--- Log opened Tue Jun 12 00:00:41 2012 | ||
-!- blackburn [~blackburn@188.168.3.9] has quit [Ping timeout: 248 seconds] | 00:10 | |
-!- wiking_ [~wiking@ip191.67-202-72.static.steadfastdns.net] has joined #shogun | 00:12 | |
-!- wiking_ [~wiking@ip191.67-202-72.static.steadfastdns.net] has quit [Changing host] | 00:12 | |
-!- wiking_ [~wiking@huwico/staff/wiking] has joined #shogun | 00:12 | |
-!- wiking [~wiking@huwico/staff/wiking] has quit [Ping timeout: 245 seconds] | 00:15 | |
-!- wiking_ is now known as wiking | 00:15 | |
-!- shogun-buildbot [~shogun-bu@7nn.de] has quit [Ping timeout: 248 seconds] | 00:20 | |
-!- shogun-buildbot_ [~shogun-bu@7nn.de] has joined #shogun | 00:29 | |
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving] | 00:36 | |
-!- emrecelikten [~emrecelik@176.40.238.218] has joined #shogun | 01:11 | |
-!- emrecelikten [~emrecelik@176.40.238.218] has quit [Remote host closed the connection] | 02:28 | |
-!- emrecelikten [~emrecelik@176.40.238.218] has joined #shogun | 02:28 | |
-!- emrecelikten [~emrecelik@176.40.238.218] has quit [Remote host closed the connection] | 02:33 | |
-!- emrecelikten [~emrecelik@176.40.238.218] has joined #shogun | 02:34 | |
-!- av3ngr [av3ngr@nat/redhat/x-whqkmpsbmkcikpnd] has joined #shogun | 04:23 | |
-!- av3ngr [av3ngr@nat/redhat/x-whqkmpsbmkcikpnd] has quit [Read error: Connection reset by peer] | 04:23 | |
-!- vikram360 [~vikram360@117.192.161.30] has quit [Ping timeout: 265 seconds] | 05:52 | |
-!- vikram360 [~vikram360@117.192.172.86] has joined #shogun | 05:52 | |
-!- wiking_ [~wiking@huwico/staff/wiking] has joined #shogun | 06:43 | |
-!- wiking [~wiking@huwico/staff/wiking] has quit [Ping timeout: 252 seconds] | 06:44 | |
-!- wiking_ is now known as wiking | 06:44 | |
-!- vikram360 [~vikram360@117.192.172.86] has quit [Ping timeout: 244 seconds] | 06:49 | |
-!- vikram360 [~vikram360@117.192.188.102] has joined #shogun | 07:18 | |
-!- vikram360 [~vikram360@117.192.188.102] has quit [Ping timeout: 240 seconds] | 07:29 | |
-!- uricamic [~uricamic@2001:718:2:1634:38f0:17b8:8038:3f25] has joined #shogun | 08:54 | |
-!- vikram360 [~vikram360@117.192.170.84] has joined #shogun | 09:28 | |
-!- emrecelikten [~emrecelik@176.40.238.218] has quit [Ping timeout: 252 seconds] | 09:31 | |
-!- wiking [~wiking@huwico/staff/wiking] has quit [Quit: wiking] | 11:48 | |
-!- wiking [~wiking@huwico/staff/wiking] has joined #shogun | 12:17 | |
-!- blackburn [~blackburn@188.168.3.9] has joined #shogun | 12:23 | |
-!- vikram360 [~vikram360@117.192.170.84] has quit [Ping timeout: 246 seconds] | 12:50 | |
-!- vikram360 [~vikram360@117.192.168.95] has joined #shogun | 12:51 | |
-!- wiking_ [~wiking@68.68.31.85.customer.bluemilenetworks.com] has joined #shogun | 13:04 | |
-!- wiking_ [~wiking@68.68.31.85.customer.bluemilenetworks.com] has quit [Changing host] | 13:04 | |
-!- wiking_ [~wiking@huwico/staff/wiking] has joined #shogun | 13:04 | |
-!- wiking [~wiking@huwico/staff/wiking] has quit [Ping timeout: 246 seconds] | 13:05 | |
-!- wiking_ is now known as wiking | 13:05 | |
-!- wiking_ [~wiking@78-23-189-112.access.telenet.be] has joined #shogun | 13:34 | |
-!- wiking_ [~wiking@78-23-189-112.access.telenet.be] has quit [Changing host] | 13:34 | |
-!- wiking_ [~wiking@huwico/staff/wiking] has joined #shogun | 13:34 | |
-!- wiking [~wiking@huwico/staff/wiking] has quit [Ping timeout: 252 seconds] | 13:34 | |
-!- wiking_ is now known as wiking | 13:34 | |
-!- romi_ [~mizobe@187.74.0.229] has joined #shogun | 13:50 | |
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 14:13 | |
-!- emrecelikten [~emrecelik@176.40.238.218] has joined #shogun | 14:29 | |
-!- nicococo [~nico@lacedcoffee.ml.tu-berlin.de] has joined #shogun | 14:31 | |
