--- Log opened Sun Mar 29 00:00:19 2015 | ||
-!- curiousguy13 [~curiousgu@120.59.99.232] has quit [Ping timeout: 255 seconds] | 04:36 | |
-!- curiousguy13 [~curiousgu@120.59.100.116] has joined #shogun | 04:49 | |
-!- pickle27 [~pickle27@192-0-134-235.cpe.teksavvy.com] has joined #shogun | 04:58 | |
-!- pickle27 [~pickle27@192-0-134-235.cpe.teksavvy.com] has quit [Remote host closed the connection] | 07:16 | |
-!- x4i [~A2o@unaffiliated/x4i] has joined #shogun | 10:15 | |
-!- besser82 [~besser82@p2003007A8D650F00F2DEF1FFFE8942D4.dip0.t-ipconnect.de] has joined #shogun | 10:22 | |
-!- besser82 [~besser82@p2003007A8D650F00F2DEF1FFFE8942D4.dip0.t-ipconnect.de] has quit [Changing host] | 10:22 | |
-!- besser82 [~besser82@fedora/besser82] has joined #shogun | 10:22 | |
-!- mode/#shogun [+o besser82] by ChanServ | 10:22 | |
-!- curiousguy13 [~curiousgu@120.59.100.116] has quit [Ping timeout: 248 seconds] | 10:36 | |
-!- x4i [~A2o@unaffiliated/x4i] has quit [Quit: irc] | 10:44 | |
-!- curiousguy13 [~curiousgu@120.57.232.169] has joined #shogun | 10:49 | |
-!- nszceta [~nszceta@p4FC8D770.dip0.t-ipconnect.de] has joined #shogun | 12:52 | |
-!- nszceta [~nszceta@p4FC8D770.dip0.t-ipconnect.de] has quit [Quit: Textual IRC Client: www.textualapp.com] | 13:24 | |
-!- HeikoS [~heiko@0543fcaf.skybroadband.com] has joined #shogun | 13:29 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 13:29 | |
-!- vortex_ape [~vortex_ap@59.177.142.50] has joined #shogun | 13:51 | |
-!- nileshK [uid75191@gateway/web/irccloud.com/x-dqeidlpjndlozdle] has quit [Quit: Connection closed for inactivity] | 14:09 | |
@HeikoS | lisitsyn: heya | 14:25 |
---|---|---|
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun | 14:26 | |
shogun-notifier- | shogun: Wu Lin :develop * 7c758d4 / src/shogun/ (10 files): https://github.com/shogun-toolbox/shogun/commit/7c758d4580ac4638fd4535c5952b978ee3c9aad9 | 14:26 |
shogun-notifier- | shogun: minor fix | 14:26 |
shogun-notifier- | shogun: Heiko Strathmann :develop * 768671c / src/shogun/ (10 files): https://github.com/shogun-toolbox/shogun/commit/768671cd34d0da24f88c85460da9bb36d8a473cc | 14:26 |
shogun-notifier- | shogun: Merge pull request #2785 from yorkerlin/develop | 14:26 |
shogun-notifier- | shogun: | 14:26 |
shogun-notifier- | shogun: minor fix | 14:26 |
-!- travis-ci [~travis-ci@ec2-54-235-29-7.compute-1.amazonaws.com] has joined #shogun | 14:28 | |
travis-ci | it's Heiko Strathmann's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: http://travis-ci.org/shogun-toolbox/shogun/builds/56294227 | 14:28 |
-!- travis-ci [~travis-ci@ec2-54-235-29-7.compute-1.amazonaws.com] has left #shogun [] | 14:28 | |
shogun-buildbot | build #259 of trusty - libshogun - viennacl is complete: Failure [failed test] Build details are at http://buildbot.shogun-toolbox.org/builders/trusty%20-%20libshogun%20-%20viennacl/builds/259 blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>, Wu Lin <yorker.lin@gmail.com> | 14:31 |
-!- nileshK [uid75191@gateway/web/irccloud.com/x-qtmjgmermnyrcnrq] has joined #shogun | 14:35 | |
nileshK | HeikoS: Hi | 14:35 |
@HeikoS | nileshK: hi | 14:35 |
@HeikoS | nileshK: how are things with the lpboost? | 14:35 |
nileshK | Yeah. I trying to get something working on lpboost | 14:36 |
