--- Log opened Sun Apr 14 00:00:28 2013 | ||
-!- gsomix [~Miranda@95.67.132.202] has quit [Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org] | 00:01 | |
-!- sumit [73f91219@gateway/web/freenode/ip.115.249.18.25] has quit [Ping timeout: 245 seconds] | 00:09 | |
-!- akshaydixi [~akshay@223.236.138.90] has joined #shogun | 00:54 | |
-!- akshaydixi [~akshay@223.236.138.90] has quit [Ping timeout: 256 seconds] | 01:16 | |
-!- akshaydixi [~akshay@115.248.50.23] has joined #shogun | 01:28 | |
-!- foulwall [~foulwall@2001:da8:215:c252:7cbf:f8a:c1e2:c5dc] has joined #shogun | 01:57 | |
-!- foulwall [~foulwall@2001:da8:215:c252:7cbf:f8a:c1e2:c5dc] has quit [Remote host closed the connection] | 01:59 | |
-!- akshaydixi [~akshay@115.248.50.23] has quit [Ping timeout: 248 seconds] | 02:41 | |
-!- akshaydixi [~akshay@115.248.50.24] has joined #shogun | 02:54 | |
-!- foulwall [~foulwall@li379-21.members.linode.com] has joined #shogun | 03:00 | |
foulwall | I've sent a pull request of shogun-web , check it, pls | 03:00 |
---|---|---|
-!- foulwall [~foulwall@li379-21.members.linode.com] has quit [Remote host closed the connection] | 03:02 | |
-!- zuhao [~wanzuhao@nusnet-184-31.dynip.nus.edu.sg] has joined #shogun | 03:11 | |
-!- akshaydixi [~akshay@115.248.50.24] has quit [Ping timeout: 248 seconds] | 03:40 | |
-!- akshaydixi [~akshay@115.248.50.23] has joined #shogun | 03:54 | |
foulwall | join #shogun | 04:27 |
-!- foulwall [~foulwall@li379-21.members.linode.com] has joined #shogun | 04:27 | |
-!- foulwall [~foulwall@li379-21.members.linode.com] has quit [Remote host closed the connection] | 04:34 | |
-!- foulwall [~foulwall@li379-21.members.linode.com] has joined #shogun | 04:59 | |
-!- Calvin-O [75ef1953@gateway/web/freenode/ip.117.239.25.83] has joined #shogun | 05:17 | |
-!- Calvin-O [75ef1953@gateway/web/freenode/ip.117.239.25.83] has quit [Quit: Page closed] | 05:27 | |
-!- foulwall [~foulwall@li379-21.members.linode.com] has quit [Remote host closed the connection] | 05:37 | |
-!- foulwall [~foulwall@2001:da8:215:c252:8da0:a027:9fb6:bae3] has joined #shogun | 06:52 | |
foulwall | sonney2k: that's weird | 06:53 |
foulwall | sonney2k: when i logoff the pull request disappeared. | 06:54 |
foulwall | sonney2k: when i logged in it's in there | 06:55 |
-!- zuhao [~wanzuhao@nusnet-184-31.dynip.nus.edu.sg] has quit [Quit: zuhao] | 06:58 | |
-!- hoijui [~hoijui@dslb-088-074-124-102.pools.arcor-ip.net] has joined #shogun | 07:00 | |
foulwall | sonney2k: I'll send a pull request again. | 07:02 |
-!- foulwall [~foulwall@2001:da8:215:c252:8da0:a027:9fb6:bae3] has quit [Remote host closed the connection] | 07:26 | |
-!- foulwall [~foulwall@2001:da8:215:6819:b464:bf90:2956:36e4] has joined #shogun | 07:57 | |
-!- sids_aquarius [~sidi@li400-124.members.linode.com] has joined #shogun | 08:06 | |
-!- zuhao [~wanzuhao@nusnet-184-31.dynip.nus.edu.sg] has joined #shogun | 08:35 | |
-!- sids_aquarius [~sidi@li400-124.members.linode.com] has quit [Quit: Lost terminal] | 08:38 | |
foulwall | sonney2k: I guess there is something wrong with my github account. I contacted the github for service. waiting for respond. Not only the disapearence of pull requests but also some weird mistakes happenened on my accounts. waiting for respond. | 09:44 |
-!- akshaydixi [~akshay@115.248.50.23] has quit [Ping timeout: 245 seconds] | 10:07 | |