n4nd0 | nicococo: hallo, how are you? | 15:08 |
---|---|---|
nicococo | hi, fine. what about you? | 15:09 |
n4nd0 | I am good | 15:10 |
n4nd0 | I got a new mosek license :D | 15:10 |
nicococo | very good .. now its working?? | 15:11 |
n4nd0 | yeah, I can solve "big" problems again | 15:11 |
nicococo | uuuhhh.. and the mc problem is solved therefore | 15:11 |
nicococo | :) | 15:11 |
n4nd0 | yeah | 15:12 |
nicococo | fantastic! | 15:12 |
n4nd0 | I have yet to checked that the results are reliable enough | 15:12 |
n4nd0 | still better than yet :P | 15:12 |
nicococo | ok ;) i think we then have to talk about the next steps right | 15:13 |
nicococo | i mean the big picture steps | 15:13 |
n4nd0 | yeah, we can talk a bit about it right now if you want | 15:13 |
n4nd0 | I will have to go in a while though | 15:13 |
nicococo | yes, because i want to go on vacation next month (for about 1 month) and would like to keep you busy ;) | 15:14 |
n4nd0 | ok | 15:14 |
n4nd0 | so hm-svm? | 15:14 |
nicococo | in your application was a timetable that states: hmsvm and crf | 15:14 |
n4nd0 | yes | 15:15 |
n4nd0 | what do you think we should go for first? | 15:15 |
nicococo | honestly, i don't know how much we can deviate from these goals (i have to ask) but i think the main goal is already achieved and there are a couple of new things. | 15:15 |
n4nd0 | I would say hm-svm | 15:15 |
n4nd0 | what are these new things? | 15:16 |
n4nd0 | the optimizer? | 15:16 |
nicococo | like: coffin, opensource solver | 15:16 |
n4nd0 | yes | 15:16 |
nicococo | what are your preferences? | 15:16 |
n4nd0 | but I still think that hm-svm is important | 15:16 |
n4nd0 | I won't be pleased until we have done that :) | 15:16 |
nicococo | yes, and this will take a long time (can be quite tricky) | 15:17 |
n4nd0 | I would say then hm-svm, coffin, solver and crf; in that order | 15:17 |
n4nd0 | oh | 15:17 |
nicococo | we can combine it with coffin.. | 15:17 |
n4nd0 | what are the tricky points with hm-svm? | 15:17 |
nicococo | its a whole bunch of little things that are not exactly hard to solve but not easy either | 15:18 |
n4nd0 | mmm | 15:18 |
n4nd0 | can you tell me a bit more in particular? :) | 15:19 |
n4nd0 | or is it better to wait? | 15:19 |
nicococo | okay what do you think about combining hm-svm with coffin?? | 15:19 |
n4nd0 | while I have more or less an idea how hm-svm would fit in the framework | 15:20 |
n4nd0 | I don't have it for coffin | 15:20 |
n4nd0 | so I am not sure about the answer | 15:20 |
n4nd0 | I understand that hm-svm should derive from CStructuredModel | 15:21 |
nicococo | in my opinion combination with coffin would be 1) a new thing, 2) maybe easier for the task i am thinking of (procaryote gene prediction) | 15:21 |
n4nd0 | it may be tricky as you say but I have clear the idea that we "just" need an argmax function, a psi and a delta loss | 15:21 |
nicococo | mmhhh.. sorry.. there is more :) | 15:22 |
n4nd0 | tell me | 15:22 |
nicococo | ever heard of plif-functions? | 15:22 |
n4nd0 | something about the labels too | 15:22 |
n4nd0 | I have seen something around in shogun yes | 15:22 |
n4nd0 | but I have no idea what they are | 15:22 |
nicococo | and you also need a state model :) | 15:22 |
n4nd0 | ? | 15:23 |
nicococo | okay, there are two possible things: | 15:23 |
nicococo | a) the one i told you two minutes ago | 15:24 |
nicococo | b) building the hmsvm matlab framework | 15:24 |
n4nd0 | what do you think? | 15:25 |
n4nd0 | combination with coffin? | 15:25 |
nicococo | i mean, you already had a look at the hmsvm-code, right? there are 30 files only for hm-svm :) its much more than argmax,delta | 15:25 |
n4nd0 | yeah | 15:25 |
