| --- Log opened Fri Feb 06 00:00:07 2015 | ||
| -!- nszceta [~nszceta@24.131.255.77] has joined #shogun | 00:33 | |
| AndreLopes | Hi | 01:26 |
|---|---|---|
| AndreLopes | ../examples/documented/java_modular/classifier_averaged_perceptron_modular.java | 01:26 |
| AndreLopes | In this example | 01:26 |
| AndreLopes | I got a warning | 01:26 |
| AndreLopes | [WARN] In file /builddir/build/BUILD/shogun-96f3cf3ce58514299f98e688b7c43e057ad4fa41/src/shogun/classifier/AveragedPerceptron.cpp line 90: Averaged Perceptron algorithm did not converge after 10 iterations. | 01:28 |
| AndreLopes | Is this normal to the example ? | 01:28 |
| AndreLopes | And why does the neural network only accepts a DynamicObjectArray ? | 01:44 |
| AndreLopes | And also this.. SwigType_p_p_shogun_CSGObject | 01:44 |
| -!- nszceta [~nszceta@24.131.255.77] has quit [Quit: Textual IRC Client: www.textualapp.com] | 02:21 | |
| -!- PirosB3 [~pirosb3@host180-35-dynamic.18-79-r.retail.telecomitalia.it] has quit [Quit: PirosB3] | 02:47 | |
| -!- AndreLopes [bae787ab@gateway/web/cgi-irc/kiwiirc.com/ip.186.231.135.171] has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client] | 03:54 | |
| shogun-buildbot | build #962 of nightly_default is complete: Failure [failed notebooks] Build details are at http://buildbot.shogun-toolbox.org/builders/nightly_default/builds/962 blamelist: sanuj <sanuj.sharma.in@gmail.com>, Kunal Arora <kunalarora.135@gmail.com>, Fernando Iglesias <fernando.iglesiasg@gmail.com>, Bj?rn Esser <bjoern.esser@gmail.com> | 04:08 |
| -!- lambday [c40f1064@gateway/web/freenode/ip.196.15.16.100] has joined #shogun | 07:02 | |
| -!- mode/#shogun [+o lambday] by ChanServ | 07:02 | |
| -!- chintak [6acef958@gateway/web/cgi-irc/kiwiirc.com/ip.106.206.249.88] has joined #shogun | 07:36 | |
| -!- chintak [6acef958@gateway/web/cgi-irc/kiwiirc.com/ip.106.206.249.88] has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client] | 08:21 | |
| -!- besser82 [~besser82@2a02:8108:8800:3d00:f2de:f1ff:fe89:42d4] has joined #shogun | 09:22 | |
| -!- besser82 [~besser82@2a02:8108:8800:3d00:f2de:f1ff:fe89:42d4] has quit [Changing host] | 09:22 | |
| -!- besser82 [~besser82@fedora/besser82] has joined #shogun | 09:22 | |
| -!- mode/#shogun [+o besser82] by ChanServ | 09:22 | |
| -!- sonne|work [~sonnenbu@ip5b415e18.dynamic.kabel-deutschland.de] has quit [Quit: Leaving.] | 11:33 | |
| @besser82 | lisitsyn, ping? ^^ | 12:09 |
| @besser82 | lisitsyn, can you please give me commit-access to your swig-fork on github? =) | 12:10 |
| -!- lambday [c40f1064@gateway/web/freenode/ip.196.15.16.100] has quit [Ping timeout: 246 seconds] | 12:38 | |
| @lisitsyn | besser82: oops sorry | 13:13 |
| @lisitsyn | yeah sure | 13:13 |
| @lisitsyn | besser82: have fun :) | 13:15 |
| @besser82 | lisitsyn, thx ^^ | 13:15 |
| -!- Netsplit *.net <-> *.split quits: aditya95 | 13:19 | |
| -!- Netsplit over, joins: aditya95 | 13:21 | |
| @besser82 | lisitsyn, force pushed rebased commit into bugfix-branch | 13:28 |
| @besser82 | lisitsyn, you should be able to resolve that by: git checkout bugfix/apple_clang_check && git fetch --all && git reset --hard origin/bugfix/apple_clang_check | 13:29 |
| -!- exploreshaifali [~shaifali@182.70.164.164] has joined #shogun | 13:38 | |
| -!- PirosB3 [~pirosb3@host180-35-dynamic.18-79-r.retail.telecomitalia.it] has joined #shogun | 14:50 | |
| -!- exploreshaifali [~shaifali@182.70.164.164] has quit [Ping timeout: 256 seconds] | 15:28 | |
| -!- exploreshaifali [~shaifali@182.70.164.164] has joined #shogun | 16:06 | |
| -!- sonne|work [~sonnenbu@ip5b415e18.dynamic.kabel-deutschland.de] has joined #shogun | 16:18 | |
| sonne|work | lisitsyn: around? | 16:19 |
| @lisitsyn | sonne|work: yes | 16:21 |
| sonne|work | lisitsyn: I have a ML kind of problem - I have a few distributions (say 10) | 16:21 |
