--- Log opened Sat Jun 25 00:00:54 2016 | ||
-!- bazdmeg is now known as BZDMG | 01:47 | |
-!- sonne|osx [~sonne@x4db339ab.dyn.telefonica.de] has joined #shogun | 05:13 | |
-!- sonne|osx_ [~sonne@x4e345856.dyn.telefonica.de] has joined #shogun | 05:18 | |
-!- sonne|osx [~sonne@x4db339ab.dyn.telefonica.de] has quit [Ping timeout: 240 seconds] | 05:20 | |
-!- sonne|osx_ is now known as sonne|osx | 05:20 | |
-!- sonne|osx [~sonne@x4e345856.dyn.telefonica.de] has quit [Quit: sonne|osx] | 05:28 | |
-!- Saurabh7__ [Saurabh7@gateway/shell/panicbnc/x-wlacyikiyzqrkfcs] has left #shogun [] | 07:02 | |
-!- sonne|osx [~sonne@x4e345856.dyn.telefonica.de] has joined #shogun | 07:03 | |
-!- sonne|osx [~sonne@x4e345856.dyn.telefonica.de] has quit [Quit: sonne|osx] | 07:14 | |
-!- sanuj [~sanuj@117.220.49.228] has joined #shogun | 07:41 | |
sanuj | HeikoS_web, there? | 07:41 |
---|---|---|
sanuj | wiking, there? | 07:42 |
-!- HeikoS [~heiko@AAubervilliers-652-1-253-126.w83-112.abo.wanadoo.fr] has joined #shogun | 07:55 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 07:55 | |
sanuj | HeikoS, yo | 07:55 |
@HeikoS | sanuj: jo, sorry gotta come back later | 07:56 |
-!- HeikoS [~heiko@AAubervilliers-652-1-253-126.w83-112.abo.wanadoo.fr] has quit [Client Quit] | 07:56 | |
-!- sanuj [~sanuj@117.220.49.228] has quit [Ping timeout: 244 seconds] | 08:01 | |
-!- sanuj [~sanuj@117.220.49.228] has joined #shogun | 08:06 | |
-!- Saurabh7__ [Saurabh7@gateway/shell/panicbnc/x-wlacyikiyzqrkfcs] has joined #shogun | 09:40 | |
-!- GandalfTheWizard [~Eva@112.10.171.169] has joined #shogun | 10:20 | |
-!- HeikoS [~heiko@AAubervilliers-652-1-253-126.w83-112.abo.wanadoo.fr] has joined #shogun | 10:23 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 10:23 | |
@HeikoS | sanuj: jo | 10:24 |
@HeikoS | now I have some itme | 10:24 |
sanuj | okay | 10:24 |
sanuj | i have a problem | 10:24 |
sanuj | wait i'll push the code | 10:25 |
@HeikoS | ok ! | 10:25 |
@HeikoS | sanuj: btw did you see my email for midterm update? | 10:27 |
@HeikoS | sanuj: can you cc me in the daily update email? | 10:27 |
@HeikoS | Saurabh7__: jo | 10:27 |
Saurabh7__ | HeikoS: yo | 10:29 |
sanuj | HeikoS, the error is caused in this region | 10:30 |
sanuj | https://github.com/shogun-toolbox/shogun/pull/3221/files#diff-4a319646835bd87fa05b41d8b1872184R22 | 10:30 |
sanuj | HeikoS, here is the error http://pastebin.com/eFrBNL51 | 10:30 |
@HeikoS | sanuj: so what shappening? | 10:31 |
sanuj | HeikoS, yeah, i'll start sending mails from monday.... | 10:31 |
@HeikoS | sanuj: cool thanks for that! | 10:32 |
sanuj | HeikoS, did you see the error? | 10:32 |
@HeikoS | sanuj: I see it, but can you point me the the things important? | 10:33 |
sanuj | i think any destructor deletes member variables | 10:33 |
sanuj | any.h destructor | 10:33 |
sanuj | HeikoS, https://github.com/shogun-toolbox/shogun/pull/3221/files#diff-8ea96286d95b52029d31636117e0fe55R149 | 10:33 |
@HeikoS | whats a minimal example? | 10:34 |
@HeikoS | is there one in the PR? | 10:34 |
@HeikoS | sanuj: https://github.com/shogun-toolbox/shogun/pull/3221/files#diff-8ea96286d95b52029d31636117e0fe55R98 | 10:34 |
@HeikoS | this line deletes memory | 10:34 |
@HeikoS | right? | 10:34 |
@HeikoS | BZDMG: jo | 10:35 |
sanuj | yes | 10:35 |
sanuj | BZDMG is away i think | 10:35 |
sanuj | HeikoS, why is the destructor being called | 10:36 |
