--- Log opened Thu May 31 00:00:29 2018 | ||
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4317 synchronized by vinx13 | 04:56 |
---|---|---|
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4316 synchronized by vinx13 | 04:56 |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4316 closed by vinx13 | 04:58 |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4316 synchronized by vinx13 | 04:58 |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4316 synchronized by vinx13 | 05:00 |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4316 reopened by vinx13 | 05:00 |
-!- witness_ [uid10044@gateway/web/irccloud.com/x-koiqjutynwflbjuq] has quit [Quit: Connection closed for inactivity] | 06:13 | |
-!- shogun-buildbot [~shogun-bu@7nn.de] has quit [Ping timeout: 240 seconds] | 06:21 | |
-!- shogun-buildbot [~shogun-bu@7nn.de] has joined #shogun | 06:24 | |
-!- witness_ [uid10044@gateway/web/irccloud.com/x-dltrrsqnhcrguliv] has joined #shogun | 06:46 | |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4316 merged by vigsterkr | 07:46 |
@sukey | [https://github.com/shogun-toolbox/shogun] vigsterkr pushed 2 commits: | 07:46 |
@sukey | https://github.com/shogun-toolbox/shogun/commit/74d8dba0ab36d1d4b16d98e38d0df63f819391d1 | 07:46 |
@sukey | https://github.com/shogun-toolbox/shogun/commit/1cc9b920b472ac3207afdf4d203b98c46f3177fe | 07:46 |
-!- travis-ci [~travis-ci@ec2-54-163-68-221.compute-1.amazonaws.com] has joined #shogun | 08:47 | |
travis-ci | it's Wuwei Lin'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/386030115 | 08:47 |
-!- travis-ci [~travis-ci@ec2-54-163-68-221.compute-1.amazonaws.com] has left #shogun [] | 08:47 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has joined #shogun | 08:53 | |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4318 opened by vinx13 | 08:54 |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client] | 09:05 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has joined #shogun | 09:05 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has quit [Client Quit] | 09:07 | |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4318 synchronized by vinx13 | 09:07 |
-!- sukey [~nodebot@ks312251.kimsufi.com] has quit [Read error: Connection reset by peer] | 09:08 | |
-!- sukey [~nodebot@ks312251.kimsufi.com] has joined #shogun | 09:09 | |
-!- mode/#shogun [+o sukey] by ChanServ | 09:10 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has joined #shogun | 09:30 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has quit [Client Quit] | 09:33 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has joined #shogun | 09:35 | |
wuwei | wiking: any idea for data locking for refactor ? | 09:50 |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client] | 10:00 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has joined #shogun | 10:01 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has quit [Ping timeout: 256 seconds] | 10:40 | |
-!- HeikoS [~heiko@host81-153-166-11.range81-153.btcentralplus.com] has joined #shogun | 10:50 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 10:50 | |
-!- HeikoS [~heiko@host81-153-166-11.range81-153.btcentralplus.com] has quit [Ping timeout: 268 seconds] | 11:00 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has joined #shogun | 11:05 | |
-!- HeikoS [~heiko@host81-153-166-11.range81-153.btcentralplus.com] has joined #shogun | 11:11 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 11:11 | |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4318 synchronized by vinx13 | 11:22 |
-!- HeikoS [~heiko@host81-153-166-11.range81-153.btcentralplus.com] has quit [Ping timeout: 248 seconds] | 11:40 | |
-!- HeikoS [~heiko@82-132-214-240.dab.02.net] has joined #shogun | 11:53 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 11:53 | |
@HeikoS | lisitsyn: around? | 12:09 |
@HeikoS | shubham808: hey | 12:20 |
@HeikoS | how are things going? | 12:20 |
shubham808 | i was working on the perceptron thing | 12:21 |
@HeikoS | ah you mean making the state change even when it is not initialised? | 12:22 |