nileshK | Since now I have got it compiling, I am trying to figure out how to create a toy example | 14:36 |
nileshK | which will just act a proof of concept and add it to the examples set | 14:37 |
@HeikoS | okok | 14:37 |
@HeikoS | nileshK: the error you got seems to be of a different source | 14:37 |
nileshK | "different source"? I dont think I understand this | 14:38 |
nileshK | Also, I needed your help on one more thing | 14:40 |
nileshK | Look at this: ui/GUIClassifier.cpp. This thing tries to use the lpboost | 14:40 |
shogun-notifier- | shogun: Sergey Lisitsyn :develop * 4e5f194 / / (23 files): https://github.com/shogun-toolbox/shogun/commit/4e5f194853e04e463efdfb67f692cd1841dac9ea | 14:41 |
shogun-notifier- | shogun: Moved hashed features into separate directory | 14:41 |
shogun-notifier- | shogun: Sergey Lisitsyn :develop * ecc6f4d / src/shogun/ (4 files): https://github.com/shogun-toolbox/shogun/commit/ecc6f4da161aa4f8bf8baeb28b2bbc8e1e6ddcd6 | 14:41 |
shogun-notifier- | shogun: Fixed some broken include paths | 14:41 |
shogun-notifier- | shogun: Sergey Lisitsyn :develop * aaaaaf3 / tests/unit/ (5 files): https://github.com/shogun-toolbox/shogun/commit/aaaaaf3c1791c1dc7a5fde9ca3773cf2c2bd55f5 | 14:41 |
shogun-notifier- | shogun: Adjusted paths to hashed features in unit tests | 14:41 |
shogun-notifier- | shogun: Sergey Lisitsyn :develop * 060bb0f / / (32 files): https://github.com/shogun-toolbox/shogun/commit/060bb0f8789641a1af4689113ba2d8cba297373e | 14:41 |
shogun-notifier- | shogun: Merge pull request #2739 from lisitsyn/refactor/hashed_to_separate_dir | 14:42 |
shogun-notifier- | shogun: | 14:42 |
shogun-notifier- | shogun: Moved hashed features into separate directory | 14:42 |
shogun-buildbot | build #3375 of deb1 - libshogun is complete: Failure [failed compile] Build details are at http://buildbot.shogun-toolbox.org/builders/deb1%20-%20libshogun/builds/3375 blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com> | 14:43 |
shogun-buildbot | build #260 of trusty - libshogun - viennacl is complete: Failure [failed compile] Build details are at http://buildbot.shogun-toolbox.org/builders/trusty%20-%20libshogun%20-%20viennacl/builds/260 blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com> | 14:43 |
-!- vortex_ape [~vortex_ap@59.177.142.50] has quit [Ping timeout: 252 seconds] | 14:46 | |
-!- vortex_ape [~vortex_ap@120.59.75.212] has joined #shogun | 14:46 | |
shogun-buildbot | build #261 of trusty - libshogun - viennacl is complete: Failure [failed test] Build details are at http://buildbot.shogun-toolbox.org/builders/trusty%20-%20libshogun%20-%20viennacl/builds/261 blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com> | 14:47 |
shogun-notifier- | shogun: Khaled Nasr :develop * 88269cd / examples/undocumented/libshogun/ (4 files): https://github.com/shogun-toolbox/shogun/commit/88269cdc73df3a82f2fb9fafbc8462f602d69a2a | 14:50 |
shogun-notifier- | shogun: fixed uncought exception problem in neuralnets c++ examples | 14:50 |
shogun-notifier- | shogun: Sergey Lisitsyn :develop * 529ab43 / examples/undocumented/libshogun/ (4 files): https://github.com/shogun-toolbox/shogun/commit/529ab4383cac9ee6ff90a59de8c098e12e64fbe8 | 14:50 |
shogun-notifier- | shogun: Merge pull request #2535 from khalednasr/bugfix | 14:50 |
shogun-notifier- | shogun: | 14:50 |