@HeikoS | sanuj: can you put the test that segfaults into its own main | 10:36 |
@HeikoS | and then run again, share the output | 10:36 |
@HeikoS | of valgrind | 10:36 |
@HeikoS | the current one is convoluted so much with gtest | 10:36 |
sanuj | okay, let me do that | 10:36 |
@HeikoS | cool | 10:36 |
@HeikoS | Saurabh7__: how are things? | 10:37 |
Saurabh7__ | HeikoS: I have kept the duplicate indices, | 10:37 |
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun | 10:37 | |
shogun-notifier- | shogun-data: Sanuj :master * 39551ae / testsuite/meta/regression/multiple_kernel_learning_regression.dat: https://github.com/shogun-toolbox/shogun-data/commit/39551ae4620ec0baffff17e9dae5b8ff0f8ea101 | 10:37 |
shogun-notifier- | shogun-data: add data for mkl regression cookbook | 10:37 |
shogun-notifier- | shogun-data: Heiko Strathmann :master * 5e64d2a / testsuite/meta/regression/multiple_kernel_learning_regression.dat: https://github.com/shogun-toolbox/shogun-data/commit/5e64d2aa2456432f3c67604fcbf8b76d0c6e4d79 | 10:37 |
shogun-notifier- | shogun-data: Merge pull request #107 from sanuj/mkl_regression | 10:37 |
shogun-notifier- | shogun-data: | 10:37 |
shogun-notifier- | shogun-data: Mkl regression | 10:37 |
Saurabh7__ | I believe they are justified | 10:37 |
Saurabh7__ | I checked | 10:37 |
@HeikoS | Saurabh7__: ok | 10:38 |
@HeikoS | can you share someevidence for that with me? :) | 10:38 |
sanuj | HeikoS, what is HeikoS_web | 10:38 |
@HeikoS | sanuj: thats my work computer logged in via webinterface | 10:38 |
@HeikoS | but it is in london | 10:38 |
@HeikoS | and I am in France | 10:38 |
@HeikoS | escaping the brexit | 10:38 |
Saurabh7__ | HeikoS: HeikoS for e.g https://en.wikipedia.org/wiki/Bootstrapping_(statistics)#Approach | 10:39 |
Saurabh7__ | so sampling with replacement | 10:39 |
Saurabh7__ | am i right? | 10:39 |
@HeikoS | ok | 10:40 |
@HeikoS | Saurabh7__: can you also check sklearn impl? | 10:40 |
Saurabh7__ | HeikoS: I checked | 10:41 |
Saurabh7__ | its very different in terms of approach, has been changed with time | 10:41 |
Saurabh7__ | dups are allowed | 10:41 |
Saurabh7__ | but they handle it with sample weights | 10:41 |
Saurabh7__ | like bincount(samples) | 10:42 |
Saurabh7__ | and use it somehow | 10:42 |
Saurabh7__ | so the idea is same tho | 10:42 |
@HeikoS | okok | 10:43 |
@HeikoS | so lets continue with things then | 10:43 |
Saurabh7__ | HeikoS: antoher thing, presortin I have kept optional still | 10:44 |
Saurabh7__ | because in sinlge tree we train on subset | 10:44 |
Saurabh7__ | so cant presort inside it | 10:44 |
Saurabh7__ | so I do it in RF | 10:44 |
@HeikoS | ok | 10:45 |
@HeikoS | thats ok then | 10:45 |
Saurabh7__ | cool | 10:45 |
Saurabh7__ | HeikoS: ok I will add some doc in t current PR and some tests are needed | 10:46 |
@HeikoS | yep | 10:47 |
sanuj | HeikoS, i'm not able to compile this | 10:52 |
sanuj | temp.cpp:(.text+0x5): undefined reference to `shogun::init_shogun_with_defaults()' | 10:52 |
sanuj | error ^ | 10:52 |
@HeikoS | dude | 10:52 |
@HeikoS | :) | 10:52 |
@HeikoS | so lets see | 10:53 |
@HeikoS | you know what an undefined reference means? | 10:53 |
@HeikoS | I mean the error, what causes it=? | 10:53 |
sanuj | no | 10:53 |
sanuj | not able to link it | 10:54 |
@HeikoS | sanuj: so that happens | 10:54 |
@HeikoS | if you can compile a program | 10:54 |