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 10:20 | |
-!- mode/#shogun [+o iglesiasg] by ChanServ | 10:20 | |
-!- hoijui [~hoijui@dslb-088-074-124-102.pools.arcor-ip.net] has quit [Ping timeout: 276 seconds] | 10:20 | |
-!- zuhao [~wanzuhao@nusnet-184-31.dynip.nus.edu.sg] has quit [Quit: zuhao] | 10:23 | |
-!- gsomix [~Miranda@188.168.14.129] has joined #shogun | 10:33 | |
gsomix | good morning | 10:33 |
-!- foulwall [~foulwall@2001:da8:215:6819:b464:bf90:2956:36e4] has quit [Remote host closed the connection] | 10:40 | |
@iglesiasg | morning | 10:43 |
@lisitsyn | iglesiasg: morgen | 10:50 |
-!- sumit [73f91219@gateway/web/freenode/ip.115.249.18.25] has joined #shogun | 10:57 | |
-!- sids_aquarius [~sidi@li400-124.members.linode.com] has joined #shogun | 11:10 | |
@sonney2k | foulwall - now it is there :) | 11:23 |
@sonney2k | https://github.com/shogun-toolbox/shogun-web/pulls | 11:23 |
@sonney2k | iglesiasg, have a look - pretty nice... | 11:24 |
@iglesiasg | sonney2k: yes, I saw the code this morning | 11:31 |
@iglesiasg | sonney2k: do you think it is ready to be put live in the website? | 11:47 |
-!- zuhao [~wanzuhao@nusnet-184-31.dynip.nus.edu.sg] has joined #shogun | 12:08 | |
-!- foulwall [~foulwall@li379-21.members.linode.com] has joined #shogun | 12:16 | |
foulwall | Maybe I write some more demos then put them in the website? | 12:44 |
foulwall | @lisitsyn | 12:45 |
foulwall | @sonney2k | 12:45 |
@lisitsyn | foulwall: yes sure | 12:45 |
foulwall | I'll go and find | 12:46 |
@lisitsyn | foulwall: I do not think we are in rush with putting them live | 12:46 |
foulwall | I'll just work on my local repo and when I finish I'll show you. | 12:48 |
@iglesiasg | cool, looking forward to seeing new demos :) | 12:48 |
foulwall | lisitsyn: btw, is there any suggestion what I write next or I'll find on my own? | 12:49 |
@lisitsyn | foulwall: just ask if you are stucked but if you have your own ideas it is even better :) | 12:49 |
@iglesiasg | foulwall: maybe you can draw some inspiration from python_modular/graphical | 12:49 |
foulwall | iglesiasg: thank you | 12:50 |
@sonney2k | foulwall, hey there | 12:50 |
foulwall | hey! sonney2k | 12:51 |
foulwall | lisitsyn: thankyou. | 12:51 |
@sonney2k | foulwall, I don't have too much time currently (kids...) but excellent work again :) | 12:51 |
@sonney2k | foulwall, so I will give you some feedback tonight and then we can even put this live! | 12:51 |
@sonney2k | foulwall, about future stuff... | 12:52 |
@sonney2k | there are a couple of things you could do | 12:52 |
@sonney2k | one thing would be the same thing with dots but not for regression but classification | 12:52 |
foulwall | sonney2k: sorry for the github trouble. | 12:52 |
@sonney2k | foulwall, no idea what went wrong but now I see 2 pull requests | 12:52 |
foulwall | sonney2k: One has closed by me. | 12:53 |
foulwall | two requests are exactly the same. | 12:54 |
-!- sumit [73f91219@gateway/web/freenode/ip.115.249.18.25] has quit [Ping timeout: 245 seconds] | 12:54 | |
@sonney2k | foulwall, or if you want to do sth different and a bit more challenging you could also re-write the applications/ocr demo | 12:57 |
@sonney2k | foulwall, that is the one on the website here | 12:57 |
@sonney2k | http://shogun-toolbox.org/page/about/examples | 12:57 |