n4nd0 | but the article gave me that idea :S | 15:26 |
nicococo | theory-application gap ;) | 15:27 |
n4nd0 | too bad | 15:27 |
n4nd0 | :D | 15:27 |
n4nd0 | anyhow | 15:28 |
n4nd0 | let me know your opinion about each of the two possibilities :) | 15:28 |
nicococo | i would like to go with the coffin-hmsvm.. | 15:28 |
n4nd0 | I am ok with that | 15:29 |
n4nd0 | but tell me why please ;) | 15:29 |
n4nd0 | it's not a why meaning that I would like the other | 15:29 |
n4nd0 | it's a why to gain understanding :) | 15:29 |
nicococo | it is a new thing.. first step towards real-large scale (high dim) problems in so | 15:30 |
nicococo | and much easier to implement in the basic version :)) | 15:30 |
n4nd0 | sounds pretty nice to me | 15:30 |
n4nd0 | all advantages :D | 15:30 |
n4nd0 | do you have any application in mind for it? | 15:30 |
nicococo | yes, i also know some people who can give you a hand when i am on vacation | 15:31 |
nicococo | so, don't worry | 15:31 |
n4nd0 | that's great | 15:31 |
nicococo | we code an gene predictor for simple procaryote organism (like bacteria) | 15:32 |
n4nd0 | no idea what's that :P | 15:32 |
nicococo | .. with real data | 15:32 |
nicococo | no toy data anymore ;) | 15:33 |
n4nd0 | yeah | 15:33 |
nicococo | okay, but first, the toy data mc-experiment | 15:33 |
n4nd0 | sure | 15:33 |
nicococo | can you test the coffin thing with that? | 15:33 |
n4nd0 | we could do that | 15:33 |
n4nd0 | but the coffind thing may take time | 15:34 |
n4nd0 | coffin | 15:34 |
nicococo | no prob | 15:34 |
n4nd0 | I have to read a bit more about it and probably ask sonney2k | 15:34 |
nicococo | did you already checked in the code for this exp? | 15:34 |
nicococo | experiment | 15:34 |
n4nd0 | for the mc example? | 15:34 |
nicococo | yes | 15:35 |
n4nd0 | not yet | 15:35 |
n4nd0 | but I am able to do it in python now | 15:35 |
n4nd0 | sonney2k found the part I was missing | 15:35 |
nicococo | super! (that was my next question) | 15:35 |
nicococo | what was the problem? | 15:35 |
n4nd0 | so I will commit the example this evening probably | 15:35 |
n4nd0 | we have to include "class CStructuredModel" and the same for CLossFunction in the header of StructuredMachine | 15:36 |
n4nd0 | it looks like a workaround for making it work with SWIG | 15:36 |
nicococo | one thing less to worry about | 15:37 |
n4nd0 | yes | 15:37 |
nicococo | so, can we meet again on thursday and you put all the stuff in git and try to get coffin working with the example.. | 15:39 |
n4nd0 | ok, let's do that | 15:39 |
nicococo | its a bit short on time maybe.. but just have a look | 15:39 |
n4nd0 | I will do my best :) | 15:39 |
nicococo | ;) | 15:40 |
nicococo | ok. then see you on thursday.. don't work too much :) | 15:40 |
n4nd0 | haha ok | 15:41 |
n4nd0 | bye | 15:41 |
-!- nicococo [~nico@lacedcoffee.ml.tu-berlin.de] has left #shogun [] | 15:45 | |
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Ping timeout: 240 seconds] | 15:50 | |
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 16:35 | |
-!- heiko1 [~heiko@host86-181-81-253.range86-181.btcentralplus.com] has joined #shogun | 16:37 | |
-!- vikram360 [~vikram360@117.192.168.95] has quit [Ping timeout: 244 seconds] | 16:42 | |
emrecelikten | http://lwn.net/Articles/501536/ | 16:43 |
emrecelikten | You guys should see this :D | 16:43 |
blackburn | hahaha | 16:44 |
n4nd0 | :D | 16:45 |
wiking | :>> | 16:50 |
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Ping timeout: 244 seconds] | 16:52 | |
emrecelikten | I should mail this to Stallman | 16:52 |
-!- wiking_ [~wiking@208.76.55.196] has joined #shogun | 16:54 | |
-!- wiking_ [~wiking@208.76.55.196] has quit [Changing host] | 16:54 | |
