--- Log opened Thu Feb 16 00:00:27 2017 | ||
@wiking | lisitsyn, you see, JIT https://haosdent.gitbooks.io/tensorflow-document/content/resources/xla_prerelease.html :) | 01:07 |
---|---|---|
@sukey | Pull Request #3608 "Clean up KNN" synchronized by MikeLing - https://github.com/shogun-toolbox/shogun/pull/3608 | 02:04 |
@sukey | Pull Request #3620 "Add time comparison for KNN_KDTREE" synchronized by MikeLing - https://github.com/shogun-toolbox/shogun/pull/3620 | 04:18 |
-!- remote [~remote@portemanteau.ca] has left #shogun [] | 04:26 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has quit [Remote host closed the connection] | 04:59 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has joined #shogun | 05:00 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has quit [Ping timeout: 240 seconds] | 05:05 | |
-!- Guest85308 [~hazzard@142.54.173.74] has joined #shogun | 06:01 | |
-!- Guest85308 [~hazzard@142.54.173.74] has quit [K-Lined] | 06:01 | |
@sukey | Pull Request #3620 "Add time comparison for KNN_KDTREE" synchronized by MikeLing - https://github.com/shogun-toolbox/shogun/pull/3620 | 06:02 |
-!- Guest78383 [~haha@223.13.66.34] has joined #shogun | 08:09 | |
-!- Guest78383 [~haha@223.13.66.34] has quit [Ping timeout: 258 seconds] | 08:18 | |
-!- Guest23603 [~ff@223.13.66.34] has joined #shogun | 09:35 | |
-!- Guest23603 [~ff@223.13.66.34] has left #shogun [] | 09:38 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has joined #shogun | 09:44 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has quit [Ping timeout: 240 seconds] | 09:49 | |
-!- abhinav_ [~abhinav@202.173.126.154] has joined #shogun | 10:09 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has joined #shogun | 10:15 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has quit [Ping timeout: 240 seconds] | 10:20 | |
-!- abhinav_ [~abhinav@202.173.126.154] has quit [Ping timeout: 260 seconds] | 10:34 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has joined #shogun | 10:46 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has quit [Ping timeout: 240 seconds] | 10:51 | |
-!- abhinav_ [~abhinav@202.173.126.154] has joined #shogun | 10:55 | |
-!- abhinav_ [~abhinav@202.173.126.154] has quit [Quit: Leaving] | 11:10 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has joined #shogun | 11:18 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has quit [Ping timeout: 240 seconds] | 11:23 | |
-!- ShellcatZero [~ShellcatZ@cpe-76-93-136-67.san.res.rr.com] has quit [Ping timeout: 264 seconds] | 12:13 | |
-!- ShellcatZero [~ShellcatZ@cpe-76-93-136-67.san.res.rr.com] has joined #shogun | 12:16 | |
-!- HeikoS [~heiko@untrust-out.swc.ucl.ac.uk] has joined #shogun | 13:11 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 13:11 | |
@sukey | Pull Request #3608 "Clean up KNN" synchronized by MikeLing - https://github.com/shogun-toolbox/shogun/pull/3608 | 13:12 |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has joined #shogun | 13:19 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has quit [Ping timeout: 240 seconds] | 13:23 | |
@sukey | Wiki page: Google-Summer-of-Code-2017-Projects edited on shogun-toolbox/shogun by karlnapf | 13:24 |
@sukey | Wiki page: Google-Summer-of-Code-2017-Projects edited on shogun-toolbox/shogun by karlnapf | 13:24 |
CaBa | HeikoS: heya | 13:26 |
CaBa | HeikoS: have a minute? | 13:27 |
@sukey | Issue #3629 "Polish class API documentation" opened by karlnapf - https://github.com/shogun-toolbox/shogun/issues/3629 | 13:28 |
@sukey | Issue #3629 "Polish class API documentation" karlnapf added label: "entrance" - https://github.com/shogun-toolbox/shogun/issues/3629 | 13:28 |