shogun-notifier- | shogun: fixed uncought exception problem in neuralnets c++ examples | 14:50 |
shogun-buildbot | build #3376 of deb1 - libshogun is complete: Success [build successful] Build details are at http://buildbot.shogun-toolbox.org/builders/deb1%20-%20libshogun/builds/3376 | 14:50 |
shogun-buildbot | build #262 of trusty - libshogun - viennacl is complete: Failure [failed test] Build details are at http://buildbot.shogun-toolbox.org/builders/trusty%20-%20libshogun%20-%20viennacl/builds/262 blamelist: Khaled Nasr <k.nasr92@gmail.com>, Sergey Lisitsyn <lisitsyn.s.o@gmail.com> | 14:53 |
-!- travis-ci [~travis-ci@ec2-54-166-62-75.compute-1.amazonaws.com] has joined #shogun | 14:55 | |
travis-ci | it's Sergey Lisitsyn's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: http://travis-ci.org/shogun-toolbox/shogun/builds/56295522 | 14:55 |
-!- travis-ci [~travis-ci@ec2-54-166-62-75.compute-1.amazonaws.com] has left #shogun [] | 14:55 | |
-!- HeikoS [~heiko@0543fcaf.skybroadband.com] has quit [Quit: Leaving.] | 15:02 | |
-!- HeikoS [~heiko@0543fcaf.skybroadband.com] has joined #shogun | 15:08 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 15:08 | |
-!- HeikoS [~heiko@0543fcaf.skybroadband.com] has quit [Quit: Leaving.] | 15:14 | |
-!- travis-ci [~travis-ci@ec2-54-235-29-7.compute-1.amazonaws.com] has joined #shogun | 15:17 | |
travis-ci | it's Sergey Lisitsyn's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: http://travis-ci.org/shogun-toolbox/shogun/builds/56295990 | 15:17 |
-!- travis-ci [~travis-ci@ec2-54-235-29-7.compute-1.amazonaws.com] has left #shogun [] | 15:17 | |
-!- curiousguy13 [~curiousgu@120.57.232.169] has quit [Ping timeout: 255 seconds] | 16:37 | |
-!- curiousguy13 [~curiousgu@59.177.201.227] has joined #shogun | 16:49 | |
nileshK | Hello anybody there | 16:59 |
nileshK | wiking:? | 17:00 |
nileshK | lisitsyn:? | 17:00 |
@lisitsyn | nileshK: yes | 17:00 |
nileshK | I was a bit doubtful regarding the implementation of LpBoost that Shogun has | 17:00 |
nileshK | Does it really work? | 17:00 |
nileshK | And has it been tested? | 17:00 |
@lisitsyn | nileshK: ohh its pretty old | 17:01 |
@lisitsyn | I think it was tested somehow | 17:01 |
@lisitsyn | nileshK: sonney2k_ is the man to ask actually | 17:02 |
nileshK | I need some help in getting an example of it working. | 17:02 |
nileshK | Can you help me right now? | 17:02 |
nileshK | He doesnt seem to be be online now. | 17:02 |
@lisitsyn | I can try | 17:03 |
nileshK | So, see LpBoost is inherited from a class CLinearMachine | 17:03 |
@lisitsyn | yes | 17:03 |
nileshK | But it doesnot have a constructed implemented which can accept the data | 17:04 |
@lisitsyn | what data? | 17:04 |
nileshK | as in for that matter even SVMOcas is also inherited from CLinearMachine | 17:04 |
@lisitsyn | yes they are all linear | 17:04 |
nileshK | data => training set | 17:05 |
nileshK | CSVMOcas* lin_svm = new CSVMOcas(C, rf_feats, labels); | 17:05 |
@lisitsyn | ah | 17:05 |
@lisitsyn | that's ok | 17:06 |
@lisitsyn | just do set_features | 17:06 |
nileshK | Also regarding doing the set_features thing | 17:06 |
nileshK | is there a specfic format in which the training data has to be presented | 17:07 |
nileshK | Just to clarify my point what kind of data files can I use with LpBoost | 17:07 |
nileshK | from the shogun-data that I downloaded | 17:08 |