@sonney2k | (scroll down to applications) | 12:58 |
@sonney2k | then there is also the clustering demo etc etc etc | 12:58 |
foulwall | Is that only for numbers? | 12:58 |
@sonney2k | foulwall, we only trained for numbers but it would work for anything | 12:59 |
foulwall | seems not so hard | 12:59 |
@sonney2k | so I don't know what you prefer - easiest is probably do clustering (same like you have but just call clustering method) | 12:59 |
@sonney2k | and then classification | 13:00 |
@sonney2k | classification needs a way to select the class label | 13:00 |
@sonney2k | so one would need some option to select some color for the dots | 13:01 |
@sonney2k | (from some fixed set) | 13:01 |
foulwall | so I'll try that tonight | 13:01 |
@sonney2k | then when showing the result it is a bit more difficult as it would be cool to have sth like | 13:01 |
@sonney2k | http://shogun-toolbox.org/bigpicture/krr.png | 13:02 |
@sonney2k | difficult only d3js wise | 13:02 |
@sonney2k | not so much shogun - we have an example for that too | 13:03 |
foulwall | sonney2k: let me have a try. | 13:03 |
@sonney2k | (python_modular/graphical/svm.py) | 13:04 |
foulwall | sonney2k: thanks | 13:04 |
@sonney2k | for clustering / classification we also have this demo interactive_clustering_demo.py interactive_svm_demo.py | 13:05 |
@sonney2k | maybe that helps already... | 13:05 |
@sonney2k | foulwall, please ask whenever you have a question - ask here or on the mailing list. you are doing a great job and it would be stupid that you get stuck over sth shogun/ machine learning wise that for us might be easy | 13:06 |
foulwall | sonney2k: oh, I know that. | 13:07 |
foulwall | sonney2k: i'll ask if i get stucked | 13:07 |
foulwall | I'll begin tonight's coding! | 13:08 |
@sonney2k | foulwall, just to give you an idea how this could go on - we have some other applications that predicts starts of genes / some gene segmentations, various dimension reduction techniques for which one can do nice visualizations, some statistical tests etc etc | 13:11 |
@sonney2k | there is no limit if you find something you would *like* to do - e.g. not just digit recognition but whole alphabet or show how shogun would work on some kaggle.com competition or even an automated way of analyzing data there are no limits and we are happy if you do stuff like that | 13:14 |
foulwall | cool | 13:16 |
foulwall | sonney2k: I've worked a system which can recognize chinese characters and divide the character into strokes,then organize them with user's customized stroke and generate a new character. | 13:18 |
foulwall | sonney2k: but the generated result are terrible... anyway, I'll try the svm tonight and then I'll try the ocr demo. | 13:19 |
@sonney2k | foulwall, I just reviewed you pull request - I am happy with it except one thing: | 13:20 |
foulwall | what's that? | 13:20 |
@sonney2k | why did you change pages/views.py ? | 13:20 |
@sonney2k | looks like you just added some whitspaces? | 13:20 |
foulwall | I add a demoHandler function. | 13:21 |
@sonney2k | yes | 13:22 |
foulwall | the whitespace is my mistake. | 13:22 |
@sonney2k | ok | 13:22 |
foulwall | I won't do that again. | 13:23 |
@sonney2k | foulwall, merged | 13:24 |
foulwall | the next pull I'll check the diff. | 13:24 |
foulwall | sonney2k: thanks , it's the first merge in my life. :-) | 13:25 |