@sukey | Issue #3629 "Polish class API documentation" karlnapf added label: "Documentation" - https://github.com/shogun-toolbox/shogun/issues/3629 | 13:28 |
CaBa | HeikoS: i'm writing that quick and dirty version of the customkernel that we were talking about the other day right now ;) | 13:28 |
@sukey | Wiki page: GSoC_2017_low_hanging_fruit edited on shogun-toolbox/shogun by karlnapf | 13:28 |
@HeikoS | CaBa: hi | 13:29 |
@HeikoS | yes | 13:29 |
CaBa | HeikoS: and i guess i'm running into problems that already came up with CustomKernel, so i thought to ask l) | 13:29 |
CaBa | HeikoS: for one, what is the scenario where lhs != rhs in a kernel? i guess that cannot be covered with precalculated kernels, right? | 13:30 |
@HeikoS | yes | 13:30 |
@HeikoS | this is when you want to apply the method to a new datapoint | 13:30 |
@HeikoS | i.e. you trained on X and then want to apply to Y that you havent seen when computing the kernel matrix for X | 13:30 |
@HeikoS | so in that case you need to precompute the kernel between X and Y before you can predict | 13:31 |
CaBa | HeikoS: so how's that handled in the case of CCustomKernel? I saw some C++ code where ->set_kernel() was called with the testing precalc matrix... | 13:31 |
@HeikoS | you need to provide the kernel if you want to predict on new data | 13:31 |
@HeikoS | as f(x) = \sum_i alpha_i k(x_new, x_i) | 13:32 |
@HeikoS | wiking: jo around? | 13:32 |
@HeikoS | I got back from gunnar! finally :) | 13:33 |
@wiking | HeikoS, yes | 13:33 |
@wiking | sup? | 13:33 |
@HeikoS | - Feb 21, 2 or 3pm- Feb 22, 10 or 11am | 13:33 |
@HeikoS | whats better=? | 13:33 |
@HeikoS | the 11am I guess? | 13:33 |
@wiking | 11 yes | 13:33 |
@HeikoS | cool | 13:33 |
@HeikoS | let me reply | 13:33 |
@wiking | but i can do 3 if need be | 13:33 |
@HeikoS | 10 is better? | 13:33 |
@HeikoS | because it is 10 or 11 | 13:33 |
@wiking | either is fine | 13:34 |
@HeikoS | ok then 11 better for me | 13:34 |
@HeikoS | wiking: cool | 13:35 |
@HeikoS | Ill write a little plan until then so we can present concrete things to gunnar and he just has to yes/no them | 13:35 |
@HeikoS | wiking: saw this https://github.com/shogun-toolbox/shogun/pull/3534#issuecomment-280154766 | 13:35 |
@HeikoS | is that the one you warn about in cmake? | 13:36 |
@HeikoS | could do that as well and disable the test for that version? | 13:36 |
@wiking | dunno yet | 13:36 |
@wiking | i have to finish somethng | 13:36 |
@HeikoS | kk | 13:36 |
CaBa | HeikoS: https://github.com/shogun-toolbox/shogun/blob/develop/examples/undocumented/libshogun/classifier_svmlight_string_features_precomputed_kernel.cpp#L110 | 13:39 |
CaBa | HeikoS: so that's the training / testing scheme for custom kernel? | 13:39 |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has joined #shogun | 13:39 | |
@sukey | New Commit "remove message about unused variable warnings as they will just confuse users" to shogun-toolbox/shogun by karlnapf: https://github.com/shogun-toolbox/shogun/commit/6599b893c731e246e54bd140b4f6a8d858eb864f | 13:40 |
@HeikoS | CaBa: kind of | 13:41 |
@HeikoS | just here all data is known at training time | 13:41 |
@HeikoS | do you can precompute the test kernel matri xas well | 13:41 |
@HeikoS | but yeah | 13:41 |
@HeikoS | you see below that the precomputed test kernel matrices are passed | 13:41 |
@HeikoS | so it is that | 13:42 |
CaBa | ah, so one approach is to concat training + testing data, then subset for training and subset for testing? | 13:42 |
@HeikoS | no | 13:44 |
@HeikoS | that is just one way to do it | 13:44 |
@HeikoS | but you could also imagine | 13:44 |
@HeikoS | you traing your model on all data | 13:44 |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has quit [Ping timeout: 240 seconds] | 13:44 | |