@lisitsyn | as far as I can see LpBoost requires sparse features | 17:08 |
nileshK | Yeah. so how do I proceed | 17:08 |
@lisitsyn | what language are you using shogun in? | 17:09 |
nileshK | I am using C++ | 17:09 |
@lisitsyn | ok I see | 17:10 |
@lisitsyn | you need to construct CSparseFeatures https://github.com/shogun-toolbox/shogun/blob/develop/src/shogun/features/SparseFeatures.h | 17:10 |
@lisitsyn | you can create sgsparsematrix or sgmatrix | 17:11 |
@lisitsyn | or load it from file | 17:11 |
nileshK | ok. Assuming I created the CSparseFeatures | 17:12 |
nileshK | The next step would be to the train function? | 17:12 |
@lisitsyn | yeah you just create CLpBoost thing | 17:12 |
@lisitsyn | call set_features(<your features>) | 17:12 |
@lisitsyn | call set_labels(<your labels>) | 17:12 |
@lisitsyn | and train | 17:13 |
nileshK | ok. | 17:14 |
nileshK | I will try this and get back to you | 17:14 |
nileshK | Till what time will you be online? | 17:14 |
@lisitsyn | nileshK: whole day but can disappear for some moments :) | 17:14 |
nileshK | Ok. I might require your help for small issues I might stuck into. :) | 17:15 |
nileshK | Thanks a lot! | 17:15 |
nileshK | I will ping you after trying this | 17:15 |
@lisitsyn | nileshK: sure | 17:17 |
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout] | 17:50 | |
-!- besser82 [~besser82@fedora/besser82] has quit [Ping timeout: 265 seconds] | 18:37 | |
nileshK | lisitsyn: Hi | 19:23 |
@lisitsyn | nileshK: hey | 19:24 |
nileshK | I was checking this file out: https://github.com/shogun-toolbox/shogun/blob/develop/applications/classification/random_fourier_classification.cpp | 19:24 |
nileshK | It had an example of reading sparefeatures | 19:24 |
@lisitsyn | ah yeah | 19:24 |
@lisitsyn | it uses libsvm | 19:25 |
nileshK | but can you point me to a dataset that could be read via that function | 19:25 |
nileshK | Please check the load_data function | 19:25 |
@lisitsyn | just a second | 19:27 |
@lisitsyn | nileshK: http://www.csie.ntu.edu.tw/~cjlin/libsvmtools/datasets/ | 19:30 |
@lisitsyn | these should work | 19:30 |
nileshK | Ok. I will try and get back. Thanks :) | 19:31 |
nileshK | Lisitsyn: The dataset worked correctly. | 20:24 |
nileshK | I was trying to modify the code to generate an example for lpBoost on lines with this code | 20:24 |
nileshK | https://www.irccloud.com/pastebin/Jw49xhzS | 20:24 |
nileshK | This code I have adapted from teh random_fourier_classfication | 20:25 |
nileshK | It is compiling well for me | 20:25 |
nileshK | g++ lpBoost_classfication.cpp -lshogun -L/home/nilesh/shogun-install/lib -I/home/nilesh/shogun-install/include | 20:25 |
nileshK | I am using this data set | 20:26 |
nileshK | http://www.csie.ntu.edu.tw/~cjlin/libsvmtools/datasets/binary.html#a1a | 20:26 |
nileshK | a1a | 20:26 |
nileshK | But on executing with this command : | 20:26 |
nileshK | ./a.out --dataset ../../data/a1a | 20:26 |
nileshK | https://www.irccloud.com/pastebin/FvO1kxdk | 20:26 |
@lisitsyn | export LD_LIBRARY_PATH=/home/nilesh/shogun-install/lib | 20:27 |
@lisitsyn | then try that again | 20:27 |
nileshK | Bingo! | 20:29 |
nileshK | Worked. | 20:29 |
@lisitsyn | nileshK: cool :) hope lpboost works as well | 20:31 |
nileshK | Yeah I am now adding the testing part to it. | 20:31 |
