--- Log opened Fri Jun 16 00:00:15 2017
-!- olinguyen [81615ad9@gateway/web/freenode/ip.] has quit [Quit: Page closed]00:47
-!- TingMiao [uid229534@gateway/web/irccloud.com/x-nkuhyptdjjkmkseh] has quit [Quit: Connection closed for inactivity]01:42
-!- mikeling [uid89706@gateway/web/irccloud.com/x-eslhbcvcdmsjzmai] has joined #shogun03:08
shogitter(iteachmachines) Hello everyone i'm priyam and i am interested in contributing to shogun as i am new in the field of open source software,can anyone guide me please?03:17
mikelingshogitter: Hey, I guess read this https://github.com/shogun-toolbox/shogun/wiki/Getting-involved is a good start to contribute :)03:27
-!- olinguyen [488ddefa@gateway/web/freenode/ip.] has joined #shogun04:29
-!- olinguyen [488ddefa@gateway/web/freenode/ip.] has quit [Client Quit]04:29
-!- TingMiao [uid229534@gateway/web/irccloud.com/x-zthrpfylkpsrvdcd] has joined #shogun04:29
mikelingwiking: ping05:14
mikelingAfter register std::vector by add_vector, what's the m_parameter for? https://pastebin.mozilla.org/902484505:16
mikelingwiking: I found we have this problem https://pastebin.mozilla.org/9024846 is because  m_parameter is actually point to nothing https://pastebin.mozilla.org/9024847, so it can't be warp as CSGObject**05:20
mikelingwiking: forget it, I was wrong06:40
-!- geektoni [~geektoni@93-34-234-212.ip52.fastwebnet.it] has joined #shogun08:26
-!- geektoni [~geektoni@93-34-234-212.ip52.fastwebnet.it] has quit [Remote host closed the connection]08:34
-!- wiking_mob [~Mutter@] has joined #shogun09:38
-!- wiking_mob [~Mutter@] has quit [Quit: Mutter: www.mutterirc.com]09:49
-!- geektoni [~Mutter@] has joined #shogun10:27
-!- geektoni [~Mutter@] has quit [Client Quit]10:27
-!- wiking_mob [~Mutter@] has joined #shogun10:45
-!- wiking_mob [~Mutter@] has quit [Remote host closed the connection]10:50
-!- wiking_mob [~Mutter@] has joined #shogun11:11
-!- wiking_mob [~Mutter@] has quit [Client Quit]11:14
shogitter(iteachmachines) @bazdmeg  Thank you!11:45
-!- TingMiao [uid229534@gateway/web/irccloud.com/x-zthrpfylkpsrvdcd] has quit [Quit: Connection closed for inactivity]14:20
-!- TingMiao [uid229534@gateway/web/irccloud.com/x-xbcsnzdqkuwblyri] has joined #shogun15:10
-!- mikeling [uid89706@gateway/web/irccloud.com/x-eslhbcvcdmsjzmai] has quit [Quit: Connection closed for inactivity]15:26
-!- mikeling [uid89706@gateway/web/irccloud.com/x-rovmvjrxlvkouswc] has joined #shogun15:42
@iglesiasgTingMiao: hey15:58
@iglesiasghow's it going?15:58
@iglesiasgTingMiao: so, let's see how are the features you are using built (the ones you mentioned in the e-mail)16:00
-!- HeikoS [~heiko@host-92-0-178-129.as43234.net] has joined #shogun16:02
-!- mode/#shogun [+o HeikoS] by ChanServ16:02
TingMiaohttps://github.com/tingpan/shogun-data#play-type-statsmost-important-stats-in-this-project these are features I am using for clustering16:03
@iglesiasgTingMiao: in the kmeans notebook, you found out a reasonable number to cluster players into different groups. That number was 1316:03
TingMiaoYes, and I have validated players are similar in each group16:03
TingMiao13 is reasonable for clustering players into different types16:03
@iglesiasgso far so good16:04
TingMiaoI was wondering if their types are important to game winning16:04
@iglesiasgnow, you mention something about the team score and player total scores16:04
TingMiaoIt may only similar, but do not influence game result:(16:04
@iglesiasgso, where do these player scores come from?16:04
TingMiaoplayer scores = time * efficiency of scoring16:05
@iglesiasgmmm ok so that only takes into account the number of points they score, right?16:05