@HeikoS | and then, afterwards, you get new test data | 13:44 |
@HeikoS | and then you compute the kernel between train and test (no subsetting) | 13:44 |
@HeikoS | and then prediuct | 13:44 |
@HeikoS | here it is done with the subsetting since all data is known from the start | 13:44 |
CaBa | yes. which will be the case for my usecase, too. sure, generally it is desirable to be able to work with new data at a later time | 13:48 |
CaBa | HeikoS: is there a code snippet for "solution 1" somewhere? | 13:49 |
@HeikoS | CaBa: no dont think so | 13:49 |
@HeikoS | but it is easy | 13:49 |
@HeikoS | you just need to create the km_test matrix in your snippet above manually | 13:50 |
@HeikoS | without subsetting | 13:50 |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has joined #shogun | 14:11 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has quit [Ping timeout: 240 seconds] | 14:16 | |
-!- travis-ci [~travis-ci@ec2-54-158-130-127.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: https://travis-ci.org/shogun-toolbox/shogun/builds/201982436 | 14:28 |
-!- travis-ci [~travis-ci@ec2-54-158-130-127.compute-1.amazonaws.com] has left #shogun [] | 14:28 | |
-!- mikeling [uid89706@gateway/web/irccloud.com/x-heyxhvoscogkbdtr] has joined #shogun | 14:33 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has joined #shogun | 14:43 | |
@sukey | Pull Request #3608 "Clean up KNN" synchronized by MikeLing - https://github.com/shogun-toolbox/shogun/pull/3608 | 14:46 |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has quit [Ping timeout: 240 seconds] | 14:47 | |
-!- travis-ci [~travis-ci@ec2-54-158-130-127.compute-1.amazonaws.com] has joined #shogun | 15:25 | |
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/202230853 | 15:25 |
-!- travis-ci [~travis-ci@ec2-54-158-130-127.compute-1.amazonaws.com] has left #shogun [] | 15:25 | |
-!- suhas2go [uid201652@gateway/web/irccloud.com/x-dozyivljzebrwbwh] has joined #shogun | 15:44 | |
CaBa | HeikoS: when CCustomKernel is initialized with a matrix, it creates CIndexFeatures somewhere internally? | 16:09 |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has joined #shogun | 16:20 | |
mikeling | hi HeikoS: I found there are several sub-projects for Shogun detox 2, is that mean I need to pick some sub-projects from them if I want to work on Shogun detox 2 as GSoC project? | 16:24 |
@HeikoS | CaBa: I think so | 16:25 |
@HeikoS | mikeling: hi! | 16:25 |
CaBa | HeikoS: i can't find that anywhere... | 16:25 |
@HeikoS | mikeling: yeah we will probably not be able to do them all | 16:25 |
@HeikoS | mikeling: this list will need to be made much more concrete | 16:25 |
mikeling | Sp, is Shogun detox 2 an independent project ? is that possible more than one students work on it? | 16:26 |
@HeikoS | mikeling: it also depends on how much of the last years projects we manage to fully integrate until gsoc 2016 starts | 16:26 |
@HeikoS | mikeling: I think it will be independent, but maybe a small subproject will be soved by another student, like the efficiency one or others | 16:26 |
@HeikoS | CaBa: why? | 16:26 |
CaBa | HeikoS: well i'm currently using 1-dimensional CDenseFeatures<uint32_t> for my own custom kernel hack, was wondering if CIndexFeatures would be more straight forward to use. so i was looking whether i can see them in action somewher | 16:28 |
@HeikoS | CaBa: I currently dont have a good idea in my head what the IndexFeatures do, but I did not write that stuff initially | 16:29 |
CaBa | ^^ | 16:29 |
mikeling | HeikoS: great! Thank you for your kindly explanation :) | 16:30 |
CaBa | HeikoS: https://github.com/shogun-toolbox/shogun/blob/develop/src/shogun/kernel/CustomKernel.cpp#L105 | 16:30 |