nileshK | Let me see how that goes through | 20:31 |
-!- pickle27 [~pickle27@192-0-134-235.cpe.teksavvy.com] has joined #shogun | 20:43 | |
-!- vortex_ape [~vortex_ap@120.59.75.212] has quit [Ping timeout: 256 seconds] | 20:46 | |
-!- vortex_ape [~vortex_ap@120.56.206.135] has joined #shogun | 20:47 | |
-!- HeikoS [~heiko@0543fcaf.skybroadband.com] has joined #shogun | 21:04 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 21:04 | |
@HeikoS | lisitsyn: around? | 21:07 |
@lisitsyn | HeikoS: yes | 21:08 |
@HeikoS | lisitsyn: hi how are things? | 21:08 |
@HeikoS | feeling better? | 21:08 |
@lisitsyn | HeikoS: yeah much better :) | 21:08 |
@lisitsyn | thanks | 21:08 |
@lisitsyn | what about you? | 21:08 |
@HeikoS | yeah same | 21:09 |
@HeikoS | parents are visiting in London | 21:09 |
@HeikoS | thats fun | 21:09 |
@lisitsyn | cool | 21:09 |
@HeikoS | hey I wanted to ask aobut the manual stuff | 21:09 |
@HeikoS | want to keep that going | 21:09 |
@HeikoS | whats needed? | 21:09 |
@lisitsyn | well you know I didn't approach that since then | 21:09 |
@lisitsyn | :) | 21:09 |
@HeikoS | I didnt know | 21:10 |
@HeikoS | ? | 21:10 |
@lisitsyn | hmm well | 21:10 |
@HeikoS | shall we try to tackle the html thingi now? | 21:10 |
@lisitsyn | I am not sure I can do that now but I can try tomorrow | 21:10 |
@lisitsyn | the thing is that we would need our own 'visit' thing | 21:11 |
@lisitsyn | that prints node | 21:11 |
@lisitsyn | because we mess with bootstrap and classes are not enough | 21:11 |
@HeikoS | shouldn not be too hard should it? | 21:11 |
@lisitsyn | should be easy | 21:11 |
@lisitsyn | HeikoS: if we go with better naming of functions it would look really amazing | 21:12 |
@lisitsyn | people will love it | 21:12 |
@lisitsyn | get_something() is non-sense in java | 21:12 |
@HeikoS | lisitsyn: yeah I like tha tone | 21:13 |
@HeikoS | lisitsyn: I think we should combine it with the manual thing | 21:13 |
@HeikoS | integrate it straight away to get these things out together | 21:13 |
@HeikoS | one *big* downside though is that the API docs dont work anymore | 21:13 |
@HeikoS | currently, syntax is the same | 21:14 |
@HeikoS | and that is good | 21:14 |
@HeikoS | so not sure about details there | 21:14 |
@lisitsyn | should be solvable | 21:14 |
@lisitsyn | I need to force myself to work on doc/distribution of shogun | 21:14 |
@lisitsyn | instead of neural nets :D | 21:14 |
@HeikoS | hehe | 21:14 |
@lisitsyn | something made me to refactor some code there | 21:15 |
@HeikoS | lisitsyn: I mean as far as I know we cannot generate API for different languages | 21:15 |
@HeikoS | but if we had the examples, then this is fine as it shows how to use the API | 21:15 |
@HeikoS | lisitsyn: do you ahve an idea how to define this visit thing? | 21:15 |
@lisitsyn | HeikoS: I have seen some example | 21:15 |
@lisitsyn | ah | 21:16 |
@lisitsyn | actually in http://sphinx-doc.org/extdev/tutorial.html they define visit_* thing | 21:16 |
@lisitsyn | but I've seen some other example | 21:16 |
@lisitsyn | how to add custom stuff there | 21:16 |
@lisitsyn | HeikoS: http://stackoverflow.com/questions/13168702/how-to-add-rel-attribute-to-docutils-sphinx-reference-in-html-output that's it | 21:19 |
@HeikoS | ok let me check | 21:21 |
@HeikoS | lisitsyn: so its just self.body.append | 21:21 |