-!- wiking_ [~wiking@huwico/staff/wiking] has joined #shogun | 16:54 | |
-!- uricamic [~uricamic@2001:718:2:1634:38f0:17b8:8038:3f25] has quit [Quit: Leaving.] | 16:55 | |
wiking_ | emrecelikten: http://www.theregister.co.uk/2012/06/12/stallman_laptop_stolen/ | 16:55 |
emrecelikten | wiking_: Yep, heard it the day it happened. He was crying | 16:56 |
-!- wiking [~wiking@huwico/staff/wiking] has quit [Ping timeout: 244 seconds] | 16:56 | |
-!- wiking_ is now known as wiking | 16:56 | |
-!- puffin444 [62e3926e@gateway/web/freenode/ip.98.227.146.110] has joined #shogun | 16:59 | |
emrecelikten | wiking: All of that config gone, that must have hurt | 17:02 |
wiking | backup :) | 17:02 |
emrecelikten | Even the OS? | 17:03 |
emrecelikten | I should backup though, I would be demolished if something like that happened to me | 17:06 |
emrecelikten | Thanks for reminding :) | 17:06 |
-!- blackburn [~blackburn@188.168.3.9] has quit [Ping timeout: 252 seconds] | 17:14 | |
-!- wiking [~wiking@huwico/staff/wiking] has quit [Ping timeout: 245 seconds] | 17:32 | |
-!- wiking [~wiking@78-23-189-112.access.telenet.be] has joined #shogun | 17:34 | |
-!- wiking [~wiking@78-23-189-112.access.telenet.be] has quit [Changing host] | 17:34 | |
-!- wiking [~wiking@huwico/staff/wiking] has joined #shogun | 17:34 | |
-!- puffin444 [62e3926e@gateway/web/freenode/ip.98.227.146.110] has quit [Ping timeout: 245 seconds] | 17:52 | |
-!- puffin444 [62e3926e@gateway/web/freenode/ip.98.227.146.110] has joined #shogun | 17:58 | |
-!- vikram360 [~vikram360@117.192.160.190] has joined #shogun | 18:08 | |
-!- vikram360 [~vikram360@117.192.160.190] has quit [Ping timeout: 244 seconds] | 18:21 | |
-!- vikram360 [~vikram360@117.192.176.105] has joined #shogun | 18:21 | |
-!- vikram360 [~vikram360@117.192.176.105] has quit [Ping timeout: 244 seconds] | 18:26 | |
-!- vikram360 [~vikram360@117.192.170.52] has joined #shogun | 18:26 | |
-!- zxtx [~zv@cpe-75-83-151-252.socal.res.rr.com] has quit [Ping timeout: 245 seconds] | 18:44 | |
-!- KMcQuisten [d8338942@gateway/web/freenode/ip.216.51.137.66] has joined #shogun | 18:51 | |
KMcQuisten | hey all. Has gsomix been around? I'm having some trouble still with DirectorKernels | 18:52 |
shogun-buildbot_ | build #606 of csharp_modular is complete: Failure [failed compile] Build details are at http://www.shogun-toolbox.org/buildbot/builders/csharp_modular/builds/606 blamelist: sonne@debian.org, heiko.strathmann@gmail.com | 18:54 |
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 19:03 | |
shogun-buildbot_ | build #590 of python_modular is complete: Failure [failed test_1] Build details are at http://www.shogun-toolbox.org/buildbot/builders/python_modular/builds/590 blamelist: sonne@debian.org, heiko.strathmann@gmail.com | 19:22 |
CIA-18 | shogun: Heiko Strathmann master * r86919c0 / (7 files in 4 dirs): Merge pull request #581 from karlnapf/master (+8 more commits...) - http://git.io/O7Pfyw | 19:33 |
-!- heiko1 [~heiko@host86-181-81-253.range86-181.btcentralplus.com] has quit [Ping timeout: 244 seconds] | 19:39 | |
-!- KMcQuisten [d8338942@gateway/web/freenode/ip.216.51.137.66] has quit [Quit: Page closed] | 19:41 | |
emrecelikten | Hahaha blamelist, never noticed it | 19:43 |
emrecelikten | Nice | 19:43 |
-!- wiking [~wiking@huwico/staff/wiking] has quit [Quit: wiking] | 19:57 | |
-!- KMcQuisten [d8338942@gateway/web/freenode/ip.216.51.137.66] has joined #shogun | 20:01 | |
-!- K0stIa [~kostia@2001:718:2:1634:76f0:6dff:fe92:4013] has joined #shogun | 20:05 | |
-!- K0stIa [~kostia@2001:718:2:1634:76f0:6dff:fe92:4013] has left #shogun [] | 20:05 | |
-!- K0stIa [~kostia@2001:718:2:1634:76f0:6dff:fe92:4013] has joined #shogun | 20:05 | |
