd0blackburn: I am suffering it22:36
wikingblackburn: sweden loose?22:36
n4nd0blackburn: but to tell the truth ... they really suck22:37
blackburnyeah22:37
wikingyeah cool22:37
wikingfinally some fun :)22:37
n4nd0yeah, against Ukraine22:37
blackburnI am not watching it just read in twitter :)22:37
n4nd0Ukraine seems a good team so imagine ....22:37
n4nd0:D22:37
wikingn4nd0: so the thing is that i'm lazy.. so i don't want to reimplement a lot of things, just because of one function change :P22:37
n4nd0wiking: yeah, I understand that22:38
wikingit's going to be russia i have 10 euros on that one :)22:38
n4nd0wiking: but my point is that you don't need to do it with CStructuredModel either22:38
blackburnn4nd0: well they have pretty strong teams in national league22:38
wikingn4nd0: currently i can only change that way22:38
blackburnand well, they have shevchenko22:38
wikingor?22:38
n4nd0wiking: yeah22:39
n4nd0wiking: you can extend the model and just implement the argmax if you want22:39
wikingextend u mean class Whatever : public CStructuredModel22:39
wikingright?22:39
n4nd0yes22:39
n4nd0sorry ... java naming :P22:39
wikingok that is a huge overhead22:40
wikingfor a simple functional change :P22:40
wikingimho22:40
wiking:)22:40
wikingbecause this will require me22:40
wikingdefining ctors and dtors22:40
wikingcalling the parent ctors22:40
wikingbut then again22:41
wikingyeah it'd be easier in case of using structs22:41
wiking;)22:41
wikingso argmax is a struct22:41
wikinghttp://www.swig.org/Doc1.3/SWIG.html#SWIG_nn3022:42
wikingwtf is this?22:42
wiking:D22:42
n4nd0haha ok22:42
n4nd0if you want to fight for function pointers then sonney2k is your man not me :)22:43
wikingno worries22:43
wikingthat'll never happen22:43
@sonney2kexactly :D22:43
wikingok go with the implementation that u and nico agreed on22:43
* wiking knows that there's no way to convince sonney2k of anything that he hasn't agreed with from the beginning :P22:44
n4nd0sonney2k: if it could work with SWIG, is it just a problem with the documentation?22:44
wikingn4nd0: "Occasionally, a C library may include functions that expect to receive pointers to functions--possibly to serve as callbacks. SWIG provides full support for function pointers provided that the callback functions are defined in C and not in the target language."22:44
n4nd0wiking: yeah, I read that22:44
wikingso one could not define a function for argmax in python22:44
@sonney2kwiking, sure I read that doc22:45
@sonney2kbut have you seen how much effort it is?22:45
wikingand pass it22:45
wikingsonney2k: it's bs22:45
@sonney2kyou will need to add some rules for every function ptr22:45
n4nd0wiking: for that we would use director classes22:45
wikingsonney2k: so i'm really not pushing this agenda :P22:45
wikingi mean it's really a nightmare22:45
@sonney2kexactly22:46
@sonney2kI have no idea how directors (which are crazy enough already) would work with this stuff22:46
blackburnI admit I was pretty near to commit code containing for_each and std::list day before yesterday22:47
n4nd0blackburn: std rebellion!!22:47
n4nd0I vote for it22:48
wikingblackburn: :DDD22:48
blackburnhowever later I have changed it to CList :(22:49
@sonney2kwhich is a good idea if you want to have serialization22:50
@sonney2kfinally apple releases a high-resolution macbook :D22:52
@sonney2k2880x1800 pixel - yay!22:52
wiking:D22:53
@sonney2kkeeping font size whole shogun fits on one screen :D22:53
wikinghahah and it can handle more than 8 gigs of ram22:53
blackburnhttp://www.tresnormale.com/components/com_virtuemart/shop_image/product/fb45191b411f14614b63ea9368cac7bf.jpg22:53
blackburnsonney2k: 2.3K$, right?22:54
@sonney2kI guess I need to go back to academia to get it funded :D22:55
blackburnhahha22:55
blackburngood reason22:55
wikingloooool22:55
@sonney2kor I have to blackmail any of you ;-)22:55
n4nd0lol22:55
wikingin KL everything was a lot cheaper than here in europe :)22:56
blackburnKL?22:56
wikingmac stuff at least by 200 euros22:56
wiking(malaysia)22:56
blackburnah22:56
wiking1715 euros vs 2300 euros for the 15-inch: 2,3-GHz Retina-display22:57
wiking:>22:57
wikingthat's almost a whooping 600 euros22:58
blackburnit will cost 3.5K here I think22:58
blackburnw/o ebay :D22:58
@sonney2kI guess the display is really pricey... >1k22:59
wikinglol wtf22:59
wikingwhy is there 2 gnu in that machine?22:59
wikingIntel HD Graphics 4000 and a NVIDIA GeForce GT 650M?22:59
wiking*gpu23:00
blackburnit is something common now23:00
wikingreally? cool i guess u can opencl the shit out of those )23:00
wiking:D23:00
blackburnshit I have exam day after tomorrow23:01
blackburnproblem is that it will be about everything :D23:05
@sonney2kwiking, yeah even my mac has this23:05
@sonney2kand it is 4 years old...23:05
n4nd0do you like mac guys?23:06
n4nd0I mean if you like mac computers ... it was a bit ambiguous :P23:06
wikingsonney2k: not mine.... :S23:06
blackburnI have good old samsung23:07
@sonney2kn4nd0, I do - excellent hardware. only the os is crap :D23:08
n4nd0sonney2k: how do manage with linux on it?23:08
n4nd0lot of problems?23:09
@sonney2kn4nd0, I am running linux on macs since 8 years or so23:09
@sonney2kfor 8 years :D23:09
@sonney2kback then (titanium powerpc based mac)23:09
@sonney2kthere was no other notebook that could handle 5 hours battery time23:09
@sonney2klong before intel built reasonable mobile cpus23:10
-!- vikram360 [~vikram360@117.192.161.30] has joined #shogun23:10
@sonney2kahh and I could hot plug a battery once the other one got empty23:10
n4nd0the battery in my computer sucks so badly...23:10
@sonney2kso I had ~8 hours of battery and that is how shogun became to existence23:11
n4nd0nice23:11
@sonney2klooong train rides23:11
n4nd0:)23:11
n4nd0did you code most of it in the train?23:11
@sonney2kn4nd0, back then almost everything23:12
n4nd0haha23:12
blackburndeutsche railroad should be in (C)23:12
@sonney2ksponsord by deutsche bahn delays :D23:12
blackburnsonney2k: did you live in poland? why so long train rides?23:12
n4nd0did people look what you were doing or something?23:12
blackburn:D23:12
@sonney2kothers where having windows notebooks23:13
@sonney2kI got asked a lot how long these macs can run on battery ...23:13
blackburnbumps while riding should determine bad formatting at some places23:13
n4nd0haha23:14
@sonney2kblackburn, I can assure you that we don't have jumping ICE trains :D23:16
blackburnah poor you have no bumpy roads :(23:17
@sonney2kcrazy!23:22
@sonney2kon this galaxy s3 the led color is fully programmable23:22
@sonney2keach app can have its own notification color...23:22
-!- heiko [~heiko@host86-177-4-176.range86-177.btcentralplus.com] has quit [Quit: Leaving.]23:48
--- Log closed Tue Jun 12 00:00:41 2012