--- Log opened Wed Oct 10 00:00:13 2012 | ||
--- Day changed Wed Oct 10 2012 | ||
-!- zxtx [~zv@rrcs-24-199-47-170.west.biz.rr.com] has joined #shogun | 00:00 | |
blackburn | heiko: any news? | 00:58 |
---|---|---|
heiko | blackburn, I am doing some more advanced two sample tests currently | 01:16 |
heiko | based on the information about the optimal kernel for that I could get 65% with a gaussian SVM :) | 01:16 |
heiko | but these things will have to run over night. | 01:16 |
heiko | next thing I will try is a more sophisticated kernel | 01:16 |
heiko | since that also works for two sample testing | 01:16 |
heiko | so I have a benchmark | 01:17 |
heiko | but going to bed now | 01:17 |
heiko | so see you :) | 01:17 |
blackburn | I see | 01:27 |
blackburn | okay I'm off too | 01:27 |
blackburn | see you | 01:27 |
-!- blackburn [~blackburn@37.61.179.175] has quit [Quit: Leaving.] | 01:29 | |
shogun-buildbot | build #128 of nightly_default is complete: Failure [failed test] Build details are at http://www.shogun-toolbox.org/buildbot/builders/nightly_default/builds/128 | 03:44 |
-!- zxtx [~zv@rrcs-24-199-47-170.west.biz.rr.com] has quit [Ping timeout: 260 seconds] | 03:51 | |
-!- zxtx [~zv@76.91.80.197] has joined #shogun | 07:05 | |
-!- zxtx [~zv@76.91.80.197] has quit [Ping timeout: 245 seconds] | 08:25 | |
-!- zxtx [~zv@cpe-75-83-151-252.socal.res.rr.com] has joined #shogun | 09:27 | |
wiking | mmm do we have api for doing a dot product of a dense and a sparse vector? | 10:31 |
sonne|work | wiking: you mean SGSparseVector * SGVector? | 10:34 |
wiking | yes | 10:34 |
wiking | or for that matter SGSparseVector * SGSparseVector | 10:34 |
sonne|work | unfortunately not... | 10:34 |
wiking | mmm | 10:35 |
sonne|work | we have that code in CSparseFeatures | 10:35 |
wiking | yes | 10:35 |
wiking | i saw that | 10:35 |
sonne|work | ...it needs to be extracted to SGSparseVector | 10:35 |
wiking | yeah | 10:35 |
wiking | i'll do that | 10:35 |
wiking | will send a PR about it | 10:35 |
wiking | dense_dot will be a good name for the feature there as well... right? | 10:36 |
wiking | lol | 10:44 |
wiking | http://math.nist.gov/spblas/ | 10:44 |
sonne|work | hmmhh last update 2006?? | 10:46 |
wiking | i wonder if there's other lib out there for efficient sparse calculations | 10:47 |
wiking | apparently | 10:48 |
wiking | MTL4 | 10:48 |
wiking | The Matrix Template Library version 4 is a generic C++ template library providing sparse and dense BLAS functionality. MTL4 establishes an intuitive interface (similar to MATLAB) and broad applicability thanks to Generic programming. | 10:48 |
wiking | eigen seems to have something for sparse matrices | 10:48 |
sonne|work | then better use eigen | 10:48 |
sonne|work | nothing hurts more than depending on some lib that gets no longer maintained | 10:49 |
wiking | mmm | 10:50 |
wiking | http://eigen.tuxfamily.org/dox/classEigen_1_1SparseMatrixBase.html#af64e012789113d938655fbb7c5f4538e | 10:50 |
wiking | so it seems there is | 10:50 |
wiking | the only thing is now that we store sparse matrices differently than in eigen :) | 10:52 |
wiking | but yeah i dont care about this now | 10:52 |
wiking | i'll optimize later | 10:52 |
wiking | i just want to see results :) | 10:53 |
sonne|work | wiking: however this needs to work even if eigen is not available so I guess extracting the function from sparse features is simplest | 10:53 |
sonne|work | wiking: and if you look at dense_dot | 10:53 |
sonne|work | the function is trivial | 10:53 |
sonne|work | 5 lines or so | 10:53 |
wiking | yeah i've did that | 10:56 |
wiking | doing tests now | 10:56 |
ptizoom | hi, I had a break through, or more like a punny leap, with my perl-swig-sg activity and unslaught your git with a "pull-request", there are lots of things to tie up, help needed! | 12:04 |