@HeikoS | cool, should be a quick fix no? | 12:22 |
shubham808 | yes yes | 12:22 |
shubham808 | actually no | 12:22 |
shubham808 | i am debugging what went wrong | 12:22 |
shubham808 | the assertion at LinearMachine line 89 failed | 12:23 |
shubham808 | it checks for correct diemnsions of weight vector | 12:23 |
@HeikoS | maybe it wasnt initialised correctly | 12:25 |
shubham808 | probably... | 12:28 |
-!- HeikoS1 [~heiko@82-132-227-229.dab.02.net] has joined #shogun | 12:30 | |
shubham808 | i will dig deeper then get back to u | 12:30 |
-!- HeikoS [~heiko@82-132-214-240.dab.02.net] has quit [Ping timeout: 240 seconds] | 12:31 | |
-!- HeikoS1 [~heiko@82-132-227-229.dab.02.net] has quit [Ping timeout: 260 seconds] | 12:46 | |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4319 opened by vinx13 | 13:02 |
-!- HeikoS [~heiko@host81-153-166-11.range81-153.btcentralplus.com] has joined #shogun | 13:03 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 13:03 | |
shubham808 | HeikoS hey | 13:10 |
shubham808 | I have a question about vectors | 13:11 |
shubham808 | around ? | 13:21 |
@HeikoS | shubham808: sure shoot | 13:32 |
shubham808 | https://pastebin.com/dD4nPbwL | 13:34 |
shubham808 | all 3 are not same | 13:34 |
shubham808 | w = get_w() and m_w are not pointing to same location are they ? | 13:36 |
wuwei | heiko: could u restart this https://travis-ci.org/shogun-toolbox/shogun/jobs/386093267 | 13:37 |
@wiking | wuwei, done | 13:42 |
wuwei | thanks! | 13:43 |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client] | 14:06 | |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4318 synchronized by vinx13 | 14:21 |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has joined #shogun | 14:31 | |
@HeikoS | shubham808: I guess you could also use set_w | 14:38 |
@HeikoS | if you wanna change ity | 14:38 |
@HeikoS | i.e. create the initialisation w lolcally | 14:38 |
@HeikoS | and then set it | 14:39 |
shubham808 | so | 14:45 |
shubham808 | we could go like | 14:45 |
shubham808 | SGVector<float64_t> w;set_w(w); | 14:45 |
shubham808 | then when i w.resize_vector(num_feat) the m_w.vector should be getting resized ? | 14:46 |
wuwei | hi, a quick question about c++ | 15:15 |
wuwei | what's the suggested way to achieve something like: | 15:16 |
wuwei | for (index, iter: DotIterator(feats)) | 15:16 |
@HeikoS | shubham808: no | 15:21 |
@HeikoS | you would do | 15:21 |
@HeikoS | w=SGVector(...) | 15:21 |
@HeikoS | set_w(w) | 15:21 |
-!- HeikoS [~heiko@host81-153-166-11.range81-153.btcentralplus.com] has quit [Ping timeout: 256 seconds] | 15:26 | |
@wiking | wuwei, index is for? | 15:35 |
@wiking | i mean i know... seen this pattern everywhere | 15:35 |
wuwei | index is for feature vector | 15:36 |
wuwei | we have iterator, but index is still used | 15:36 |
@wiking | yeah | 15:36 |
@wiking | one super annoying way | 15:37 |
@wiking | you expose the idx via iterator | 15:37 |
@wiking | :D | 15:37 |
@wiking | wuwei, https://faouellet.github.io/enumerate/ | 15:40 |
@wiking | https://infektor.net/posts/2017-03-31-range-based-enumerate.html#range-based-enumerate | 15:41 |
@wiking | but that's c++17 | 15:41 |
wuwei | yeah :) | 15:41 |
wuwei | that would be good, when we have c++17 | 15:42 |
@wiking | https://skebanga.github.io/structured-bindings/ | 15:42 |
@wiking | :> | 15:42 |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4320 opened by shubham808 | 15:50 |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4320 | 15:51 |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4287 synchronized by shubham808 | 16:01 |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client] | 16:09 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has joined #shogun | 16:12 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has quit [Client Quit] | 16:14 | |