| sonne|work | and some mixture of these 10 distributions with unknown coefficients | 16:22 |
| sonne|work | I am now looking for a way to estimate the coefficients | 16:22 |
| sonne|work | and these given distributions are just based on counts - so histograms | 16:23 |
| @lisitsyn | sonne|work: any assumption on distributions? :) | 16:24 |
| sonne|work | none | 16:24 |
| @lisitsyn | why not to fit GMM? | 16:25 |
| sonne|work | I could smooth them all :) | 16:25 |
| sonne|work | well the base distribs are not gaussians | 16:25 |
| @lisitsyn | then choose something general and fit it with l-bfgs maybe? | 16:25 |
| sonne|work | or some gradient decent kind of thing ... hmmhh sure would work | 16:26 |
| @lisitsyn | sonne|work: scipy has cool fmin_bfgs just give it a task | 16:27 |
| sonne|work | lisitsyn: yeah that would do it | 16:28 |
| @lisitsyn | sonne|work: can you assume gamma distribution | 16:29 |
| @lisitsyn | ? | 16:29 |
| sonne|work | lisitsyn: no the histograms can be anything even some multimodal stuff | 16:29 |
| sonne|work | lisitsyn: but that's good enough | 16:30 |
| @lisitsyn | sonne|work: fitting multimodal mixture sounds scary :D | 16:30 |
| sonne|work | lisitsyn: relax all 1d - not so scary | 16:31 |
| @lisitsyn | ah | 16:31 |
| @lisitsyn | why not to assume # of distributions equals # of modes? | 16:31 |
| @lisitsyn | then gaussian is good | 16:32 |
| sonne|work | all I know is that the resulting distribution is a mixture of the others - and I would 'just' want to know the most likely mixture coefficients | 16:35 |
| sonne|work | if the match is good enough then I can infer voila the distribution did get generated from the following sources | 16:35 |
| -!- abinashm [cb6ef316@gateway/web/freenode/ip.203.110.243.22] has joined #shogun | 16:47 | |
| -!- abinashm [cb6ef316@gateway/web/freenode/ip.203.110.243.22] has quit [Quit: Page closed] | 17:28 | |
| -!- hargup [uid10222@gateway/web/irccloud.com/x-omkchnohteteinmb] has joined #shogun | 18:12 | |
| -!- abinashm [abinashmeh@gateway/shell/waartaa/x-giqmdrqtnbgvniot] has joined #shogun | 18:18 | |
| -!- abinashm [abinashmeh@gateway/shell/waartaa/x-giqmdrqtnbgvniot] has left #shogun [] | 18:20 | |
| -!- abinashm_ [uid66179@gateway/web/irccloud.com/x-dhzqbuwhaexvulil] has joined #shogun | 18:23 | |
| -!- besser82 [~besser82@fedora/besser82] has quit [Ping timeout: 252 seconds] | 19:48 | |
| @wiking | hey | 21:14 |
| @wiking | anybody here? | 21:14 |
| -!- PirosB3 [~pirosb3@host180-35-dynamic.18-79-r.retail.telecomitalia.it] has quit [Quit: PirosB3] | 21:15 | |
| @lisitsyn | wiking: ja | 21:16 |
| @wiking | lisitsyn: how do u use numpy array as a RealFeatures? | 21:17 |
| @wiking | btw pytthon modular is totally broken still on osx | 21:17 |
| @wiking | segfaults:< | 21:17 |
| @lisitsyn | wiking: yeah I know | 21:17 |
| @lisitsyn | RealFeatures(array) | 21:17 |
| -!- sonne|osx [~sonne@f055012204.adsl.alicedsl.de] has joined #shogun | 21:30 | |
| @wiking | lisitsyn: here? | 22:10 |
| @wiking | lisitsyn: how the hell do u matplotlib a 2d numpy array, where each element is [a,b]. and the whole numpy array is like [[a,b],[x,v]...] | 22:11 |
| @lisitsyn | wiking: hmm | 22:24 |
| @lisitsyn | wiking: dont' get it | 22:24 |
| @lisitsyn | you need 2xN matrix? | 22:24 |
| @wiking | i just want to plot a 2d graph | 22:27 |
| @wiking | where the input is | 22:27 |
| @wiking | [[[a,b],[x,v]...] | 22:27 |
| @wiking | imean [[a,b],[x,v]...] | 22:28 |
| @wiking | but if i do axis.plot(x...) | 22:28 |
| @wiking | then i just get one part of each element of the arrays (the first) as y and x is basicaly the lenght of the array | 22:28 |
| -!- exploreshaifali [~shaifali@182.70.164.164] has quit [Quit: Leaving] | 22:36 | |
| --- Log closed Sat Feb 07 00:00:09 2015 | ||
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!