@iglesiasgit does not use the play types you used in the clustering16:06
TingMiaoactually I represent each team with the average score(which is the player score) and played-time percentage for each type.16:06
TingMiaoso every team has 2 * 13 feature16:07
TingMiao13 is number if types16:07
@iglesiasgthat is not clear to me16:07
TingMiaoEach team would have 13 types of player16:08
TingMiaoI am not using score of each player16:08
TingMiaoI use score of each type of player, so that each team would have 13 types and each types would have percent and scores16:08
@iglesiasgall right16:09
TingMiaoIs this clear enough? Sorry for previous description I think I missed some important points:(16:09
@iglesiasgnot yet16:10
@iglesiasgit is fine, that's why we are talking about it now :)16:10
@iglesiasgso right now I would understand that a team has a score that consists of 13 numbers16:11
@iglesiasgone for each of the player types16:11
@iglesiasgwhere does the 2*13 or 2*2*13 come from?16:12
TingMiaoI used other attribute called 'time percent of type' so each type would have two features16:13
TingMiaothis shows how much time one type of player plays in one team16:13
TingMiaothis is 2*1316:13
@iglesiasggot it16:13
TingMiaoand 2*2*13 because there are 2 teams in one game16:13
@iglesiasg\o/ understood16:13
TingMiaoso every game would have 52 features, is this too many?16:14
@iglesiasgno, it is not16:14
TingMiaoprediction of training set is good(around 90%) but prediction of test set is only 67%. That's why I want to find out some way to fix the overfited16:15
@iglesiasg52 features is not really even high-dimensional data16:15
@iglesiasgthis type of high/low dimensional classification is always a bit relative; but yeah, 52 features it not that many, don't worry about that for the moment16:16
@iglesiasgyeah, that gap between train set and test set looks like there could be some overfitting16:16
@iglesiasgbut before getting into how we can address that16:16
@iglesiasgfirst, let me understand a bit more how do you come up with the team scores16:17
@iglesiasgyou mentioned: player scores = time * efficiency of scoring16:18
@iglesiasgdoes that mean that the played time (by a player) is considered twice?16:18
TingMiaoyes, and I will add up this score of players of each types so this is the 'player type score'16:19
TingMiaoI try only use percent but result was not good as well16:19
TingMiaopercent I mean played time16:19
@iglesiasgok, I see. We could revisit this thing of using the time twice later16:20
@iglesiasgthe played time is different for each match, right?16:21
@iglesiasgso that the feature vector corresponding to Team A is different in say Match 1 and Match 216:21
@iglesiasglet me put that a bit more detailed16:21
@iglesiasgMatch1: Team A - Team B16:21
@iglesiasgMatch2: Team A - Team C16:21
TingMiaoI will use the total play time of a season and convert it to percent of total team play time16:21
@iglesiasgoh I see16:22
TingMiaobecause I could not get the play time of a game before it start:(16:22
@iglesiasgI don't really understand why that is an issue16:22
@iglesiasglet me think about it16:23
TingMiaothanks for your time:)16:23
@iglesiasgsure, np16:23
@iglesiasgmm ok16:24
@iglesiasgso, right now a certain Team always gets the same features for every match it participates in16:24
@iglesiasgTingMiao: yes ^?16:26
TingMiaoyes since they are all history data16:26
@iglesiasgok ok, that's good16:26
@iglesiasgthat could be something that could be done differently when applying this approach to predict the results of future matches16:27
@iglesiasgas you said, it is not reasonable to assume we know the time each player will play in a match beforehand16:27