-!- HeikoS [~heiko@eduroam-int-pat-8-0.ucl.ac.uk] has joined #shogun | 16:14 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 16:14 | |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4287 synchronized by shubham808 | 16:18 |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has joined #shogun | 16:19 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has quit [Client Quit] | 16:20 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has joined #shogun | 16:29 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has quit [Client Quit] | 16:30 | |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4320 synchronized by shubham808 | 16:34 |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has joined #shogun | 16:34 | |
shubham808 | HeikoS hey | 16:34 |
@HeikoS | shubham808: hi | 16:34 |
shubham808 | about the set_w | 16:34 |
shubham808 | we want inside the if right? | 16:34 |
shubham808 | because if we want to initialize we will need to set it | 16:35 |
@HeikoS | you only want to overwrite it if m_initialise is true | 16:35 |
shubham808 | but we can just get the user initialized one in the else ? | 16:35 |
@HeikoS | otherwise, you just want to get it | 16:35 |
@HeikoS | i.e. you dont wanna change it if initilize=false | 16:35 |
shubham808 | i see i see | 16:35 |
@HeikoS | so the user can initialize themselves | 16:35 |
shubham808 | i made an update | 16:39 |
shubham808 | now what it says is | 16:39 |
shubham808 | if(inititialize_hyperplane)set_w(w); | 16:39 |
shubham808 | else w = get_w(); //initialize is false so we can use the user provided | 16:40 |
shubham808 | look okay ? | 16:43 |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client] | 16:45 | |
-!- mikeling [uid89706@gateway/web/irccloud.com/x-ntbuqnkyepoqwknl] has joined #shogun | 16:47 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has joined #shogun | 16:48 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has quit [Client Quit] | 16:50 | |
-!- HeikoS [~heiko@eduroam-int-pat-8-0.ucl.ac.uk] has quit [Ping timeout: 244 seconds] | 16:51 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has joined #shogun | 16:51 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client] | 16:57 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has joined #shogun | 16:57 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has quit [Client Quit] | 16:57 | |
-!- HeikoS [~heiko@untrust-out.swc.ucl.ac.uk] has joined #shogun | 17:07 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 17:07 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has joined #shogun | 17:50 | |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4320 synchronized by shubham808 | 17:54 |
shubham808 | HeikoS hey | 17:56 |
@HeikoS | shubham808: hi | 17:56 |
shubham808 | i wrote a test | 17:56 |
@HeikoS | cool | 17:57 |
shubham808 | its a bit copy paste atm | 17:57 |
@HeikoS | fixtures are your friend there | 17:57 |
@HeikoS | and moving shared code into functions | 17:57 |
@HeikoS | but show :) | 17:57 |
shubham808 | pushed | 17:57 |
@HeikoS | reading it | 17:57 |
@HeikoS | ah ok | 17:57 |
@HeikoS | so the thing i had in mind is | 17:57 |
@HeikoS | two tests | 17:58 |
@HeikoS | 1. user wants to specify a custom initialization, which is used as a start from training | 17:58 |
@HeikoS | 2. user doesnt care | 17:58 |
@HeikoS | 2. is something that we probably test already | 17:58 |
@HeikoS | so you need to write a test for 1 | 17:58 |
shubham808 | yep | 17:58 |
shubham808 | so i initialized the weights to 1 | 17:59 |
@HeikoS | ok cool | 17:59 |
@HeikoS | so how do you know it worked? | 17:59 |
@HeikoS | the test doesnt really check that | 17:59 |
@HeikoS | I would maybe initialise it to something that you know is a solution (maybe from a previous run) | 17:59 |
@HeikoS | and then run one iteration | 17:59 |
shubham808 | i see | 18:00 |
@HeikoS | and then make sure it didnt change (i.e. it converged immediately)? | 18:00 |
@HeikoS | something like this | 18:00 |
@HeikoS | quite minimal | 18:00 |
@HeikoS | evaluating accuracy is a bit of an overkill | 18:00 |
@HeikoS | all you want is to ensure that the algorithm is "hot started" | 18:00 |
@HeikoS | and also | 18:00 |
shubham808 | yeah thats true | 18:00 |
@HeikoS | so maybe | 18:01 |
@HeikoS | train model to convergence | 18:01 |
@HeikoS | then copy that w | 18:01 |
@HeikoS | create new model | 18:01 |