@lisitsyn | looks like | 21:22 |
@lisitsyn | lets try | 21:22 |
@HeikoS | Ill just pase his code in ours and see what happens | 21:22 |
@HeikoS | mmh | 21:25 |
@lisitsyn | HeikoS: I just tried to self.body.append | 21:25 |
@HeikoS | lisitsyn: doesnt seem like its doing something | 21:25 |
@lisitsyn | yeap | 21:25 |
@HeikoS | lisitsyn: but he does someting as self.body.append and then self.context.append | 21:26 |
@lisitsyn | hah stop I was doing some useless stuff | 21:27 |
@HeikoS | lisitsyn: ? | 21:28 |
@HeikoS | lisitsyn: the code is actually not executed | 21:28 |
@HeikoS | I mean the visit code | 21:28 |
@lisitsyn | yeah I tried to modfiy visit_sgexample | 21:28 |
@lisitsyn | which is not used anyway | 21:29 |
@HeikoS | lisitsyn: mmh | 21:29 |
@HeikoS | why is that? | 21:29 |
@HeikoS | was the orginal example different? | 21:30 |
@lisitsyn | HeikoS: we use ShogunExample | 21:30 |
@lisitsyn | but not sgexample | 21:30 |
@HeikoS | lisitsyn: its this phase 4 they talk about | 21:30 |
@HeikoS | doesnt seem to change anything? | 21:31 |
@HeikoS | lisitsyn: http://sphinx-doc.org/extdev/tutorial.html | 21:32 |
@lisitsyn | oka | 21:32 |
@HeikoS | check description of phase 4 | 21:32 |
@lisitsyn | I managed to add a tag | 21:32 |
@HeikoS | This phase converts the resolved doctrees to the desired output format, such as HTML or LaTeX. This happens via a so-called docutils writer that visits the individual nodes of each doctree and produces some output in the process. | 21:32 |
@lisitsyn | we are close | 21:32 |
@lisitsyn | HeikoS: ok I can commit now | 21:35 |
@HeikoS | ok go | 21:35 |
@lisitsyn | but it doesn't work | 21:35 |
@lisitsyn | :D | 21:35 |
@HeikoS | ? | 21:35 |
@HeikoS | what did you do then? | 21:35 |
@lisitsyn | HeikoS: I added divs with proper role | 21:36 |
@lisitsyn | around the code | 21:36 |
@HeikoS | cool | 21:36 |
@HeikoS | why doesnt it work? | 21:36 |
@lisitsyn | some bootstrap stuff | 21:36 |
@lisitsyn | it breaks due to some strange layout I guess | 21:36 |
@lisitsyn | or maybe no ids | 21:36 |
@lisitsyn | let me check | 21:37 |
@HeikoS | kk | 21:38 |
@lisitsyn | HeikoS: we have multiple tab panel | 21:38 |
@HeikoS | ill wait for the push | 21:38 |
@lisitsyn | this is tricky | 21:38 |
@HeikoS | really? | 21:38 |
@HeikoS | should not be | 21:38 |
@HeikoS | because we just iterate? | 21:38 |
@HeikoS | and its all tree structure? | 21:38 |
@lisitsyn | HeikoS: yeah our template makes it very tricky | 21:39 |
@lisitsyn | HeikoS: I mean | 21:40 |
@lisitsyn | we have tabpanel that serves python/java/.. | 21:40 |
@lisitsyn | and it has tab-content that's basically whole page | 21:40 |
@lisitsyn | and inside of that we have another tab-content | 21:40 |
@lisitsyn | this gets really tricky | 21:41 |
@lisitsyn | okay anyway that's html problem now | 21:41 |
@lisitsyn | I'll commit now and continue later | 21:42 |
@lisitsyn | pushed | 21:42 |
@HeikoS | cool Ill check | 21:46 |
@HeikoS | nono | 21:47 |
@HeikoS | lisitsyn: wait sergey | 21:47 |
@HeikoS | its different | 21:47 |
@HeikoS | its not a whole page | 21:47 |
@HeikoS | we changed that | 21:47 |
@HeikoS | the template still has the thing globally | 21:47 |
@HeikoS | but only the code blocks are duplicated ,the page itself is not | 21:47 |
@lisitsyn | HeikoS: yes and it won't work globally | 21:47 |