foulwall | sonney2k: does it work? | 13:28 |
@sonney2k | foulwall, yes it works | 13:31 |
@sonney2k | foulwall, we will add more blink lanker | 13:31 |
@sonney2k | later | 13:31 |
foulwall | sonney2k: ok | 13:31 |
@sonney2k | like the curve will be animated | 13:31 |
@sonney2k | and some grid | 13:31 |
@sonney2k | but it is cool already | 13:31 |
foulwall | I'll do that later. | 13:32 |
@sonney2k | foulwall, I have some code for the grid in d3 somewhere | 13:33 |
@sonney2k | so we could just use that and the transition in d3 is easy when you just place the blue line down at (x,0) and then call transition ... | 13:34 |
foulwall | sonney2k: there is a lot of girds already in google, | 13:34 |
foulwall | sonney2k: that's easy. | 13:34 |
foulwall | Just give the work to me, | 13:35 |
foulwall | i'll do that | 13:35 |
@sonney2k | cool! | 13:35 |
@sonney2k | thanks | 13:35 |
foulwall | I'll leave for a dinner | 13:35 |
foulwall | sonney2k: bye sonney2k | 13:37 |
@sonney2k | foulwall, Bon Appétit | 13:37 |
foulwall | sonney2k: thankyou | 13:37 |
-!- foulwall [~foulwall@li379-21.members.linode.com] has quit [Remote host closed the connection] | 13:38 | |
-!- sids_aquarius [~sidi@li400-124.members.linode.com] has quit [Quit: leaving] | 13:58 | |
-!- sids_aquarius [~quassel@14.139.122.114] has joined #shogun | 13:59 | |
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: Reconnecting] | 14:38 | |
-!- heiko [~heiko@94.3.63.154] has joined #shogun | 14:41 | |
@lisitsyn | heiko: answered ;) | 14:42 |
heiko | lisitsyn: just saw :_) | 14:43 |
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 14:43 | |
-!- mode/#shogun [+o iglesiasg] by ChanServ | 14:43 | |
@lisitsyn | heiko: is P square? | 14:43 |
heiko | so I know about the A*x for standard eigenvalues | 14:43 |
heiko | yes | 14:43 |
heiko | P is an incomplete cholesky | 14:43 |
heiko | a preconditioner for another krylov method to work | 14:43 |
heiko | faster | 14:43 |
@lisitsyn | heiko: so you can solve Px pretty fast? | 14:43 |
heiko | the matrix is P*P' though | 14:44 |
heiko | let me try | 14:44 |
@lisitsyn | heiko: the thing is that if you need to solve P*P' x you may solve P (P' x) | 14:44 |
heiko | yes I see | 14:45 |
@lisitsyn | heiko: I just have to check the interface of arpack | 14:45 |
@lisitsyn | heiko: so let us have Ax = M p x eigenproblem | 14:49 |
heiko | lisitsyn: just realised that P in fact is triangular | 14:49 |
@lisitsyn | heiko: to compute largest you would need | 14:49 |
@lisitsyn | inv(M) v | 14:49 |
heiko | so I can tell ARPACK to explot this | 14:49 |
heiko | yes | 14:49 |
@lisitsyn | and A v | 14:49 |
@lisitsyn | I don't know what is the interface in matlab for that | 14:50 |
@lisitsyn | but it looks like something possible | 14:50 |
@lisitsyn | heiko: there is a wrapper for ARPACK + Eigen3 if you are interested | 14:50 |
heiko | lisitsyn: this can be useful for the GSoC | 14:50 |
heiko | currently I am using Matlab, since working with statistitians :) | 14:51 |
@lisitsyn | heiko: arpack wrapper in tapkee is kind of specialized for standard eigenproblems | 14:51 |
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has quit [Ping timeout: 248 seconds] | 14:51 | |
@lisitsyn | heiko: but I use a trick here | 14:52 |
@lisitsyn | heiko: for example in diffusion maps one needs to compute eigenvalues of KK' | 14:52 |