TingMiaoif I am predicting future games16:27
@iglesiasgbut it is reasonable to assume which players will play a match16:27
TingMiaoI would update feature of each team before every game basic on recent games16:27
@iglesiasgbut for the time being it is ok16:27
TingMiaobase not basic16:28
@iglesiasggot it16:28
@iglesiasgnow we focus on this exercise of predicting match results for the previous season using the data for the complete season, that's good16:28
@iglesiasgso, how is the score of a player computed?16:29
TingMiaoyes and currently I am using 90% of games in one season as training set and 10% for test.16:29
TingMiaoscore is time * efficiency. Actually it is total points that a player get.16:30
-!- olinguyen [81615ad9@gateway/web/freenode/ip.] has joined #shogun16:30
@iglesiasgTingMiao: all right, so it is only based on the number of points the player scored16:30
@iglesiasgok, I think the full picture is clear to me now16:31
@iglesiasgI think it will be a good idea to add cross-validation to make a good selection of the parameters you mentioned in the e-mail16:33
@iglesiasgTingMiao: I think this is a good video introducing the concept: https://www.youtube.com/watch?v=hihuMBCuSlU16:34
TingMiaooh thanks I will check this:)16:35
@iglesiasgthat is only a bit in general to get familiar with the concept16:35
@iglesiasgI am looking for an example so you can see how to do this in Shogun16:35
@iglesiasgalso what are the parameters that could be interesting tuning for you random forest16:36
TingMiaomaybe i could read this one http://www.shogun-toolbox.org/notebook/latest/xval_modelselection.html16:37
@iglesiasgHeikoS: ping16:39
@HeikoSiglesiasg:  jo16:40
@iglesiasgHeikoS: the num_bags of CRadomForest is automagically tuned internally?16:43
@iglesiasg"The feature for calculating out-of-box error is also provided to help determine the16:43
@HeikoSI dont know16:43
@iglesiasg * appropriate number of bags.16:43
@iglesiasgok oko16:43
@HeikoSiglesiasg: you mentored parijat :D16:43
@iglesiasgHeikoS: it fallsback to bagging :P16:44
@iglesiasgCRandomForest is actually quite barebones16:44
@HeikoSiglesiasg: if you find problems, please open issues16:44
@iglesiasgI didn't ask the question very well16:44
@HeikoSso we can fix things16:44
@iglesiasgI meant more from a conceptual point of view if that makes sense16:44
@iglesiasgor perhaps from a practical point of view actually16:44
@iglesiasgwhen using RandomForest for classification in general (let's now not say only in Shogun)16:45
@HeikoSiglesiasg: I think it follows sklearns basic implementation now?16:45
@iglesiasgis the number of bags / weak classifiers / trees selected by the algo16:45
@iglesiasgor is something normally chosen with xval?16:45
@HeikoSnumber of trees is set to "large"16:48
@HeikoSsince they usually dont overfit16:48
@HeikoSnumber of bags I dont know, I think there are heuristics16:48
@HeikoSbut yeah x-validation16:48
@HeikoSbut the algorithm is very robust e.g. for the number of trees used16:48
@iglesiasgmmm I think in at least CRandomForest #trees=#bags16:49
@iglesiasgTingMiao: what is the subset size you mentioned in the e-mail?16:51
TingMiao52 ^ 0.516:52
olinguyenHeikoS: is there a way to get probability scores/confidence values from RandomForest in shogun?16:54
@HeikoSolinguyen: RF are quite bad in their confidence estimates16:54
olinguyenwhat's the better way to compute the auROC for RandomForest?16:54
@HeikoSolinguyen: I dont know, but if they did, I wouldnt trust them ;)16:54
@iglesiasgTingMiao: so, I think it is worth to experiment a bit with these two parameters to get a feeling how they impact the performance16:55
@HeikoSolinguyen: check the "get_values" of the labels produced by the RF16:55