@HeikoS | mikeling: no worries, looking forward to your application | 16:30 |
CaBa | HeikoS: looks as if CDummyFeatures are used and not CIndexFeatures | 16:30 |
@HeikoS | mikeling: definitely have a look at the cereal / tags / linalg branch from last years gsoc | 16:30 |
mikeling | HeikoS: great! Thank you for your kindly explanation :) | 16:31 |
CaBa | HeikoS: whereas i don't understand why this is initialized with 2 dummy features, one for kmatrix rows and one for cols... it should reflect the number of feature vectors, right? O.o | 16:33 |
@sukey | Pull Request #3608 "Clean up KNN" merged by vigsterkr - https://github.com/shogun-toolbox/shogun/pull/3608 | 16:50 |
@sukey | New Commit "Merge pull request #3608 from MikeLing/clean_up_KNN | 16:50 |
@sukey | Clean up KNN" to shogun-toolbox/shogun by vigsterkr: https://github.com/shogun-toolbox/shogun/commit/988fe1e4c3a39e1ca32f91d4a1933661bdbe2f07 | 16:50 |
@wiking | ok this was a big one | 16:50 |
@wiking | :) | 16:50 |
CaBa | subsetting... is that a kernel thing or a feature thing? | 16:57 |
CaBa | feature it seems. and seems like dummyfeatures don't really support it but indexfeatures do. in a rather expensive fashion O.o | 16:59 |
-!- travis-ci [~travis-ci@ec2-54-147-160-215.compute-1.amazonaws.com] has joined #shogun | 17:43 | |
travis-ci | it's Viktor Gal'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/202296292 | 17:43 |
-!- travis-ci [~travis-ci@ec2-54-147-160-215.compute-1.amazonaws.com] has left #shogun [] | 17:43 | |
@sukey | Issue #3630 "Shogun error in compilation at 68% " opened by ahmedmazari - https://github.com/shogun-toolbox/shogun/issues/3630 | 17:50 |
@sukey | Issue #3630 "Shogun error in compilation at 68% "- https://github.com/shogun-toolbox/shogun/issues/3630 | 18:13 |
@sukey | Issue #3630 "Shogun error in compilation at 68% "- https://github.com/shogun-toolbox/shogun/issues/3630 | 18:13 |
-!- mikeling [uid89706@gateway/web/irccloud.com/x-heyxhvoscogkbdtr] has quit [Quit: Connection closed for inactivity] | 18:39 | |
@HeikoS | CaBa: well both! | 18:59 |
@HeikoS | CaBa: we are always happy to take patches :) | 19:08 |
CaBa | HeikoS: hm, i know. and i'm sure you don't want what i'm cooking right now :D | 19:08 |
@HeikoS | hehe ok | 19:08 |
@sukey | Pull Request #3617 "Port sgd svm example to meta" closed by karlnapf - https://github.com/shogun-toolbox/shogun/pull/3617 | 19:10 |
lisitsyn | HeikoS: I'm trying to rebase feature/tags now | 19:33 |
lisitsyn | on top of develop | 19:33 |
lisitsyn | uhm | 19:35 |
lisitsyn | why | 19:35 |
lisitsyn | hahah | 19:35 |
lisitsyn | I totally forgot we merged it in | 19:36 |
lisitsyn | https://github.com/shogun-toolbox/shogun/blob/develop/src/shogun/loss/SmoothHingeLoss.h#L9 | 19:39 |
lisitsyn | TechnosmoothHingey | 19:39 |
lisitsyn | O_o | 19:40 |
-!- abhinav [~abhinav@103.211.19.116] has joined #shogun | 19:49 | |
-!- HeikoS [~heiko@untrust-out.swc.ucl.ac.uk] has quit [Ping timeout: 240 seconds] | 20:06 | |
@sukey | Pull Request #3597 "Add Scala Interface" synchronized by abhinavrai44 - https://github.com/shogun-toolbox/shogun/pull/3597 | 20:12 |
@sukey | Issue #3459 "Compilation error with g++ 6.1" closed by vigsterkr - https://github.com/shogun-toolbox/shogun/issues/3459 | 20:23 |
@sukey | Issue #3631 "How to import shogun in python ?" opened by ahmedmazari - https://github.com/shogun-toolbox/shogun/issues/3631 | 20:42 |
-!- abhinav [~abhinav@103.211.19.116] has quit [Ping timeout: 240 seconds] | 20:45 | |
@sukey | Pull Request #3599 "fix invalid use of pre-declared class (#3459)" closed by besser82 - https://github.com/shogun-toolbox/shogun/pull/3599 | 21:21 |
--- Log closed Fri Feb 17 00:00:28 2017 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!