heiko | ok | 14:52 |
heiko | and then? | 14:52 |
@lisitsyn | so instead of storing it I do K(K'x) | 14:52 |
@lisitsyn | more operations but less storage | 14:52 |
@lisitsyn | heiko: it pays off even better in landmark isomap | 14:53 |
@lisitsyn | heiko: if you have say 100000 vectors and only 100 landmarks | 14:53 |
@lisitsyn | it avoids to compute 100000 x 100000 matrix | 14:54 |
@lisitsyn | which would be infeasible anyway | 14:54 |
@lisitsyn | heiko: the same thing is possible for generalized I think | 14:56 |
heiko | yes I see | 14:56 |
heiko | using the same tricks for the logdets | 14:56 |
heiko | for pcg etc | 14:56 |
@lisitsyn | heiko: arpack is awesome because it doesn't care about our matrices | 14:56 |
heiko | yes, only A*x | 14:56 |
@lisitsyn | it needs only products | 14:56 |
@lisitsyn | heiko: that's why I made tapkee callback-based | 14:57 |
heiko | indeed | 14:57 |
heiko | shogun could use this for kernels | 14:57 |
@lisitsyn | heiko: how? | 14:57 |
heiko | I mean callbacks | 14:58 |
heiko | not matrix vetor products | 14:58 |
@lisitsyn | heiko: yes, but how? | 14:58 |
heiko | lisitsyn: nevermind, was just a quick thought | 14:58 |
-!- akshaydixi [~akshay@115.248.50.23] has joined #shogun | 14:59 | |
heiko | so how is the parallel thing going? :D | 14:59 |
@lisitsyn | heiko: I thought you had an idea :) I am always curious about such extensions | 14:59 |
heiko | lisitsyn: btw I would like to parallelise the grid-search model-selection | 14:59 |
heiko | is this easy to do? | 14:59 |
@lisitsyn | heiko: with openmp yes | 14:59 |
heiko | because every loop midifies the features | 14:59 |
heiko | I guess it would only work if one would create a new instance for the machine/data in every iteration? | 15:01 |
@lisitsyn | heiko: let me check the code | 15:02 |
@lisitsyn | heiko: yes probably | 15:03 |
@lisitsyn | or we get in trouble :) | 15:03 |
heiko | lisitsyn: its sad | 15:03 |
heiko | in order to make shogun ready for cluster stuff - a rewrite is necessary :( | 15:03 |
@lisitsyn | yes, of course | 15:04 |
@lisitsyn | heiko: writing code is pretty easy actually | 15:04 |
@lisitsyn | just recall how much time has been spent on _design_ | 15:04 |
heiko | yes | 15:04 |
heiko | maybe at some point ;) | 15:04 |
heiko | lisitsyn: ok, thanks for your tips, going outside now, we actually have sunny wheter :) | 15:05 |
@lisitsyn | heiko: finally? | 15:06 |
@lisitsyn | :) | 15:06 |
heiko | yes :D | 15:06 |
@lisitsyn | enjoy the weather then :) | 15:06 |
heiko | thx :) | 15:07 |
-!- heiko [~heiko@94.3.63.154] has quit [Quit: Leaving.] | 15:14 | |
-!- zuhao [~wanzuhao@nusnet-184-31.dynip.nus.edu.sg] has left #shogun [] | 15:17 | |
-!- zuhao [~wanzuhao@nusnet-184-31.dynip.nus.edu.sg] has joined #shogun | 15:20 | |
-!- sids_aquarius [~quassel@14.139.122.114] has quit [Read error: Connection reset by peer] | 16:06 | |
-!- gsomix [~Miranda@188.168.14.129] has quit [Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org] | 16:14 | |
-!- sids_aquarius [~Siddharth@14.139.122.114] has joined #shogun | 16:14 | |
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 16:55 | |
-!- mode/#shogun [+o iglesiasg] by ChanServ | 16:55 | |
-!- akshaydixi [~akshay@115.248.50.23] has quit [Ping timeout: 256 seconds] | 17:21 | |
-!- akshaydixi [~akshay@115.248.50.23] has joined #shogun | 17:21 | |