@HeikoS | that is, turn the program.c into program.o | 10:54 |
@HeikoS | so all header files were available | 10:55 |
@HeikoS | in the linking step, all the .o files are put together into the binary | 10:55 |
@HeikoS | now, | 10:55 |
@HeikoS | if there is a method that does not appear in one of the .o files linked together (that is, no implementation of the defined function) | 10:56 |
@HeikoS | then you get linker error | 10:56 |
@HeikoS | if the method is in a third party lib, like your init_shogun_with_defaults | 10:56 |
@HeikoS | then, this means you linker cannot see any library that contains an implementation of the method | 10:56 |
@HeikoS | sanuj: I suggest you read up on how c/c++ compiler work | 10:56 |
@HeikoS | in practice, it means your forgot -lshogun | 10:56 |
sanuj | HeikoS, i compiled with => g++ --std=c++11 -I/home/sanuj/Projects/shogun/src/ -L/home/sanuj/Projects/shogun/buildpy/src/shogun -lshogun temp.cpp | 10:57 |
@HeikoS | this tells the linker where to look for .so files | 10:57 |
@HeikoS | if you put the -lshogun but still get the error | 10:57 |
@HeikoS | this means that you LD_LIBRARY_PATH does not contain the file | 10:57 |
@HeikoS | so either your -L option is wrong | 10:57 |
@HeikoS | or there is no file there | 10:58 |
@HeikoS | it is one of the most basic errors when compiling c++ | 10:58 |
sanuj | my LD_LIBRARY_PATH is /home/sanuj/Projects/shogun/buildpy/src/shogun:/home/sanuj/torch/install/lib: | 10:58 |
@HeikoS | so no reason to ask more than once ;) | 10:58 |
sanuj | and the path in -L is also correct | 10:58 |
@HeikoS | sanuj: nope | 10:58 |
@HeikoS | because if it was, it would wortk | 10:58 |
@HeikoS | does your -L path contain modshogun.so? | 10:59 |
sanuj | that path has libshogun.so | 10:59 |
@HeikoS | ah yeah | 10:59 |
@HeikoS | well, I dont know, there are 100 mini things that can be set wrongly | 11:00 |
@HeikoS | but the rationale is usper simple | 11:00 |
@HeikoS | linker cannot find init_shogun_with_defaults | 11:00 |
@HeikoS | thats it | 11:00 |
@HeikoS | you have to tell it where it is | 11:00 |
@HeikoS | and it will work | 11:00 |
sanuj | HeikoS, this worked | 11:00 |
sanuj | g++ --std=c++11 -I/home/sanuj/Projects/shogun/src/ -L/home/sanuj/Projects/shogun/buildpy/src/shogun temp.cpp -lshogun | 11:00 |
sanuj | swapping temp.cpp and -lshogun | 11:00 |
@HeikoS | sanuj: ah yeah that also is a thing to consider | 11:01 |
@HeikoS | sanuj: try to read up on why | 11:01 |
@HeikoS | so next time someone asks here, you can explain them | 11:01 |
sanuj | HeikoS, okay thanks | 11:01 |
@HeikoS | ;) | 11:01 |
sanuj | :) | 11:01 |
sanuj | btw i know c++ linker and compiler basics | 11:01 |
sanuj | everything was correct but it was not working that's why i asked here | 11:02 |
@HeikoS | sanuj: sure | 11:02 |
@HeikoS | sanuj: I only explained | 11:02 |
@HeikoS | since the question | 11:03 |
@HeikoS | to make sure you are on track with this | 11:03 |
@HeikoS | since there is only one answer to the question: the linker doesnt see libshogun.so | 11:03 |
@HeikoS | but all good | 11:04 |
@HeikoS | the mkl cookbooks look nice btw | 11:04 |
shogun-notifier- | shogun: OXPHOS :develop * 9a8de01 / doc/ (3 files): https://github.com/shogun-toolbox/shogun/commit/9a8de01ba3744728b4a1612ab0de39375e0a299b | 11:04 |
shogun-notifier- | shogun: - update ecoc intro - remove python modular example | 11:04 |