@HeikoS | set that w | 18:01 |
@HeikoS | train | 18:01 |
@HeikoS | assert that it didnt change (i.e. immediate convergence) | 18:01 |
shubham808 | set max_iter to 1 | 18:01 |
shubham808 | (to make sure it converges immediately) | 18:04 |
@HeikoS | no need | 18:05 |
@HeikoS | if w is already a solution, the algorithm should converge automatically right? | 18:05 |
@HeikoS | shubham808: perceptron checks if there are any classification errors, and if no, it stops | 18:05 |
@HeikoS | only if there are errors, it changes w | 18:05 |
shubham808 | righ right | 18:06 |
shubham808 | we just make sure w didnt change then | 18:06 |
shubham808 | after training | 18:06 |
@HeikoS | yep | 18:11 |
@HeikoS | it is basically a test for "hot starting" | 18:11 |
@HeikoS | we could think about generalising this to other models later, but for now let's just do it for the perceptron | 18:11 |
shubham808 | one more thing | 18:15 |
shubham808 | suppose that the custom initialization didnt work | 18:17 |
shubham808 | the assertion will still pass | 18:17 |
shubham808 | do u see ? | 18:17 |
shubham808 | because even if its not hot started it will still reach the correct weights vector on convergence | 18:20 |
shubham808 | right ? | 18:20 |
@HeikoS | hey | 18:24 |
@HeikoS | true | 18:25 |
@HeikoS | in that case, I think your idea of restricting the number of iterations to small will help | 18:25 |
@HeikoS | shubham808: good catch! : | 18:25 |
shubham808 | yeah but | 18:25 |
shubham808 | u see perceptron only generates a warning if the number of max iters exceeds | 18:26 |
@HeikoS | thats fine | 18:26 |
shubham808 | how do we unit test that ? | 18:26 |
@HeikoS | you just set it to 1 | 18:26 |
@HeikoS | max iter = 1 | 18:26 |
@HeikoS | then you call train | 18:26 |
@HeikoS | and then afterwards you check that w wasnt changed (i.e. it converged with less than one iteration) | 18:26 |
@HeikoS | might need to set it to 2 | 18:26 |
shubham808 | right right | 18:27 |
shubham808 | yep got it | 18:27 |
shubham808 | iters = 0 in begining so max_iter =1 should be fine | 18:28 |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4320 synchronized by shubham808 | 18:31 |
@wiking | wuwei, r u here by any chance? | 18:51 |
wuwei | wiking: yes | 18:52 |
@wiking | wuwei, we forgot one thing :) | 18:52 |
@wiking | in case of transformers | 18:52 |
@wiking | inverse transform :) | 18:53 |
wuwei | u mean we should implemnet inverse transform? | 18:54 |
@wiking | interface as well as implement | 18:54 |
wuwei | well, it's not always easy to implement inverse | 18:57 |
@wiking | sure | 18:57 |
wuwei | like sorting | 18:57 |
@wiking | but currently we dont even have an interface for it :) | 18:57 |
@wiking | meaning even if it's trivial | 18:58 |
@wiking | there's no standard way of doing it | 18:58 |
@wiking | (say scaling) | 18:58 |
wuwei | yeah | 18:59 |
@wiking | but it'd be good to have | 18:59 |
@wiking | what's your current task at hand? | 18:59 |
@wiking | pipeline? | 18:59 |
wuwei | machine refactor | 19:00 |
@wiking | fit/predict? | 19:00 |
wuwei | yeah | 19:00 |
@wiking | https://github.com/shogun-toolbox/shogun/pull/4318 can be merged right? | 19:01 |
wuwei | yeah | 19:01 |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4318 merged by vigsterkr | 19:01 |
@sukey | [https://github.com/shogun-toolbox/shogun] vigsterkr pushed 5 commits: | 19:01 |
@sukey | https://github.com/shogun-toolbox/shogun/commit/246819f932cd982a9ea30e7fcf4e24157a573748 | 19:01 |
@sukey | https://github.com/shogun-toolbox/shogun/commit/fae9e76fa4eeae3dc9312581eb499962969aa2da | 19:02 |
@sukey | https://github.com/shogun-toolbox/shogun/commit/f42efa2a142da4b685f88d386f45133ed45a3ac7 | 19:02 |
@sukey | https://github.com/shogun-toolbox/shogun/commit/35089ea8e81a666edc677b6af72f6dcc5b4a1543 | 19:02 |
@sukey | https://github.com/shogun-toolbox/shogun/commit/20f7808dc649e98bf52f84d9fc2eb1ea3d9c2e02 | 19:02 |
-shogun-buildbot:#shogun- Build trusty - libshogun - viennacl #499 is complete: Failure [failed update shogun (failure)] - http://buildbot.shogun-toolbox.org:8080/#builders/6/builds/499 | 19:02 | |