@HeikoSthey dont have scores?16:55
olinguyenthey seem to be null after .apply16:56
@iglesiasgTingMiao: using the extra validation set should anyway help a bit with the overfitting16:56
olinguyenbut the get_labels have values16:56
olinguyenHeikoS: https://gist.github.com/olinguyen/ade8e2bd0899b2ec9bee00d03fbb517216:57
olinguyenreproduced here16:57
@HeikoSolinguyen: then I guess we currently dont output that16:57
@iglesiasgTingMiao: so, my suggestion is first fix one of those two (#bags, subset size) parameters and make a small sweep in the other parameter using the x-val framework.16:58
@HeikoSolinguyen: then you need to do accuracy16:58
@HeikoSolinguyen: Ill open an issue16:58
@HeikoSolinguyen: maybe another voting strategy allows it?17:00
@HeikoSIm just checking17:00
olinguyeni'll check17:00
@iglesiasgTingMiao: get a feeling of how each of them affects the performance. And then we can make x-val choosing both or so17:00
@HeikoSolinguyen: you could try the "MeanRule" and see what happens (but also make sure you understand what that changes ;) )17:01
@iglesiasgTingMiao: the goal here is to reduce the gap between test and train accuracy17:01
@iglesiasgTingMiao: you could include this type of an analysis in an upcoming blog post too17:02
@iglesiasgTingMiao: Does this all make sense to you?17:02
@sukeyIssue #3850 "Random forest probability output " karlnapf added label: "development tasks" - https://github.com/shogun-toolbox/shogun/issues/385017:02
@sukeyIssue #3850 "Random forest probability output " karlnapf added label: "entrance" - https://github.com/shogun-toolbox/shogun/issues/385017:02
@sukeyIssue #3850 "Random forest probability output " opened by karlnapf - https://github.com/shogun-toolbox/shogun/issues/385017:02
TingMiaoI made plot of accuracy when num_of_tree or subset_size changing https://usercontent.irccloud-cdn.com/file/kFruerWE/%E5%B1%8F%E5%B9%95%E5%BF%AB%E7%85%A7%202017-06-16%20%E4%B8%8B%E5%8D%8810.59.34.png https://usercontent.irccloud-cdn.com/file/t4I7JQ86/%E5%B1%8F%E5%B9%95%E5%BF%AB%E7%85%A7%202017-06-16%20%E4%B8%8B%E5%8D%8810.55.00.png17:03
TingMiaoyes I will try17:03
TingMiaothese images are produced in my previous experiment17:04
@iglesiasgTingMiao: give it a shot using the validation set, that might help a bit17:05
@HeikoSolinguyen: check the last two issues I opened17:05
@sukeyIssue #3851 "Implement a probability calibration scheme" karlnapf added label: "entrance" - https://github.com/shogun-toolbox/shogun/issues/385117:05
@sukeyIssue #3851 "Implement a probability calibration scheme" opened by karlnapf - https://github.com/shogun-toolbox/shogun/issues/385117:05
@sukeyIssue #3851 "Implement a probability calibration scheme" karlnapf added label: "development tasks" - https://github.com/shogun-toolbox/shogun/issues/385117:05
TingMiaogot it17:05
@HeikoSolinguyen: we should at some point look at some of those, and maybe plan for you to implement the most important ones, so you can get some C++ into GSoC :)17:05
@iglesiasgTingMiao: cool cool17:06
@HeikoSolinguyen: depends a bit on you and what you are most interested in, but I would be excited about improving Shogun in an application driven way17:06
@HeikoSmicmn: jo17:17
@HeikoSnice blog post!17:17
@HeikoSTingMiao, olinguyen, micmn, mikeling, did Gina get back to you guys about the numfocus feature?17:17
micmnHeikoS: thx!17:18
@sukeyIssue #3849 "Pickle on a BinaryLabels object does not save the values parameter " karlnapf added label: "entrance" - https://github.com/shogun-toolbox/shogun/issues/384917:21
@sukeyIssue #3849 "Pickle on a BinaryLabels object does not save the values parameter " karlnapf added label: "BUG" - https://github.com/shogun-toolbox/shogun/issues/384917:21
