--- Log opened Wed Nov 23 00:00:51 2011 | ||
-!- in3xes [~in3xes@180.149.49.230] has joined #shogun | 02:25 | |
-!- in3xes [~in3xes@180.149.49.230] has quit [Ping timeout: 260 seconds] | 04:16 | |
-!- in3xes [~in3xes@180.149.49.230] has joined #shogun | 05:00 | |
-!- in3xes [~in3xes@180.149.49.230] has quit [Ping timeout: 240 seconds] | 05:19 | |
-!- in3xes [~in3xes@180.149.49.230] has joined #shogun | 05:46 | |
-!- in3xes [~in3xes@180.149.49.230] has quit [Ping timeout: 252 seconds] | 05:53 | |
-!- in3xes [~in3xes@180.149.49.230] has joined #shogun | 06:37 | |
-!- in3xes [~in3xes@180.149.49.230] has quit [Ping timeout: 240 seconds] | 06:43 | |
CIA-37 | shogun: Soeren Sonnenburg master * recdb955 / (3 files in 3 dirs): rb_require narray on modshogun module load - http://git.io/JHWmTw | 08:53 |
---|---|---|
-!- policecar [5ce7be6e@gateway/web/freenode/ip.92.231.190.110] has joined #shogun | 10:24 | |
policecar | hey hey, i'm manually installing shogun 1.0.0 on snowleopard and trying to make it work with a macports installed python2.6. do you happen to know where i need to put the shogun binaries for this python to be able to see the module? basically i'm looking for a macport not fink path version of /usr/local/lib/python2.6/dist-packages/ ... | 10:33 |
-!- policecar_ [~policecar@g231190110.adsl.alicedsl.de] has joined #shogun | 10:36 | |
-!- policecar [5ce7be6e@gateway/web/freenode/ip.92.231.190.110] has quit [Quit: Page closed] | 10:37 | |
-!- policecar_ is now known as policecar | 10:37 | |
sonne|work | policecar: please try shogun from git | 11:52 |
sonne|work | or did it compile OK already | 11:52 |
sonne|work | (I did a lot of fixes recently) | 11:53 |
policecar | it compiled just fine. but thanks for the hint. | 11:53 |
sonne|work | ok then | 11:53 |
sonne|work | just search for e.g. modshogun.py | 11:53 |
policecar | i just solved the problem by adding /usr/local/lib/python2.6/dist-packages to the pythonpath a moment ago | 11:53 |
policecar | which is not nice but it works for now | 11:53 |
sonne|work | exactly what you should be doing :) | 11:53 |
policecar | maybe i missed the instruction | 11:53 |
policecar | thanks for getting back to me! | 11:54 |
sonne|work | policecar: well no - in principle it should work out of the box but everyone is cooking his own soup | 11:54 |
policecar | yep | 11:55 |
-!- policecar [~policecar@g231190110.adsl.alicedsl.de] has quit [Quit: policecar] | 14:29 | |
-!- policecar [~policecar@g231190110.adsl.alicedsl.de] has joined #shogun | 14:40 | |
-!- policecar [~policecar@g231190110.adsl.alicedsl.de] has quit [Client Quit] | 14:43 | |
-!- policecar [~policecar@g231190110.adsl.alicedsl.de] has joined #shogun | 15:05 | |
-!- policecar [~policecar@g231190110.adsl.alicedsl.de] has left #shogun [] | 15:06 | |
-!- blackburn [~blackburn@188.168.4.253] has joined #shogun | 15:12 | |
blackburn | sonney2k: so is GLPK the only way to make MKL regression work? | 20:03 |
CIA-37 | shogun: Soeren Sonnenburg master * rbd6dcde / (5 files in 3 dirs): | 20:07 |
CIA-37 | shogun: get ruby back to work | 20:07 |
CIA-37 | shogun: - add configure option --ruby to be able to specify the ruby version to be used | 20:07 |
CIA-37 | shogun: - use rb_cObject instead of ruby_class to fix undefined symbol error | 20:07 |
CIA-37 | shogun: - beautify example - http://git.io/5PAJog | 20:07 |
@sonney2k | blackburn, no - we have mkl_regression_3sine_waves.m for example | 20:08 |
@sonney2k | so it works with regression too | 20:08 |
blackburn | sonney2k: what is the problem then? | 20:08 |
@sonney2k | I need to reboot again to check if things work under osx | 20:08 |
@sonney2k | cu | 20:08 |
blackburn | ok | 20:08 |
@sonney2k | blackburn, yes works all nice now | 20:19 |
@sonney2k | :D | 20:19 |
blackburn | sonney2k: ruby? | 20:20 |
@sonney2k | yes | 20:20 |
blackburn | nice | 20:20 |
@sonney2k | really weird hack I am using... | 20:22 |
@sonney2k | we no longer link to narray.so | 20:22 |
@sonney2k | but instead open it with dlopen | 20:22 |
@sonney2k | and then extract the few needed functions | 20:22 |
@sonney2k | with dlsym | 20:22 |