-!- ptizoom [~ptizoom@85.210.93.120] has quit [Remote host closed the connection] | 12:52 | |
-!- romi_ [~mizobe@187.57.2.253] has joined #shogun | 13:10 | |
-!- romi_ [~mizobe@187.57.2.253] has quit [Remote host closed the connection] | 13:56 | |
-!- romi_ [~mizobe@187.57.2.253] has joined #shogun | 14:09 | |
-!- ptizoom [~christian@85.210.93.120] has joined #shogun | 15:08 | |
-!- ptizoom [~christian@85.210.93.120] has left #shogun [] | 15:12 | |
-!- ptizoom [~christian@85.210.93.120] has joined #shogun | 15:13 | |
sonne|work | ptizoom: to get you right - it is OK for you if we just start cherry picking patches right? | 15:22 |
sonne|work | ptizoom: you don't get credit in git logs, only some AUTHORS file later - that is why I am asking | 15:23 |
ptizoom | hi | 15:24 |
ptizoom | haaa, credits! | 15:24 |
ptizoom | I am not there yet. | 15:24 |
ptizoom | I was saying that my pull-request is not really appropriate to a master. | 15:26 |
sonne|work | ptizoom: yes I read that and it is true | 15:26 |
ptizoom | but, yet I would like to be endorsed by this community! | 15:27 |
sonne|work | however it does not hurt to add perl configure / make automagic and typemaps for perl | 15:27 |
sonne|work | the interface will be marked pre alpha or so anyways | 15:28 |
sonne|work | once some more examples work and we have no leaks we can go to beta :D | 15:28 |
ptizoom | hold the horses | 15:29 |
ptizoom | there is a chain of picks to climbs, and lots of crevasses on the way. | 15:29 |
ptizoom | the stuff I started is very rough, and I let lots of noise in the files, but the whole structure is there , I guess. | 15:32 |
ptizoom | and it needs more than one soul to complete... | 15:33 |
ptizoom | hence trying to advert it on a branch of your shogun-toolbox/perl... | 15:35 |
ptizoom | alternatively I made enough noise so you might direct other benevolence to the ptizoom/shogun/perl_swig_120924? | 15:36 |
ptizoom | here some example of work: | 15:41 |
ptizoom | make run-testsuite | 15:41 |
ptizoom | make -C src run-testsuite | 15:41 |
ptizoom | make[1]: Entering directory `/usr/src/shogun/src' | 15:41 |
ptizoom | make -C interfaces/perldl_modular run-testsuite-perldl_modular && true | 15:41 |
ptizoom | make[2]: Entering directory `/usr/src/shogun/src/interfaces/perldl_modular' | 15:41 |
ptizoom | cd ../../../tests/integration/perldl_modular && (\ | 15:41 |
ptizoom | LD_LIBRARY_PATH=/usr/local/lib PERLDL=pdl PERLDLPATH="/usr/local/lib/perl/5.14.2" \ | 15:41 |
ptizoom | LD_LIBRARY_PATH="/usr/src/shogun/src/interfaces/perldl_modular:/usr/src/shogun/src/shogun" \ | 15:41 |
ptizoom | ./test_all.sh ) | 15:41 |
ptizoom | #!/bin/bash -vx | 15:41 |
ptizoom | if test -z "$PERL" | 15:41 |
ptizoom | then | 15:41 |
ptizoom | PERL=perl | 15:41 |
ptizoom | fi | 15:41 |
ptizoom | + test -z '' | 15:41 |
ptizoom | + PERL=perl | 15:41 |
ptizoom | DATAPATH='../data' | 15:41 |
ptizoom | + DATAPATH=../data | 15:41 |
ptizoom | function test_all () { | 15:41 |
ptizoom | datapath="$1" | 15:41 |
ptizoom | if echo "$datapath" | grep -q '\.\./data/tests' | 15:41 |
ptizoom | then | 15:41 |
ptizoom | continue | 15:41 |
ptizoom | fi | 15:41 |
ptizoom | echo "*** Testing in $datapath" | 15:41 |
ptizoom | sleep 1 | 15:41 |
ptizoom | for file in $datapath; do | 15:41 |
ptizoom | echo -n "$file" | 15:41 |
ptizoom | echo -n -e "\t\t" | 15:41 |
ptizoom | output=`./test_one.pl "$file"` | 15:42 |
ptizoom | ret=$? | 15:42 |
ptizoom | if [ $ret -eq 0 ] ; then | 15:42 |
ptizoom | echo 'OK' | 15:42 |
ptizoom | else | 15:42 |
ptizoom | echo 'ERROR' | 15:42 |
ptizoom | echo $output | 15:42 |
ptizoom | fi | 15:42 |
ptizoom | done | 15:42 |
ptizoom | sleep 1 | 15:42 |
ptizoom | echo | 15:42 |
ptizoom | } | 15:42 |
ptizoom | if [ -n "$1" ]; then | 15:42 |
ptizoom | test_all "$DATAPATH/$1/*.m" | 15:42 |
ptizoom | else | 15:42 |
ptizoom | for i in $DATAPATH/*; do | 15:42 |