@lisitsyn | I don't think it can work with nested tab-content | 21:48 |
@HeikoS | but do we need that even? | 21:48 |
@HeikoS | ah I see | 21:48 |
@HeikoS | mmmh | 21:48 |
@lisitsyn | there is only one section for tabs | 21:48 |
@lisitsyn | but multiple tab sets | 21:49 |
@lisitsyn | like for every snippet | 21:49 |
@lisitsyn | but it's other problem | 21:49 |
@lisitsyn | I think I can fix that | 21:49 |
@lisitsyn | needs some javascript | 21:49 |
@HeikoS | really? | 21:50 |
@HeikoS | dont know anything about this | 21:50 |
@HeikoS | how doe sthat work? | 21:50 |
@lisitsyn | HeikoS: I think we need to add callbacks for these 'global' tabs | 21:53 |
@lisitsyn | and invisible 'local' tabs | 21:54 |
@lisitsyn | invisible tab headers I mean | 21:54 |
@lisitsyn | so when one selects some global tab | 21:54 |
@lisitsyn | all corresponding local tabs are selected | 21:54 |
@lisitsyn | do you follow the idea? | 21:55 |
nileshK | Lisitsyn: I am still having one more issue. | 21:57 |
@lisitsyn | nileshK: what's it? | 21:57 |
nileshK | https://www.irccloud.com/pastebin/odLBYcdO | 21:57 |
nileshK | This is the code we have for LpBoost example | 21:57 |
nileshK | I am not able to collect the output after training | 21:58 |
nileshK | the object with the data. | 21:58 |
@lisitsyn | I don't get it | 21:58 |
@lisitsyn | what output? | 21:58 |
nileshK | https://github.com/shogun-toolbox/shogun/blob/develop/applications/classification/random_fourier_classification.cpp | 21:59 |
nileshK | Like in this example if you check line no: | 21:59 |
nileshK | 50 | 21:59 |
nileshK | 150* | 21:59 |
nileshK | You see they are getting the weights w after training | 21:59 |
@lisitsyn | you want the learned weight vector? | 22:00 |
nileshK | I dont see how to collect the training output | 22:00 |
nileshK | Yea | 22:00 |
@lisitsyn | get_w | 22:00 |
nileshK | But then if you move to line : 159 | 22:00 |
nileshK | will just giving that w be good enough | 22:00 |
@lisitsyn | that's random features specific | 22:01 |
nileshK | I am trying to find an analogy b/w this code and lpboost example and create a similar one | 22:01 |
@lisitsyn | you don't need to do anything with weights | 22:01 |
@lisitsyn | once learned you can apply lpboost | 22:01 |
@lisitsyn | it will use learned weights | 22:01 |
nileshK | then how do you check the output for the test samples | 22:02 |
@lisitsyn | nileshK: line 176 of your example | 22:02 |
@lisitsyn | it returns binary labels with predicted classes | 22:02 |
nileshK | that code is not workgin | 22:03 |
@lisitsyn | what error? | 22:03 |
@HeikoS | lisitsyn: actually, not sure | 22:04 |
@HeikoS | but I dont know much javascript | 22:04 |
@HeikoS | can you gist an example? | 22:04 |
nileshK | https://www.irccloud.com/pastebin/lilEQF03 | 22:06 |
nileshK | This is the error which comes while training it | 22:06 |
nileshK | This is the gdb output | 22:10 |
nileshK | https://www.irccloud.com/pastebin/eeWh0MqH | 22:10 |
-!- pickle27 [~pickle27@192-0-134-235.cpe.teksavvy.com] has quit [Remote host closed the connection] | 22:30 | |
@lisitsyn | HeikoS: yeah a bit later | 22:33 |
@lisitsyn | nileshK: uh then it looks broken | 22:33 |
@lisitsyn | may be some cplex version incompatibility.. | 22:34 |
nileshK | How strong are you about that? | 22:34 |
@lisitsyn | no clue :) | 22:35 |