shogun-notifier- | shogun: Heiko Strathmann :develop * a39216f / doc/ (3 files): https://github.com/shogun-toolbox/shogun/commit/a39216f1a0825a2f28936b6b5ad865dce74b9066 | 11:04 |
shogun-notifier- | shogun: Merge pull request #3296 from OXPHOS/cookbook_ecoc | 11:04 |
shogun-notifier- | shogun: | 11:04 |
shogun-notifier- | shogun: ECOC cookbook update | 11:04 |
sanuj | HeikoS, thanks | 11:05 |
shogun-buildbot | build #738 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/738 blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>, OXPHOS <engelzora@gmail.com> | 11:05 |
sanuj | HeikoS, why did you get out of UK because of brexit? | 11:05 |
@HeikoS | haha | 11:06 |
@HeikoS | sanuj: I was afraid that idiotism is infectious | 11:06 |
sanuj | HeikoS, haha, i meant is that the only reason you are in france? | 11:07 |
sanuj | HeikoS, updated error from main http://pastebin.com/fjAZs2pJ | 11:07 |
@HeikoS | sanuj: I also wanted to use the channel tunnel for the last time before it is blown up | 11:07 |
sanuj | they will blow up the channel tunnel? | 11:08 |
shogun-notifier- | shogun: OXPHOS :develop * a4e49b3 / / (4 files): https://github.com/shogun-toolbox/shogun/commit/a4e49b3d2260754cc804a04151cf467efee2bd36 | 11:08 |
shogun-notifier- | shogun: relaxed_tree cookbook | 11:08 |
shogun-notifier- | shogun: Heiko Strathmann :develop * e001ab5 / / (4 files): https://github.com/shogun-toolbox/shogun/commit/e001ab52b0f2d128db8a31cac633ea281691bb75 | 11:08 |
shogun-notifier- | shogun: Merge pull request #3286 from OXPHOS/cookbook_relaxedtree | 11:08 |
shogun-notifier- | shogun: | 11:08 |
shogun-notifier- | shogun: Cookbook - relaxedtree classifier | 11:08 |
@HeikoS | sanuj: :D | 11:08 |
@HeikoS | just in case | 11:08 |
sanuj | :P | 11:08 |
@HeikoS | checking the error | 11:08 |
sanuj | cool | 11:08 |
@HeikoS | can you put this as a gist, and also put the example code | 11:08 |
sanuj | okay | 11:09 |
shogun-buildbot | build #739 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/739 blamelist: Heiko Strathmann <heiko.strathmann@gmail.com>, OXPHOS <engelzora@gmail.com> | 11:09 |
@HeikoS | sanuj: gist allows you to have multiple files | 11:09 |
@HeikoS | and no ads | 11:09 |
sanuj | haha | 11:10 |
sanuj | HeikoS, https://gist.github.com/sanuj/695d881295c060d5fbb99abd3552fcbb | 11:10 |
@HeikoS | sanuj: ok thanks, thats much easier to look at | 11:11 |
sanuj | HeikoS, did you guys go to deepmind? | 11:12 |
@HeikoS | not yet | 11:12 |
@HeikoS | probably next week | 11:13 |
sanuj | okay :D | 11:13 |
@HeikoS | is the code in your PR exactly the same as when you ran the example? | 11:13 |
sanuj | i read some of their research papers....cool stuff | 11:13 |
@HeikoS | yeah they do interesthing work | 11:13 |
@HeikoS | and lots of really good people | 11:13 |
sanuj | HeikoS, code is same as PR | 11:14 |
@HeikoS | sanuj: what happens if you remove the line 16 in your example? | 11:14 |
@HeikoS | same error? | 11:14 |
sanuj | it segfaults | 11:14 |
sanuj | HeikoS, i think the error is in init_params() of mock object | 11:15 |
@HeikoS | sanuj: ^ | 11:15 |
sanuj | HeikoS, any destructor deletes the member variable | 11:16 |
sanuj | any.h destructor | 11:16 |
sanuj | not sure though | 11:16 |
sanuj | ==15743== at 0x4026FD: shogun::Any::~Any() (in /home/sanuj/Projects/shogun/buildpy/a.out) | 11:16 |