ptizoom | test_all "$i/*.m" | 15:42 |
ptizoom | done | 15:42 |
ptizoom | fi | 15:42 |
ptizoom | + '[' -n '' ']' | 15:42 |
ptizoom | + for i in '$DATAPATH/*' | 15:42 |
ptizoom | + test_all '../data/classifier/*.m' | 15:42 |
ptizoom | + datapath='../data/classifier/*.m' | 15:42 |
ptizoom | + echo '../data/classifier/*.m' | 15:42 |
ptizoom | + grep -q '\.\./data/tests' | 15:42 |
ptizoom | + echo '*** Testing in ../data/classifier/*.m' | 15:42 |
ptizoom | *** Testing in ../data/classifier/*.m | 15:42 |
ptizoom | + sleep 1 | 15:42 |
ptizoom | + for file in '$datapath' | 15:43 |
ptizoom | + echo -n ../data/classifier/LibLinear_0017_1en05_1_True.m | 15:43 |
ptizoom | ../data/classifier/LibLinear_0017_1en05_1_True.m+ echo -n -e '\t\t' | 15:43 |
ptizoom | ./test_one.pl "$file" | 15:43 |
ptizoom | ++ ./test_one.pl ../data/classifier/LibLinear_0017_1en05_1_True.m | 15:43 |
ptizoom | + output='LibLinear - yesclassified: 5.949422e-02, bias: 1.784163e-14 <--- accuracy: 1.000000e-04' | 15:43 |
ptizoom | + ret=0 | 15:43 |
ptizoom | + '[' 0 -eq 0 ']' | 15:43 |
ptizoom | + echo OK | 15:43 |
ptizoom | OK | 15:43 |
ptizoom | + for file in '$datapath' | 15:43 |
ptizoom | + echo -n ../data/classifier/LibLinear_023_1en05_1_True.m | 15:43 |
ptizoom | ../data/classifier/LibLinear_023_1en05_1_True.m+ echo -n -e '\t\t' | 15:43 |
ptizoom | ./test_one.pl "$file" | 15:43 |
ptizoom | ++ ./test_one.pl ../data/classifier/LibLinear_023_1en05_1_True.m | 15:43 |
ptizoom | + output='LibLinear - yesclassified: 3.048510e-01, bias: 1.043610e-14 <--- accuracy: 1.000000e-04' | 15:43 |
ptizoom | + ret=0 | 15:43 |
ptizoom | + '[' 0 -eq 0 ']' | 15:43 |
ptizoom | + echo OK | 15:43 |
ptizoom | OK | 15:43 |
ptizoom | + for file in '$datapath' | 15:43 |
ptizoom | + echo -n ../data/classifier/LibLinear_15_1en05_1_True.m | 15:43 |
ptizoom | ../data/classifier/LibLinear_15_1en05_1_True.m+ echo -n -e '\t\t' | 15:43 |
ptizoom | ./test_one.pl "$file" | 15:43 |
ptizoom | ++ ./test_one.pl ../data/classifier/LibLinear_15_1en05_1_True.m | 15:43 |
ptizoom | + output='LibLinear - yesclassified: 9.949554e-01, bias: 1.378421e-05 <--- accuracy: 1.000000e-04' | 15:43 |
ptizoom | + ret=0 | 15:43 |
ptizoom | + '[' 0 -eq 0 ']' | 15:43 |
ptizoom | + echo OK | 15:44 |
ptizoom | OK | 15:44 |
ptizoom | note:I think it is not converging because I coked up the transfer of vector into a sv_pv memory... (sigh) | 15:44 |
sonne|work | ptizoom: good enough to me. I will have a look | 15:48 |
ptizoom | check out swig_typemap.i, pdl.i | 15:49 |
ptizoom | anyway, | 15:52 |
wiking | eh this was a big spam :) | 15:53 |
ptizoom | sorry | 15:53 |
sonne|work | I will | 15:55 |
wiking | sonne|work: SGVector<T>::add(SGSParseVector<T> v) ? | 16:03 |
wiking | or at least we should have an SGVector(const SGSparseVector<T>& orig) copy constructor in SGVector | 16:04 |
sonne|work | both make sense! | 16:05 |
wiking | or of course SGVector<T> SGSParseVector::construct_dense_vector () const; function | 16:05 |
wiking | hehe yeah | 16:06 |
wiking | ok will extend them | 16:06 |
sonne|work | just do it :D | 16:06 |
wiking | doh | 16:24 |
wiking | note: candidate function not viable: no known conversion from 'SGSparseVector<float64_t> *' to | 16:24 |
wiking | 'const SGSparseVector<double>' for 1st argument; dereference the argument with * | 16:24 |
wiking | void add(const SGSparseVector<T> x); | 16:24 |
wiking | ah ok my bad | 16:25 |
wiking | ah ok my bad | 16:25 |
-!- blackburn [~blackburn@37.61.180.250] has joined #shogun | 17:29 | |
blackburn | ptizoom: you appear to be so poetic in your explanations :) | 18:06 |
-!- heiko [~heiko@host86-176-125-28.range86-176.btcentralplus.com] has left #shogun [] | 19:25 | |
ptizoom | nothing worse than failing the turing test... | 19:45 |
--- Log closed Thu Oct 11 00:00:17 2012 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!