@sonney2k | regarding mkl + regression I suspect kernels are not properly normalized | 20:22 |
blackburn | sonney2k: similar to java I would say? | 20:23 |
blackburn | sonney2k: have you seen tikz? completely crazy stuff http://www.texample.net/tikz/examples/seismic-focal-mechanism-in-3d-view/ | 20:24 |
@sonney2k | one potential fix would be to use the sqrtdiag kernel normalizer | 20:25 |
@sonney2k | latex wizardry... | 20:26 |
blackburn | sonney2k: the problem I have now is after changing script there is an index error.. | 20:27 |
blackburn | i.e. before train one kernel is used | 20:28 |
@sonney2k | I am not sure what you are saying | 20:28 |
blackburn | sonney2k: mkl is trained with one combined kernel | 20:28 |
blackburn | then another is set | 20:28 |
@sonney2k | yes | 20:28 |
blackburn | and apply is called | 20:28 |
@sonney2k | yes | 20:28 |
blackburn | and I receive index errors | 20:28 |
blackburn | in apply | 20:28 |
@sonney2k | in the example I pushed? | 20:28 |
blackburn | no | 20:28 |
@sonney2k | in which? | 20:29 |
blackburn | in the example I sent to you with dropbox | 20:29 |
blackburn | if you have time please take a look | 20:29 |
blackburn | hmm | 20:30 |
blackburn | then if I change this line: | 20:30 |
blackburn | cols,rows = numpy.meshgrid(test_idxs2,train_idxs2) | 20:30 |
blackburn | it segfaults.. | 20:30 |
@sonney2k | which line? | 20:30 |
@sonney2k | which file | 20:31 |
blackburn | mkl_regression_cv.py | 20:31 |
blackburn | 77 | 20:31 |
@sonney2k | damn complex example | 20:32 |
blackburn | sonney2k: sorry, it is a port of example madeleine uses | 20:32 |
blackburn | in java it was even worse | 20:32 |
blackburn | sonney2k: I rewrote whole example trying to solve her issue | 20:33 |
blackburn | and it leaded to another issues.. | 20:33 |
@sonney2k | GLPK can only be used with mkl_norm=1 | 20:33 |
@sonney2k | great | 20:33 |
@sonney2k | I guess this is kind of spaghetti research code | 20:33 |
blackburn | sonney2k: it is CV in fact | 20:34 |
@sonney2k | blackburn, why are you using meshgrid | 20:34 |
blackburn | sonney2k: what to use? | 20:34 |
@sonney2k | just index oeprations | 20:34 |
@sonney2k | operations | 20:35 |
blackburn | ehmm won't it be diags? | 20:35 |
@sonney2k | like m[idx][:,idx] ? | 20:35 |
shogun-buildbot | build #235 of ruby_modular is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/ruby_modular/builds/235 | 20:36 |
blackburn | sonney2k: is it a share file? did you file changed anyhow? | 20:36 |
blackburn | your* | 20:36 |
@sonney2k | sry but I don't have time now - just ask I will answer late | 20:37 |
@sonney2k | are | 20:37 |
blackburn | sonney2k: ok sure | 20:38 |
@sonney2k | blackburn, any news? | 23:49 |
blackburn | sonney2k: no, I'm still pretty busy with my semester work :( | 23:52 |
blackburn | sonney2k: ah yes, I've got with segfault labels | 23:53 |
blackburn | with this example | 23:53 |
blackburn | ==15121== Invalid read of size 1 | 23:54 |
blackburn | ==15121== at 0x8B2F49F: make_contiguous(_object*, int*, int, int, bool) (modshogun_wrap.cxx:4177) | 23:54 |
blackburn | ==15121== by 0x8E56143: _wrap_new_Labels__SWIG_2 (modshogun_wrap.cxx:261084) | 23:54 |
blackburn | ==15121== by 0x8E5671C: _wrap_new_Labels (modshogun_wrap.cxx:261181) | 23:54 |
@sonney2k | impressive that you always find sth new... | 23:56 |
@sonney2k | no way to do anything about it w/o seeing line 4177 | 23:56 |
blackburn | sonney2k: I guess it is free flag issue | 23:57 |
blackburn | 4177 if (PyArray_ISFARRAY(ary) && !force_copy) | 23:58 |
blackburn | 4178 { | 23:58 |
blackburn | 4179 array = ary; | 23:58 |
blackburn | 4180 *is_new_object = 0; | 23:58 |
blackburn | 4181 } | 23:58 |
blackburn | 4173 static PyObject* make_contiguous(PyObject* ary, int* is_new_object, | 23:59 |
blackburn | 4174 int dims, int typecode, bool force_copy=false) | 23:59 |
blackburn | 4175 { | 23:59 |
blackburn | 4176 PyObject* array; | 23:59 |
blackburn | sonney2k: not really better I would say :) | 23:59 |
--- Log closed Thu Nov 24 00:00:13 2011 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!