@HeikoS | sanuj: can you remove the set and tell me what happens? | 11:16 |
@HeikoS | line 16 | 11:17 |
sanuj | 5 mins | 11:17 |
sanuj | doorbell | 11:17 |
@HeikoS | kk | 11:18 |
sanuj | removing all sets | 11:19 |
@HeikoS | sanuj: | 11:20 |
@HeikoS | no | 11:20 |
@HeikoS | just line 16 in example | 11:20 |
@HeikoS | and add it to the gist | 11:21 |
@HeikoS | dont overwrite the old | 11:21 |
@HeikoS | so that we can compare | 11:21 |
sanuj | oh i had removed set in MockObject and built it | 11:29 |
sanuj | undoing and rebuilding | 11:29 |
sanuj | HeikoS, do you know someone in UCL who works on deep learning related stuff and can take interns like me for 6 months or so | 11:32 |
sanuj | or anywhere else also apart from UCL | 11:32 |
@HeikoS | sanuj: if you wanna do deeplearning, better go to deepmind | 11:32 |
@HeikoS | they took all good people | 11:32 |
@HeikoS | but I know a few folds at UCL | 11:32 |
@HeikoS | whod o ml | 11:32 |
sanuj | HeikoS, deepmind wont take people with just undergraduate degrees | 11:33 |
@HeikoS | i see | 11:35 |
@HeikoS | internships are good | 11:35 |
@HeikoS | need to find some good group | 11:36 |
@HeikoS | that matches you | 11:36 |
sanuj | HeikoS, really? | 11:37 |
sanuj | i have been mailing profs but no one replies back :P | 11:37 |
@HeikoS | sanuj: haha! | 11:38 |
@HeikoS | people are busy | 11:38 |
@HeikoS | and they get mails like this all the time | 11:38 |
@HeikoS | who did you write? | 11:38 |
sanuj | well i say that | 11:38 |
sanuj | i have read your recent work | 11:38 |
sanuj | respect what you do and would like to work with you to be a part of it | 11:39 |
sanuj | and send my CV | 11:39 |
sanuj | :D | 11:39 |
@HeikoS | who, not what | 11:39 |
sanuj | i updated the gist | 11:39 |
sanuj | oh | 11:39 |
sanuj | :D | 11:39 |
sanuj | lol | 11:40 |
sanuj | David hsu from NUS - works on robotics and reinforcement learning | 11:40 |
sanuj | Nando de freitas | 11:40 |
sanuj | honglak lee | 11:41 |
-!- GandalfTheWizard [~Eva@112.10.171.169] has quit [Quit: Leaving.] | 11:41 | |
@HeikoS | I see | 11:41 |
@HeikoS | they are all too busy | 11:41 |
sanuj | ^ univ of michigan | 11:41 |
sanuj | satinder singh from univ of michigan too | 11:42 |
sanuj | and some other profs who work on deep reinforcement learning and related stuff | 11:42 |
sanuj | now i have switched to mailing PhDs working in good labs to ask if they have a vacancy | 11:43 |
@HeikoS | I see | 11:43 |
@HeikoS | they usually cannot decide | 11:43 |
@HeikoS | ah it is hard | 11:43 |
@HeikoS | to stick out of the noise | 11:43 |
@HeikoS | I get lots of emails from indian students actually | 11:43 |
sanuj | HeikoS, haha | 11:43 |
@HeikoS | but takes too much time to answer all | 11:43 |
sanuj | yeah | 11:43 |
sanuj | HeikoS, tell me if you find a good group, i would like to join after gsoc | 11:44 |
@HeikoS | sanuj: what do you want to do? | 11:44 |
@HeikoS | that is the biggest question | 11:44 |
sanuj | HeikoS, i like what deepmind is doing | 11:45 |
sanuj | anything related to that | 11:45 |
sanuj | but i'm flexible | 11:45 |
@HeikoS | I see | 11:46 |
@HeikoS | mmh | 11:46 |
@HeikoS | and which place? | 11:47 |
@HeikoS | you like? | 11:47 |
sanuj | USA, Canada, UK, Europe, Singapore | 11:47 |
sanuj | HeikoS, will it be difficult in germany if i don't know german? :D | 11:48 |