@HeikoSmicmn: ok stay tuned17:21
@HeikoSmicmn: re the un-templated linalg stuff17:21
@HeikoSmicmn: I think what we should do is to start a feature branch where we play that thing through for a simple example case17:21
@HeikoSmicmn: like linalg::dot(Matrix m, Vector v)17:22
@HeikoSor even better17:22
@HeikoSlinalg::dot(Matrix, Matrix)17:22
@HeikoSso we just need to draft one17:22
micmnyeah :P17:22
@HeikoSand then think how the linalg::dot would look like (i.e. how it determines the type to call dot_impl(SGMatrix<T>, SGMatrix<T>)17:23
@HeikoSmake it work, think a bit about scalability17:23
@HeikoSand then iterate17:23
@HeikoSand then add it to the current interface17:24
@HeikoSso that we can start using it, but also still have the old methods (so we can continuously move to the new system)17:24
@HeikoSmicmn: do you want to take a lead on this? Or are you currently busy with another thing?17:24
@HeikoSmicmn: I can start a feature branch that you can send PRs against17:24
@HeikoSOXPHOS might also be able to get involved with this stuff, she wrote most of the linalg17:25
micmnit's fine with me17:25
@HeikoSolinguyen: could you start your doc soon and put the github issues that you and me created in there?17:26
@HeikoSmicmn: ok17:26
@HeikoSmicmn: so let me create that branch17:26
olinguyenHeikoS: will do17:26
@HeikoSmicmn: keep in mind this is drafting only, so doesnt have to be super clean, nor the most beautiful, it is really about the concept17:26
@HeikoSmicmn: so rather update often and discuss often17:26
micmnspeaking of the Matrix class17:27
@sukeyNew Commit "Merge pull request #3846 from micmn/feature/linalg-refactor17:27
@sukeySplit Eigen3's linalg backend into header and implementation." to shogun-toolbox/shogun by karlnapf: https://github.com/shogun-toolbox/shogun/commit/3b35df370192d734e19c90305e4b69e8142de31b17:27
@sukeyNew branch feature/linalg_untemplated created on shogun-toolbox/shogun17:27
@HeikoSmicmn: here is your branch17:28
micmnawesome :D17:28
@HeikoSmicmn: about the Matrix class?17:28
micmnhow do we memorize the data? I mean just a void* pointer?17:29
mikelingHeikoS: hey17:29
micmnand cast it as needed?17:29
@HeikoSmicmn: good question :D17:29
@HeikoSmicmn: void pointer would be the bare bones approach, yes17:29
mikelingHad you see the email?17:29
@HeikoSmikeling: was just about to answer17:29
mikelingHeikoS: Oh, thank you17:30
@HeikoSmicmn: I think maybe something like the CSGObject::add_vector does, it stores the void pointer, and enums for the type17:30
@HeikoSmicmn: for runtime type information, we enum is the bare bones approach17:30
@HeikoSnot sure whether c++11 can do something for us there?17:30
@HeikoSsome reflection?17:30
@HeikoSmikeling: I am not sure I get your question ;)17:31
micmnnot sure yesterday I was doing some reasearch into stuff like type erasure but didn't find anything useful for now17:31
@HeikoSmicmn: first draft could use enums17:32
@HeikoSmicmn: which will result in more MACRO HACKS (tm)17:32
@HeikoSmicmn: so there might be another option17:32
@HeikoSwhich is registering function pointers17:33
@HeikoSnot totally sure if/how that would work17:33
@HeikoSbut here is the idea in a nutshell:17:33
@HeikoS1. algorithm tells linalg that it will call dot 1000 times with float6417:33
mikelingHeikoS: Actually I and wiking are talking about which length should been used for the second parameter of add_vector17:33
@HeikoS2. linalg internally registers the corresponding (typed float64) function for dot17:34
mikelingyou know, we have two length in DynamicArray17:34
mikelingone is the number of elements17:34