-shogun-buildbot:#shogun- Build deb1 - libshogun #470 is complete: Failure [failed update shogun (failure)] - http://buildbot.shogun-toolbox.org:8080/#builders/10/builds/470 | 19:02 | |
-shogun-buildbot:#shogun- Build trusty - libshogun #404 is complete: Failure [failed update shogun (failure)] - http://buildbot.shogun-toolbox.org:8080/#builders/33/builds/404 | 19:02 | |
-shogun-buildbot:#shogun- Build deb3 - interfaces #359 is complete: Failure [failed update shogun (failure)] - http://buildbot.shogun-toolbox.org:8080/#builders/37/builds/359 | 19:03 | |
-shogun-buildbot:#shogun- Build FC23 - libshogun #504 is complete: Failure [failed update shogun (failure)] - http://buildbot.shogun-toolbox.org:8080/#builders/27/builds/504 | 19:03 | |
@wiking | :O | 19:04 |
@wiking | wtf | 19:04 |
@wiking | :D | 19:04 |
@wiking | shogun data? | 19:04 |
-shogun-buildbot:#shogun- Build zesty - libshogun #492 is complete: Failure [failed update shogun (failure)] - http://buildbot.shogun-toolbox.org:8080/#builders/13/builds/492 | 19:04 | |
-shogun-buildbot:#shogun- Build yakkety - libshogun #496 is complete: Failure [failed update shogun (failure)] - http://buildbot.shogun-toolbox.org:8080/#builders/1/builds/496 | 19:04 | |
-shogun-buildbot:#shogun- Build xenial - libshogun #493 is complete: Failure [failed update shogun (failure)] - http://buildbot.shogun-toolbox.org:8080/#builders/16/builds/493 | 19:04 | |
wuwei | need merge data pr:) | 19:04 |
wuwei | yeah | 19:05 |
@wiking | done | 19:05 |
@wiking | ok next push to develop will correct this :D | 19:06 |
-!- Farouk [9cc2876f@gateway/web/freenode/ip.156.194.135.111] has joined #shogun | 19:12 | |
@wiking | Farouk, y0 | 19:15 |
Farouk | wiking, hey | 19:15 |
Farouk | wiking, just looked at the feedback. I am assuming that the .sg file is fine now so I can send a pr to shogun-data correct? | 19:16 |
-!- HeikoS [~heiko@untrust-out.swc.ucl.ac.uk] has quit [Ping timeout: 256 seconds] | 19:17 | |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4320 synchronized by shubham808 | 19:28 |
shubham808 | wiking: hey | 19:32 |
-!- witness_ [uid10044@gateway/web/irccloud.com/x-dltrrsqnhcrguliv] has quit [Quit: Connection closed for inactivity] | 19:33 | |
shubham808 | had a question about meta examples :) | 19:34 |
@wiking | hi | 19:39 |
@wiking | sorry folks | 19:39 |
@wiking | i need about 10m | 19:39 |
@wiking | and i;'ll get back here | 19:39 |
@wiking | brb | 19:39 |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client] | 19:47 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has joined #shogun | 19:47 | |
shubham808 | wiking: cool :) | 20:03 |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client] | 20:19 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has joined #shogun | 20:19 | |
-!- Farouk [9cc2876f@gateway/web/freenode/ip.156.194.135.111] has quit [Ping timeout: 260 seconds] | 20:38 | |
@wiking | ok pepes i'm here | 20:40 |
shubham808 | hey hey | 20:52 |
@wiking | shubham808, ok shoot what was the question | 20:53 |
shubham808 | wiking: about meta examples | 20:53 |
shubham808 | i was planning on porting FeatureBlockLogisticRegression to meta | 20:53 |
@wiking | ic | 20:53 |
shubham808 | https://github.com/shogun-toolbox/shogun/blob/20f7808dc649e98bf52f84d9fc2eb1ea3d9c2e02/examples/undocumented/python/classifier_featureblock_logistic_regression.py#L22 | 20:54 |
shubham808 | got stuck with hstack here | 20:54 |
@wiking | ohoooo | 20:56 |
@wiking | i mean that's just a trick in our case | 20:56 |
@wiking | how we can generate easily the 2 blocks | 20:57 |
-!- mikeling [uid89706@gateway/web/irccloud.com/x-ntbuqnkyepoqwknl] has quit [Quit: Connection closed for inactivity] | 20:57 | |
shubham808 | i see | 20:57 |
shubham808 | how do we translate this to .sg | 20:57 |
@wiking | yeah actually that should not be translated | 20:58 |
@wiking | as obviously this (hstack) only available in py | 20:58 |
shubham808 | yep yep | 20:58 |
@wiking | i'm just trying to quickly figure out | 20:58 |
shubham808 | i meant a work around maybe ? | 20:58 |
