| --- Log opened Sun Sep 11 00:00:17 2011 | ||
| shogun-buildbot | build #209 of csharp_modular is complete: Failure [failed compile] Build details are at http://www.shogun-toolbox.org/buildbot/builders/csharp_modular/builds/209 blamelist: blackburn91@gmail.com | 00:11 |
|---|---|---|
| shogun-buildbot | build #218 of octave_modular is complete: Failure [failed compile] Build details are at http://www.shogun-toolbox.org/buildbot/builders/octave_modular/builds/218 blamelist: blackburn91@gmail.com | 00:15 |
| blackburn | Speicherzugriffsfehler | 00:16 |
| blackburn | I guess it is something like PieceOfShitError? | 00:16 |
| blackburn | :D | 00:16 |
| shogun-buildbot | build #219 of octave_modular is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/octave_modular/builds/219 | 00:25 |
| shogun-buildbot | build #198 of ruby_modular is complete: Failure [failed test_1] Build details are at http://www.shogun-toolbox.org/buildbot/builders/ruby_modular/builds/198 blamelist: blackburn91@gmail.com | 01:02 |
| shogun-buildbot | build #199 of ruby_modular is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/ruby_modular/builds/199 | 01:16 |
| -!- blackburn [~blackburn@31.28.44.65] has quit [Quit: Leaving.] | 01:18 | |
| shogun-buildbot | build #210 of csharp_modular is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/csharp_modular/builds/210 | 01:26 |
| -!- blackburn [~blackburn@31.28.44.65] has joined #shogun | 10:33 | |
| @sonney2k | blackburn, yes | 12:05 |
| CIA-3 | shogun: Soeren Sonnenburg master * rb1edffb / src/shogun/kernel/LocalAlignmentStringKernel.cpp : use CALLOC instead of MALLOC - http://git.io/Xo1Xag | 20:25 |
| CIA-3 | shogun: Soeren Sonnenburg master * r73e9943 / examples/undocumented/python_modular/graphical/svm.py : use axis tight to improve figure - http://git.io/05RWKg | 20:25 |
| blackburn | sonney2k: wow did you fix that? | 20:26 |
| @sonney2k | blackburn, did you already commit your fast LLE version? | 20:26 |
| blackburn | sonney2k: yes | 20:26 |
| @sonney2k | blackburn, don't know | 20:26 |
| blackburn | sonney2k: I've spent the whole day trying to get packing numbers work nicely, no success yet | 20:26 |
| blackburn | now continued integrating superlu | 20:27 |
| @sonney2k | I don't know what packing numbers are | 20:27 |
| blackburn | sonney2k: not important, just told you what is I'm working on :) | 20:28 |
| shogun-buildbot | build #219 of r_static is complete: Failure [failed git] Build details are at http://www.shogun-toolbox.org/buildbot/builders/r_static/builds/219 blamelist: sonne@debian.org | 20:29 |
| shogun-buildbot | build #217 of octave_static is complete: Failure [failed git] Build details are at http://www.shogun-toolbox.org/buildbot/builders/octave_static/builds/217 blamelist: sonne@debian.org | 20:29 |
| shogun-buildbot | build #231 of cmdline_static is complete: Failure [failed git] Build details are at http://www.shogun-toolbox.org/buildbot/builders/cmdline_static/builds/231 blamelist: sonne@debian.org | 20:29 |
| shogun-buildbot | build #212 of python_static is complete: Failure [failed git] Build details are at http://www.shogun-toolbox.org/buildbot/builders/python_static/builds/212 blamelist: sonne@debian.org | 20:29 |
| blackburn | sonney2k: local alignment still failing | 20:51 |
| @sonney2k | blackburn, with which version did you compare last? | 20:52 |
| blackburn | 0.9.0 | 20:52 |
| @sonney2k | I mean did it work in 0.8.sth? | 20:52 |