@HeikoS3. the dot call from the algorithm than is just forward to that registered function17:34
@HeikoSmicmn: see what I mean?17:34
@HeikoSmicmn: you kind of have a global state of linalg that remembers the word-length, and all subsequent calls will just static cast17:34
@HeikoSmicmn: but not too sure about that, I think a first draft would use enums and switches17:35
mikelingthe other is the array size, which include the number of elements been used and not used17:35
micmnmmm I'll think about it17:35
@HeikoSmikeling: yes I know what you mean17:35
@HeikoSmikeling: we in fact fixed that recently17:35
@HeikoSfor DynArray17:35
@HeikoSmicmn: the thing here is that the size of the allocated memory is not registered17:35
@HeikoSonly the number of used elements17:35
@HeikoSso then you have a vector of size 10, but only 5 elements are filled, then serialization only stores the 5 values to disk, and de-serializing allocated a new vector with 5 elements17:36
@HeikoSand clone in fact ignores the "allocated size" parameter17:36
@HeikoSbecause it is not registered17:36
@HeikoSmikeling: that is how it currently works17:36
mikelingHeikoS: ok, so we only need use add_vector(&(m_array.data()), &num_elements)?17:37
@HeikoSyep that is OK17:37
@HeikoSloading into the std::vector might be more tricky17:37
@HeikoSsince de-serialization currently works like this17:37
@HeikoS1. allocate memory of appropriate size (as read from file)17:38
@HeikoS2. load elements one-by one17:38
@HeikoS3. update the pointer in the class17:38
@HeikoSmikeling: does that help?17:38
mikelingHeikoS: mmm,  yeah. But I found we have a m_parameter in the m_parameters like here https://pastebin.mozilla.org/902484517:40
mikelingline 1917:40
mikelingit's a void pointer17:41
mikelingbut I don't know where it been init? and which parameter it will been pointer to17:42
mikelingHeikoS: mmm, so here is my issue actually17:42
@HeikoSmikeling: I dont understand what you are asking17:43
@HeikoSmikeling: can you explain again?17:43
mikelingI replace DynArry with std::vector, and register things like https://pastebin.mozilla.org/902492217:44
mikelingsure, I'm typing17:44
mikelingbut I got https://pastebin.mozilla.org/902484617:45
mikelingit seems like the m_parameter point to somewhere haven't been allocated17:46
@HeikoSwhen does it happen?17:47
@HeikoSwhich test?17:49
@HeikoSwhich exact case?17:49
mikelingso it actually error out immediately17:50
mikelingHeikoS: Do you want me update the pr so you can check the issue locally?17:53
@HeikoSmikeling: can you run only that test, and then put the full valgrind output somewhere, along with the source code of your changed SG_ADD ?17:53
mikelingHeikoS: Oh, for now I can't run valgrind locally, due to I'm using osx. Sorry17:55
mikelingI could push it to remote and fetch it to my server(I got a remote server running Ubuntu 16.04 with valgrind)17:56
mikelingand then build and debug it17:56
@sukeyPull Request #3824 "Replace DynArray with std::vector"  synchronized by MikeLing - https://github.com/shogun-toolbox/shogun/pull/382417:58
shogun-buildbotbuild #754 of cookbook - PR is complete: Failure [failed shell]  Build details are at http://buildbot.shogun-toolbox.org/builders/cookbook%20-%20PR/builds/754  blamelist: MikeLing18:00
mikelingHeikoS: I'm building it :)18:03
-!- iglesiasg [~iglesiasg@] has quit [Quit: leaving]18:04
@HeikoSmikeling: whatever works18:06
@HeikoSI just want to see the valgrind output18:06
@HeikoSand the code you changed18:06
@HeikoSmicmn: just had another thought18:08
@HeikoSmicmn: I think we should actually keep both the templated and the untemplated linalg interface for now (as said before)18:09