@wiking | how we can do this with a trick | 20:58 |
@wiking | but worst case we add a new data :D | 20:58 |
@wiking | that has this :P | 20:58 |
@wiking | that way it's just csvfile etc | 20:59 |
shubham808 | yeah doable | 20:59 |
shubham808 | (worst case) | 20:59 |
@wiking | i mean to be honest | 20:59 |
@wiking | we just need something where we can divide the input data into two blocks | 21:00 |
@wiking | fm_train_real.dat is usually 2 features | 21:00 |
@wiking | i wonder if we would divide into 2 blocks | 21:00 |
@wiking | what happens :) | 21:00 |
shubham808 | :S not sure i get it | 21:04 |
@wiking | ok so | 21:06 |
@wiking | once you read in the toy data | 21:06 |
@wiking | you can just do the same as in case of the hstack | 21:06 |
@wiking | define the 2 blocks | 21:06 |
@wiking | one for each feature | 21:06 |
@wiking | so simply read in the default toy dataset for integration tests | 21:07 |
@wiking | and define blcoks like | 21:07 |
@wiking | 0, n_samples/2 | 21:07 |
@wiking | block1((0,n_features/2) | 21:07 |
@wiking | block2(n_features/2, n_features) | 21:07 |
@wiking | and see how it goes | 21:07 |
@wiking | w/o the stacking | 21:07 |
shubham808 | so essentially dividing the toy data in half? | 21:09 |
@wiking | yes | 21:09 |
shubham808 | okay | 21:10 |
shubham808 | might work | 21:10 |
shubham808 | yes :) | 21:18 |
shubham808 | works fine | 21:18 |
shubham808 | Thanks | 21:20 |
-!- iglesiasg [~iglesias@f119189.upc-f.chello.nl] has joined #shogun | 21:55 | |
-!- Farouk [9cc2876f@gateway/web/freenode/ip.156.194.135.111] has joined #shogun | 21:57 | |
Farouk | wiking, is there a link for the Zoom meeting? | 21:58 |
Farouk | or just the old link? | 21:58 |
@wiking | old one | 21:58 |
@wiking | ok lemme try to log in | 21:58 |
Farouk | it says its waiting for the host to start the meeting | 22:01 |
iglesiasg | yep | 22:02 |
iglesiasg | for me too | 22:02 |
@wiking | imin | 22:04 |
@wiking | try now | 22:04 |
@wiking | sorry | 22:04 |
@wiking | Farouk, i guess you cannot hear us? | 22:07 |
@wiking | because certainly we cannot hear u | 22:07 |
Farouk | Yea can't hear you | 22:07 |
Farouk | ill restart windows and call you | 22:08 |
Farouk | sorry for the delay | 22:08 |
@wiking | kk | 22:08 |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client] | 22:09 | |
-!- Farouk [9cc2876f@gateway/web/freenode/ip.156.194.135.111] has quit [Ping timeout: 260 seconds] | 22:12 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has joined #shogun | 22:16 | |
iglesiasg | wiking, https://github.com/shogun-toolbox/shogun/pull/4309/files#diff-850db6d5082e9cf6bbf0ddb8e5ad2ab3R53 | 22:16 |
-!- Farouk [9cc2876f@gateway/web/freenode/ip.156.194.135.111] has joined #shogun | 22:16 | |
iglesiasg | https://github.com/shogun-toolbox/shogun/pull/4309/files#diff-850db6d5082e9cf6bbf0ddb8e5ad2ab3R53 | 22:17 |
@sukey | [https://github.com/shogun-toolbox/shogun] Issue https://github.com/shogun-toolbox/shogun/issues/4321 opened by vigsterkr | 22:20 |
@sukey | [https://github.com/shogun-toolbox/shogun] Issue https://github.com/shogun-toolbox/shogun/issues/4321 | 22:20 |
@sukey | [https://github.com/shogun-toolbox/shogun] Issue https://github.com/shogun-toolbox/shogun/issues/4321 vigsterkr added label: "SWIG" | 22:20 |
@sukey | [https://github.com/shogun-toolbox/shogun] Issue https://github.com/shogun-toolbox/shogun/issues/4321 vigsterkr added label: "BUG" | 22:20 |
Farouk | https://pastebin.com/Yxf6sebY | 22:28 |
-!- iglesias [~iglesias@f119189.upc-f.chello.nl] has joined #shogun | 22:43 | |
-!- iglesiasg [~iglesias@f119189.upc-f.chello.nl] has quit [Ping timeout: 245 seconds] | 22:45 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client] | 23:01 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has joined #shogun | 23:10 | |
-!- shubham808 [0e8bf0fb@gateway/web/cgi-irc/kiwiirc.com/ip.14.139.240.251] has quit [Client Quit] | 23:13 | |
-!- Farouk [9cc2876f@gateway/web/freenode/ip.156.194.135.111] has quit [Ping timeout: 260 seconds] | 23:38 | |
--- Log closed Fri Jun 01 00:00:31 2018 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!