| blackburn | sonney2k: I didn't test it | 20:53 |
| @sonney2k | blackburn, I guess it might be sufficient to do a git diff agains 0.8 | 20:53 |
| shogun-buildbot | build #13 of nightly_default is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/nightly_default/builds/13 | 21:52 |
| blackburn | sonney2k: lle is really fast now - 223s for 10000 vectors | 22:01 |
| blackburn | but superlu to go | 22:01 |
| @sonney2k | so is it up to par to scikits now? | 22:02 |
| @sonney2k | or faster? | 22:02 |
| blackburn | sonney2k: much faster | 22:02 |
| @sonney2k | how fast is scikits? | 22:02 |
| blackburn | well I didn't test on 10k but on 8k it was 2 times slower | 22:02 |
| blackburn | sonney2k: hidden feature of my impl is parallelization - it can really impact on high-dimensional data | 22:03 |
| blackburn | I got crazy with superlu, kind of complicated thing | 22:04 |
| blackburn | sonney2k: 3k examples: | 22:07 |
| blackburn | SHOGUN Took 9.007753s | 22:07 |
| blackburn | SKLEARN Took 11.569473s | 22:07 |
| @sonney2k | about the same | 22:09 |
| blackburn | sonney2k: wait a minute for 6k | 22:09 |
| @sonney2k | blackburn, if you write a paper you should have a figure like this showing how much faster your stuff is somewhere (potentially even in shogun docs for dim red stuff) | 22:10 |
| blackburn | I know | 22:10 |
| blackburn | sonney2k: I will improve perfomance with superlu, then it would be faster, may be significantly | 22:11 |
| blackburn | SKLEARN Took 91.296676s | 22:11 |
| blackburn | SHOGUN Took 61.362044s | 22:11 |
| blackburn | sonney2k: ^ not really the same | 22:11 |
| blackburn | and now I turn on parallel things | 22:12 |
| blackburn | SHOGUN Took 50.653053s | 22:13 |
| @sonney2k | 2 cores? | 22:13 |
| blackburn | yes | 22:14 |
| blackburn | I believe it would be faster on really multiprocessor machine | 22:14 |
| @sonney2k | what is the bottleneck currently? | 22:14 |
| blackburn | sonney2k: sparse dot and eigenproblem almost the same | 22:15 |
| blackburn | neighbor searching is fast | 22:15 |
| @sonney2k | sparse dot is parallel? | 22:15 |
| blackburn | yes | 22:15 |
| @sonney2k | but eigenproblem is not right? | 22:15 |
| blackburn | eigenproblem is not | 22:15 |
| @sonney2k | is this with lapack? | 22:15 |
| blackburn | arpack | 22:15 |
| @sonney2k | ahh - is arpack much faster? | 22:15 |
| blackburn | yes, twice | 22:16 |
| blackburn | scikits uses arpack too | 22:16 |
| @sonney2k | (because there is the parallel scalapack) | 22:16 |
| @sonney2k | ok | 22:16 |
| blackburn | sonney2k: arpack could impact from parallel lapack | 22:16 |
| blackburn | it uses lapack ops | 22:16 |
| blackburn | e.g. it provides a pointer to a vector you should product with matrix :) | 22:17 |
| blackburn | or solve the system | 22:17 |
| blackburn | now I'm using LU/Cholesky decomposition from lapack | 22:17 |
| blackburn | it is pretty fast but doesn't take any advantages of matrix being \sparse | 22:18 |
| blackburn | sonney2k: well finally, arpack have parallel version too :) | 22:19 |
| blackburn | P_ARPACK | 22:19 |
| blackburn | no idea how to work with it though | 22:19 |
| blackburn | it lacks any docs | 22:19 |
| @sonney2k | nite | 22:31 |
| blackburn | see you | 22:32 |
| -!- blackburn [~blackburn@31.28.44.65] has quit [Quit: Leaving.] | 23:01 | |
| --- Log closed Mon Sep 12 00:00:22 2011 | ||
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!