@HeikoSmicmn: so the best thing was if we could just generate the method for the untemplated ones using a similar mechanism as for the templated ones18:09
@HeikoSmicmn: so that we only have to "add" the method ones, and then it is available for all SGMatrix<T> and Matrix18:10
@HeikoSmicmn: was just going through all your linalg adds btw, love them, finally we see something moving there18:12
micmnglad to hear that :)18:13
@HeikoSmikeling:  and?18:24
mikeling23% :(18:24
-!- travis-ci [~travis-ci@ec2-54-162-244-111.compute-1.amazonaws.com] has joined #shogun18:24
travis-ciit'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/24371090318:24
-!- travis-ci [~travis-ci@ec2-54-162-244-111.compute-1.amazonaws.com] has left #shogun []18:24
@HeikoSmikeling: are you using ccache?18:24
@HeikoSwith a large cache size?18:25
@HeikoSbecause that makes compiling very fast18:25
@HeikoSsee DEVELOPING.md18:25
mikelingyeah, I do install CCache18:25
mikelingbut I don't know how to check the config18:25
@HeikoS ccache -s18:33
@HeikoSmikeling: google is your friend :D18:33
mikelingHeikoS: What's the time in your there? BTW18:41
mikelingI will cc you the result after things done :)18:42
@HeikoSmikeling: cool thx18:43
@HeikoSmikeling: btw just sent another email18:43
-!- HeikoS [~heiko@host-92-0-178-129.as43234.net] has quit [Quit: Leaving.]19:34
-!- HeikoS [~heiko@host-92-0-178-129.as43234.net] has joined #shogun19:35
-!- mode/#shogun [+o HeikoS] by ChanServ19:35
-!- HeikoS [~heiko@host-92-0-178-129.as43234.net] has quit [Client Quit]19:35
@sukeyPull Request #3843 "Add linalg methods needed by FisherLDA and KernelPCA (CPU-only)"  synchronized by micmn - https://github.com/shogun-toolbox/shogun/pull/384320:01
shogun-buildbotbuild #755 of cookbook - PR is complete: Success [build successful]  Build details are at http://buildbot.shogun-toolbox.org/builders/cookbook%20-%20PR/builds/75520:02
@sukeyPull Request #3843 "Add linalg methods needed by FisherLDA and KernelPCA (CPU-only)" - https://github.com/shogun-toolbox/shogun/pull/384320:03
shogitter(iteachmachines) Hey why am i seeing only @bazdmeg  your messages?20:37
@sukeyPull Request #3843 "Add linalg methods needed by FisherLDA and KernelPCA (CPU-only)"  synchronized by micmn - https://github.com/shogun-toolbox/shogun/pull/384320:47
-!- HeikoS [~heiko@] has joined #shogun21:39
-!- mode/#shogun [+o HeikoS] by ChanServ21:39
@HeikoSmikeling: updates?21:39
@HeikoSmicmn:  there?21:42
-!- olinguyen [81615ad9@gateway/web/freenode/ip.] has quit [Quit: Page closed]21:54
@HeikoSrcurtin: how are things going gsoc wise?21:54
rcurtinHeikoS: I'd say things are going quite well, but maybe not so well for my free time! :)21:58
@HeikoScool! good to hear21:58
rcurtinlots of nice work on the benchmarking system21:58
rcurtinwe are running the shogun benchmarks now (I think 5.0, but we'll have to upgrade and run again)21:58
rcurtinonce those are done, we can deploy the new database to the website, and then we can see all the things that the benchmarking scripts did wrong :)21:58
-!- HeikoS [~heiko@] has quit [Read error: Connection reset by peer]22:05
-!- HeikoS [~heiko@] has joined #shogun22:08
-!- mode/#shogun [+o HeikoS] by ChanServ22:08
-!- HeikoS [~heiko@] has quit [Quit: Leaving.]22:25
-!- HeikoS [~heiko@] has joined #shogun22:25
-!- mode/#shogun [+o HeikoS] by ChanServ22:25
@HeikoSlisitsyn: you around?22:30
-!- mikeling [uid89706@gateway/web/irccloud.com/x-rovmvjrxlvkouswc] has quit [Quit: Connection closed for inactivity]22:46
-!- TingMiao [uid229534@gateway/web/irccloud.com/x-xbcsnzdqkuwblyri] has quit [Quit: Connection closed for inactivity]23:30
-!- HeikoS [~heiko@] has quit [Ping timeout: 255 seconds]23:31
--- Log closed Sat Jun 17 00:00:16 2017