@HeikoS | lisitsyn: cool thanks! | 22:36 |
@HeikoS | lisitsyn: its a pitty we have no tests for thouse cplex things | 22:36 |
@lisitsyn | yeah | 22:37 |
@HeikoS | lisitsyn: sonney2k_ suggested to drop them as not open source | 22:37 |
@HeikoS | any other way to implement such linear programing things? | 22:37 |
@lisitsyn | well it is written before jesus christ | 22:37 |
-!- curiousguy13 [~curiousgu@59.177.201.227] has quit [Ping timeout: 256 seconds] | 22:37 | |
@lisitsyn | in 1953 or so | 22:37 |
@lisitsyn | :D | 22:37 |
@lisitsyn | don't know | 22:37 |
@lisitsyn | I guess there is no unified cool open source solution for LP | 22:38 |
nileshK | HeikoS: what do you mean by other way? | 22:38 |
@lisitsyn | w/o cplex | 22:38 |
nileshK | Cplex is good right? | 22:38 |
nileshK | There are other libraries lpsolve | 22:38 |
@lisitsyn | cplex is good but not open source yeah | 22:38 |
nileshK | Lpsolve is good option in that case | 22:39 |
@HeikoS | nileshK: do you know other lp solvers? | 22:39 |
@HeikoS | nileshK, lisitsyn could be a cool GSoC project | 22:39 |
nileshK | I think yes. Because the amount of work that is currently required to get these things working | 22:40 |
nileshK | is quite enough | 22:40 |
nileshK | HeikoS just confirm one thing, do know whether lpBoost is tested? | 22:41 |
@HeikoS | nileshK: no idea | 22:41 |
@HeikoS | this is all very old stuff before all of us were around | 22:42 |
@HeikoS | so you will have to try, sorry | 22:42 |
@HeikoS | but any findings are absolutely welcome | 22:42 |
@HeikoS | so document them on github issues | 22:42 |
@HeikoS | this increases probability of getting them fixed and working again | 22:42 |
nileshK | Yeah. As soon as i get everything working right and I am sure. I will write some examples and unit tests | 22:43 |
nileshK | for these. Also is there any time when Soren is available? | 22:43 |
nileshK | lisitsyn: Are you going to be around? | 22:45 |
-!- curiousguy13 [~curiousgu@120.59.108.178] has joined #shogun | 22:49 | |
@HeikoS | nileshK: s?ren is not available atm | 23:09 |
@HeikoS | you can try to email him, but its unlikely that he will reply, he is too busy currently | 23:10 |
-!- pickle27 [~pickle27@192-0-134-235.cpe.teksavvy.com] has joined #shogun | 23:10 | |
@HeikoS | nileshK: is always around | 23:10 |
@HeikoS | pickle27: hi! | 23:10 |
@HeikoS | pickle27: did you see our take on sphinx and the manual? | 23:10 |
@HeikoS | pickle27: we are almost there, just some things missing, pull from the repo and run make preview in the sphinx dir | 23:10 |
@HeikoS | if you have an idea on the html tags, that would be useful | 23:10 |
pickle27 | I did see, looks like its getting close | 23:11 |
pickle27 | I've never used sphinx so I cant really add much there | 23:11 |
@HeikoS | python docutils is the thing we need to understand | 23:14 |
@HeikoS | going to bed now. see you! | 23:14 |
nileshK | Gn HeikoS | 23:14 |
pickle27 | okay! I'll try and see if I have any time to look at it - got deadlines at work atm | 23:15 |
@HeikoS | nileshK: night! | 23:15 |
@HeikoS | pickle27: same ;) | 23:15 |
@HeikoS | bye | 23:15 |
-!- HeikoS [~heiko@0543fcaf.skybroadband.com] has quit [Quit: Leaving.] | 23:15 | |
--- Log closed Mon Mar 30 00:00:21 2015 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!