@HeikoS | sanuj: some unis have english speaking groups | 11:48 |
@HeikoS | sanuj: ok so we get the same error when you remobed the set | 11:49 |
sanuj | yes | 11:49 |
@HeikoS | ok | 11:49 |
sanuj | it is from init_param() | 11:49 |
@HeikoS | so my hunch is: your tag implementation tries to delete the member variable | 11:49 |
@HeikoS | since its happening in Any destructor | 11:49 |
sanuj | yes, i think the same | 11:50 |
sanuj | but we need a way around it | 11:50 |
sanuj | i thought about removing that from any.h destructor and using something like SG_REMOVE | 11:50 |
@HeikoS | I dont understand the problem fully yet | 11:51 |
sanuj | SG_FREE* | 11:51 |
sanuj | okay | 11:51 |
@HeikoS | can you maybe re-build your example without optimizations? | 11:54 |
@HeikoS | so it doesnt say | 11:54 |
@HeikoS | at 0x4024A9: shogun::Any::~Any() (in /home/sanuj/Projects/shogun/buildpy/a.out) | 11:54 |
-!- GandalfTheWizard [~Eva@112.10.171.169] has joined #shogun | 11:54 | |
@HeikoS | but gives the line number at the end? | 11:54 |
sanuj | okay | 11:54 |
@HeikoS | I dont understand why the desctructor is called | 11:54 |
sanuj | HeikoS, debug mode in cmake? | 11:54 |
@HeikoS | nono | 11:54 |
@HeikoS | sanuj: check the error output | 11:55 |
@HeikoS | shogun's line numbers are there | 11:55 |
@HeikoS | but not the ones of your example | 11:55 |
@HeikoS | there if only sais a.out | 11:55 |
@HeikoS | but no line number | 11:55 |
sanuj | oh i see | 11:55 |
@HeikoS | sanuj: so the problem is: registering member variables in the parameter map | 11:56 |
@HeikoS | we are already down to that | 11:56 |
sanuj | yeah | 11:57 |
@HeikoS | updated? | 12:00 |
sanuj | HeikoS, i used g++ -O0 | 12:01 |
sanuj | not getting the line numbers | 12:01 |
@HeikoS | there is a debug option | 12:01 |
@HeikoS | that includes line number infos | 12:01 |
sanuj | HeikoS, updating | 12:03 |
sanuj | HeikoS, updated | 12:04 |
@HeikoS | checking | 12:05 |
@HeikoS | the example line numbers dont matcht he listing you put | 12:06 |
@HeikoS | line 17 in error message is nothing | 12:06 |
@HeikoS | I think since you remove a line | 12:06 |
sanuj | HeikoS, it's line 12 in the error message for temp.cpp | 12:07 |
@HeikoS | ah yeah | 12:08 |
@HeikoS | sorry | 12:08 |
@HeikoS | ok then | 12:08 |
@HeikoS | so why is desctructor called? | 12:08 |
@HeikoS | of any | 12:08 |
@HeikoS | CSGObject::485 | 12:09 |
@HeikoS | erase_type | 12:09 |
@HeikoS | why is that done? | 12:09 |
@HeikoS | did you maybe update the "overwritign" machanism and forget to consider the case where something is newly added to the parameter map? | 12:10 |
sanuj | HeikoS, erase_type creates an object of Any | 12:10 |
sanuj | using the value from the argument | 12:11 |
sanuj | because parameter map stores string --> any | 12:11 |
sanuj | HeikoS, sorry, it stores BaseTag ---> any | 12:11 |
sanuj | HeikoS, the diff between public set and protected set is that public set only sets if the provided basetag already exists in the map | 12:13 |
@HeikoS | sanuj: maybe a better name should be chosen | 12:13 |
@HeikoS | like "add" | 12:13 |
@HeikoS | or so | 12:13 |
@HeikoS | and "set" is for public use | 12:13 |
sanuj | currently it is _set | 12:13 |
sanuj | i'll use add | 12:13 |
@HeikoS | so why is any destructor called? | 12:14 |
sanuj | even i'm asking the same question | 12:14 |
@HeikoS | from erase_value? | 12:15 |