-!- sids_aquarius [~Siddharth@14.139.122.114] has quit [Read error: Connection reset by peer] | 19:17 | |
-!- sids_aquarius [~Siddharth@14.139.122.114] has joined #shogun | 19:17 | |
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has quit [Read error: Operation timed out] | 19:53 | |
-!- sids_aquarius [~Siddharth@14.139.122.114] has quit [Read error: Connection reset by peer] | 20:01 | |
@sonney2k | lisitsyn, hurray nice weather | 20:09 |
@sonney2k | I am sitting in the garden just now :D | 20:09 |
@lisitsyn | sonney2k: I'd catch a cold if I was in some garden now | 20:09 |
@lisitsyn | :) | 20:09 |
@lisitsyn | +3or so | 20:09 |
@sonney2k | 16 now here | 20:10 |
@sonney2k | ...yes I wear pants today | 20:10 |
@sonney2k | exception to the rule... | 20:10 |
@lisitsyn | sonney2k: the usual case is w/o? | 20:11 |
@sonney2k | lisitsyn, yes of course | 20:13 |
@sonney2k | danger of overheating | 20:13 |
@sonney2k | lisitsyn, should we assign names to tasks this year? I mean potential candidates? | 20:15 |
@lisitsyn | sonney2k: tasks? | 20:15 |
@lisitsyn | like gsoc tasks? | 20:15 |
@sonney2k | lisitsyn, I meant students -> task | 20:16 |
@lisitsyn | gsoc or entrance? | 20:16 |
@sonney2k | gsoc | 20:16 |
@lisitsyn | ahh you mean put candidate names to the list? | 20:16 |
@sonney2k | question is whether we should have that list public or not | 20:16 |
@lisitsyn | no I don't think so | 20:16 |
@lisitsyn | that demotivates people | 20:17 |
@sonney2k | ok | 20:17 |
@lisitsyn | if I was student and there was some guy I'd think I have no chances | 20:17 |
@lisitsyn | sonney2k: like in opencv or scikits where all places are already 'sold' :D | 20:17 |
@sonney2k | yeah.. | 20:18 |
@lisitsyn | sonney2k: in russian words 'ticket' and 'exam paper' are the same | 20:18 |
@lisitsyn | there is a joke | 20:18 |
@sonney2k | though it is not clear how many students will be applying this year | 20:19 |
@lisitsyn | 'exam won't happen as all (that word) were sold' | 20:19 |
@sonney2k | we have lots of tough stuff this year IMHO | 20:19 |
@sonney2k | hehe | 20:19 |
@lisitsyn | sonney2k: my task is easy I have no idea why they don't want to do that :) | 20:19 |
@lisitsyn | sonney2k: ??????? ?? ?????????, ??? ?????? ??????? | 20:19 |
@lisitsyn | if it makes any sense | 20:20 |
@sonney2k | cool I see bats hunting :) | 20:21 |
@lisitsyn | sonney2k: have you seen crazy belarussian beaver? | 20:22 |
@lisitsyn | sonney2k: http://www.youtube.com/watch?feature=player_detailpage&v=z10EpPbfSoQ | 20:22 |
@sonney2k | uhh a cat is hunting too | 20:23 |
@lisitsyn | sonney2k: what a cat is hunting for? | 20:24 |
@lisitsyn | bats? | 20:24 |
@lisitsyn | :D | 20:24 |
@sonney2k | lisitsyn, http://shogun-toolbox.org/demo/svr/entrance | 20:28 |
@sonney2k | lisitsyn, birds I guess | 20:28 |
@lisitsyn | sonney2k: oh we have a PR on GPs | 20:28 |
@sonney2k | I wonder where the bats live here | 20:29 |
@lisitsyn | sonney2k: where do birds live? | 20:29 |
@lisitsyn | I guess they all live under roofs | 20:29 |
@lisitsyn | sonney2k: how do you link shogun-toolbox.org with the server? | 20:31 |
@lisitsyn | A record? | 20:31 |
@sonney2k | what? | 20:32 |
@lisitsyn | sonney2k: domain name -> IP? | 20:32 |
@lisitsyn | or domain name -> domain name? | 20:32 |
@sonney2k | lisitsyn, the former | 20:34 |