@HeikoS | erase_type( | 12:15 |
sanuj | HeikoS, i had used the exact same method in the old implementation | 12:15 |
@HeikoS | https://github.com/shogun-toolbox/shogun/pull/3221/files#diff-8ea96286d95b52029d31636117e0fe55R210 | 12:15 |
sanuj | i just made it protected | 12:15 |
@HeikoS | there you return by value | 12:16 |
sanuj | yes | 12:16 |
@HeikoS | sanuj: so the way to solve this | 12:16 |
@HeikoS | is to use a debugger | 12:16 |
@HeikoS | and to go through the code step by step | 12:16 |
@HeikoS | understand what it does | 12:16 |
@HeikoS | I cannot do that from the PR | 12:16 |
@HeikoS | since I cannot brwose the code | 12:16 |
sanuj | i'll do that | 12:17 |
@HeikoS | too many redirections for my mind :) | 12:17 |
@HeikoS | sanuj: you know how to use gdb? | 12:17 |
@HeikoS | or ddd maybe? | 12:17 |
sanuj | now i'll find learn it from somewhere | 12:17 |
@HeikoS | if you already compiled with debug symbols it is easy | 12:17 |
@HeikoS | which OS? | 12:17 |
sanuj | mint | 12:17 |
sanuj | which is ubunut | 12:17 |
sanuj | ubuntu | 12:17 |
@HeikoS | ok install "ddd" | 12:17 |
-!- GandalfTheWizard [~Eva@112.10.171.169] has quit [Quit: Leaving.] | 12:19 | |
@HeikoS | sanuj: and then start it, and just load your a.out | 12:19 |
@HeikoS | then set a breakpoint before the line 12 | 12:19 |
@HeikoS | and then run | 12:19 |
@HeikoS | and then to "single steps" | 12:19 |
sanuj | HeikoS, btw can you merge https://github.com/shogun-toolbox/shogun/pull/3283 | 12:19 |
@HeikoS | if it goes to far in, you can always "step return" | 12:20 |
sanuj | this is holding my neural net cookbook since 2 weeks | 12:20 |
@HeikoS | sanuj: did you ever use any debugger? | 12:20 |
@HeikoS | i.e. you know "step next" "step return" etc? | 12:20 |
@HeikoS | and breakpoints | 12:20 |
@HeikoS | ? | 12:20 |
sanuj | yeah | 12:20 |
sanuj | i know | 12:20 |
@HeikoS | cool | 12:20 |
@HeikoS | so ddd does that for your c | 12:20 |
@HeikoS | binary | 12:20 |
@HeikoS | given it is compiled with debug mode | 12:20 |
sanuj | okay | 12:20 |
sanuj | oh wow, it has a gui | 12:21 |
@HeikoS | yeah | 12:22 |
@HeikoS | sanuj: so use that to find out whats wrong | 12:22 |
@HeikoS | you can just read what all the variables are | 12:22 |
@HeikoS | good exercise to understand c memory errors ;) | 12:23 |
sanuj | doing | 12:23 |
@HeikoS | sanuj: I am here for a bit more, let me know your findings | 12:23 |
sanuj | okay | 12:23 |
@HeikoS | sanuj: and? | 12:29 |
sanuj | i set a breakpoint on line 11 | 12:30 |
sanuj | 10 actually | 12:31 |
sanuj | then i do next | 12:31 |
sanuj | and it reaches the any destructor | 12:31 |
sanuj | HeikoS, ^ | 12:33 |
@HeikoS | you gotta do single steps | 12:33 |
@HeikoS | or put the breakpoint earlier, sometimes ddd is confusing | 12:33 |
@HeikoS | "step in" | 12:34 |
@HeikoS | not "step over" | 12:34 |
@HeikoS | step over exectues all code by the active line | 12:34 |
@HeikoS | but you want to jump inside the call | 12:34 |
sanuj | HeikoS, there is "step" and "next" | 12:34 |
@HeikoS | sanuj: I dont know how they are called | 12:34 |
sanuj | "next" says = proceeding through subroutine calls | 12:34 |
@HeikoS | but if it goes to destructor, it obvioulsy tries to execute the whole line | 12:34 |
sanuj | yeah | 12:34 |
sanuj | oh okay | 12:35 |
sanuj | wait | 12:35 |
@HeikoS | gotta go | 12:39 |
@HeikoS | sanuj: will check back later tonight | 12:39 |