@lisitsyn | sonney2k: so the IP is assumed to be static? | 20:34 |
@sonney2k | it is static | 20:39 |
-!- lambday [3ba21749@gateway/web/freenode/ip.59.162.23.73] has joined #shogun | 20:42 | |
@lisitsyn | I see | 20:42 |
lambday | heiko? :( | 20:53 |
-!- lambday [3ba21749@gateway/web/freenode/ip.59.162.23.73] has quit [] | 21:06 | |
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun | 21:12 | |
shogun-notifier- | shogun: Roman Votyakov :develop * 9f96593 / src/shogun/regression/gp/ExactInferenceMethod.cpp: https://github.com/shogun-toolbox/shogun/commit/9f96593b8a78cb60f3127182b3e3bf4b46c2c1e3 | 21:12 |
shogun-notifier- | shogun: Fixed recomputing of Cholesky factor in ExactInferenceMethod class | 21:12 |
shogun-notifier- | shogun: Sergey Lisitsyn :develop * e98d84b / src/shogun/regression/gp/ExactInferenceMethod.cpp: https://github.com/shogun-toolbox/shogun/commit/e98d84b66f7d1e2414027a75fff92160dcfc1c92 | 21:12 |
shogun-notifier- | shogun: Merge pull request #972 from votjakovr/develop | 21:12 |
shogun-notifier- | shogun: | 21:12 |
shogun-notifier- | shogun: Fixed recomputing of Cholesky factor in ExactInferenceMethod class | 21:12 |
-!- akshaydixi [~akshay@115.248.50.23] has quit [Ping timeout: 264 seconds] | 21:25 | |
@lisitsyn | sonney2k: come on! | 21:31 |
@lisitsyn | ah | 21:31 |
@lisitsyn | I thought you commited to master of shogun | 21:31 |
@lisitsyn | not shogun-web | 21:31 |
@lisitsyn | :D | 21:31 |
@lisitsyn | sorry my bad | 21:31 |
-!- akshaydixi [~akshay@27.63.9.74] has joined #shogun | 21:37 | |
-!- hoijui [~hoijui@dslb-088-074-124-102.pools.arcor-ip.net] has joined #shogun | 21:56 | |
-!- gsomix [~gsomix@83.234.54.155] has joined #shogun | 21:59 | |
gsomix | good evening, guys | 21:59 |
@sonney2k | lisitsyn, :P | 22:20 |
-!- iglesiasg [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 22:31 | |
-!- mode/#shogun [+o iglesiasg] by ChanServ | 22:31 | |
@iglesiasg | sonney2k: it seems something broke with the demo? | 22:32 |
@sonney2k | iglesiasg, I just fixed it | 22:34 |
@sonney2k | iglesiasg, it is actaully working ok | 22:34 |
@sonney2k | it was not on the server because gsomix' buffer protocol support was not working there (too old numpy version...) | 22:35 |
@sonney2k | so labels.get_labels() was necessary | 22:35 |
@sonney2k | instead of directly doing labels[i] | 22:35 |
gsomix | sonney2k, at first I was like "O_O", but then I just finished reading the string | 22:38 |
gsomix | *sentence | 22:38 |
@iglesiasg | sonney2k: aham ok, I see | 22:39 |
@lisitsyn | gsomix: catched you, robot | 22:39 |
@iglesiasg | sonney2k: BTW, I have seen that the URL is http://shogun-toolbox.org/demo/svr/entrance but didn't find links from the homepage | 22:40 |
@iglesiasg | are there any? | 22:40 |
gsomix | lisitsyn, I just remembered: http://www.youtube.com/watch?v=HGSVYgcy24Q | 22:40 |
@sonney2k | iglesiasg, no links yet | 22:41 |
@sonney2k | lets first all try it out and get it rock stable before proceeding | 22:41 |
@iglesiasg | good idea | 22:43 |
-!- akshaydixi [~akshay@27.63.9.74] has quit [Ping timeout: 245 seconds] | 22:52 | |
-!- zuhao [~wanzuhao@nusnet-184-31.dynip.nus.edu.sg] has quit [Quit: zuhao] | 23:29 | |
gsomix | nite | 23:53 |
@iglesiasg | night gsomix | 23:54 |
--- Log closed Mon Apr 15 00:00:28 2013 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!