@HeikoS | feel free to send an email | 12:39 |
@HeikoS | should be easy to track the problem with the ddd | 12:39 |
@HeikoS | sanuj: see you! | 12:39 |
sanuj | HeikoS, cool | 12:39 |
sanuj | thanks | 12:39 |
@HeikoS | thanks to you too! :) | 12:39 |
sanuj | :D | 12:39 |
-!- HeikoS [~heiko@AAubervilliers-652-1-253-126.w83-112.abo.wanadoo.fr] has quit [Quit: Leaving.] | 12:39 | |
-!- sanuj [~sanuj@117.220.49.228] has quit [Ping timeout: 244 seconds] | 12:44 | |
-!- lambday [569de6b3@gateway/web/freenode/ip.86.157.230.179] has joined #shogun | 12:46 | |
-!- mode/#shogun [+o lambday] by ChanServ | 12:46 | |
@wiking | Saurabh7__: | 13:03 |
@wiking | ping | 13:03 |
-!- sanuj [~sanuj@117.220.49.228] has joined #shogun | 13:13 | |
-!- lambday [569de6b3@gateway/web/freenode/ip.86.157.230.179] has quit [Ping timeout: 250 seconds] | 13:15 | |
-!- travis-ci [~travis-ci@ec2-54-234-46-43.compute-1.amazonaws.com] has joined #shogun | 13:41 | |
travis-ci | it's Heiko Strathmann's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/140201417 | 13:41 |
-!- travis-ci [~travis-ci@ec2-54-234-46-43.compute-1.amazonaws.com] has left #shogun [] | 13:41 | |
sanuj | BZDMG, hey | 14:03 |
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout] | 14:08 | |
Saurabh7__ | wiking: yo | 14:44 |
shogun-buildbot | build #569 of deb1 - libshogun - PR is complete: Failure [failed cookbook] Build details are at http://buildbot.shogun-toolbox.org/builders/deb1%20-%20libshogun%20-%20PR/builds/569 blamelist: sanuj | 15:41 |
shogun-buildbot | build #570 of deb1 - libshogun - PR is complete: Success [build successful] Build details are at http://buildbot.shogun-toolbox.org/builders/deb1%20-%20libshogun%20-%20PR/builds/570 | 15:44 |
-!- sanuj [~sanuj@117.220.49.228] has quit [Ping timeout: 264 seconds] | 16:49 | |
-!- sanuj [~sanuj@117.220.49.228] has joined #shogun | 17:21 | |
-!- Saurabh7__ [Saurabh7@gateway/shell/panicbnc/x-wlacyikiyzqrkfcs] has left #shogun [] | 18:57 | |
sanuj | BZDMG, there? | 19:27 |
-!- sanuj [~sanuj@117.220.49.228] has quit [Remote host closed the connection] | 19:37 | |
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun | 22:21 | |
shogun-notifier- | shogun: Sanuj :develop * 4375d99 / src/interfaces/modular/NeuralNets.i,src/shogun/neuralnets/NeuralLayers.h: https://github.com/shogun-toolbox/shogun/commit/4375d996461069838c21ee0608b22fad425c2dbb | 22:21 |
shogun-notifier- | shogun: add %newobject in neuralnet interface | 22:21 |
shogun-notifier- | shogun: Sergey Lisitsyn :develop * 8cb801e / src/interfaces/modular/NeuralNets.i,src/shogun/neuralnets/NeuralLayers.h: https://github.com/shogun-toolbox/shogun/commit/8cb801eaba4c563282e3aac57575921e63cacfa0 | 22:21 |
shogun-notifier- | shogun: Merge pull request #3283 from sanuj/nn_cleanup | 22:21 |
shogun-notifier- | shogun: | 22:21 |
shogun-notifier- | shogun: add %newobject in neuralnet interface | 22:21 |
shogun-buildbot | build #740 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/740 blamelist: Sanuj <sanuj.sharma.in@gmail.com>, Sergey Lisitsyn <lisitsyn.s.o@gmail.com> | 22:21 |
-!- sonne|osx [~sonne@x4e345856.dyn.telefonica.de] has joined #shogun | 23:36 | |
-!- sonne|osx [~sonne@x4e345856.dyn.telefonica.de] has quit [Quit: sonne|osx] | 23:57 | |
--- Log closed Sun Jun 26 00:00:56 2016 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!