-!- K0stIa [~kostia@2001:718:2:1634:76f0:6dff:fe92:4013] has left #shogun [] | 20:05 | |
puffin444 | Hey has anyone here had experience with clapack_dpotrf? | 20:08 |
@sonney2k | KMcQuisten, what troubles? | 20:09 |
-!- wiking [~wiking@we02c096.ugent.be] has joined #shogun | 20:09 | |
-!- wiking [~wiking@we02c096.ugent.be] has quit [Changing host] | 20:09 | |
-!- wiking [~wiking@huwico/staff/wiking] has joined #shogun | 20:09 | |
KMcQuisten | Just posted a description of it to the mailing list. | 20:10 |
KMcQuisten | I made a trivial DircetorKernel for testing, provided it to LibSVR, and I get lots of garbage collection and segfault errors | 20:10 |
@sonney2k | KMcQuisten, how many CPU cores does your machine have? | 20:11 |
KMcQuisten | 2 | 20:11 |
@sonney2k | set number of threads to 1 and try again | 20:11 |
-!- emrecelikten [~emrecelik@176.40.238.218] has quit [Ping timeout: 265 seconds] | 20:15 | |
@sonney2k | KMcQuisten, did that help? | 20:15 |
KMcQuisten | Seemed to. My nontrivial example worked as well | 20:15 |
KMcQuisten | Shame I can't seem to take advantage of multicore. | 20:16 |
KMcQuisten | Oh well. at least it works | 20:16 |
KMcQuisten | thanks | 20:17 |
-!- gsomix [~gsomix@188.168.5.114] has joined #shogun | 20:18 | |
gsomix | hi all | 20:19 |
KMcQuisten | Hey there | 20:19 |
@sonney2k | KMcQuisten, reason is that python doesn't support threads | 20:19 |
@sonney2k | you know the GIL (global interpreter lock...) | 20:19 |
@sonney2k | so no way | 20:19 |
gsomix | It seems that all problems have decided on without me. :) Nice. | 20:19 |
KMcQuisten | So the builtins can leverage parallel, but there's no way to make directorkernels work like that? | 20:20 |
@sonney2k | KMcQuisten, please post the answer to the mailinglist - and give us some hint how we can warn the user ... | 20:20 |
@sonney2k | KMcQuisten, yes no way | 20:20 |
KMcQuisten | Hmm. | 20:21 |
KMcQuisten | i wonder if the multiprocessing package could be used to help this. | 20:22 |
@sonney2k | no | 20:22 |
KMcQuisten | tried it? | 20:22 |
@sonney2k | as I said it is a limitation of python | 20:22 |
@sonney2k | at some europython conference I talked to guido the BDL about this ... no way | 20:23 |
-!- wiking [~wiking@huwico/staff/wiking] has quit [Ping timeout: 265 seconds] | 20:28 | |
-!- wiking [~wiking@78-23-189-112.access.telenet.be] has joined #shogun | 20:44 | |
-!- wiking [~wiking@78-23-189-112.access.telenet.be] has quit [Changing host] | 20:44 | |
-!- wiking [~wiking@huwico/staff/wiking] has joined #shogun | 20:44 | |
-!- puffin444 [62e3926e@gateway/web/freenode/ip.98.227.146.110] has quit [Ping timeout: 245 seconds] | 21:05 | |
-!- KMcQuisten [d8338942@gateway/web/freenode/ip.216.51.137.66] has quit [Ping timeout: 245 seconds] | 21:06 | |
n4nd0 | sonney2k: hey! | 21:54 |
n4nd0 | about wiking and michal's pull requests | 21:54 |
n4nd0 | I think it could be a good idea if they use the latest code of the SO framework | 21:54 |
n4nd0 | it has changed a bit from the last PR you merged | 21:54 |
shogun-buildbot_ | build #608 of csharp_modular is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/csharp_modular/builds/608 | 22:09 |
@sonney2k | n4nd0, sure | 22:26 |
@sonney2k | they should | 22:26 |
@sonney2k | problem really is that when they wait longer they have to merge and merge over and over again with all the changes coming... | 22:26 |
n4nd0 | yeah, I understand | 22:35 |
shogun-buildbot_ | build #596 of java_modular is complete: Failure [failed compile] Build details are at http://www.shogun-toolbox.org/buildbot/builders/java_modular/builds/596 blamelist: blackburn91@gmail.com, pluskid@gmail.com, sonne@debian.org, gsomix@gmail.com, heiko.strathmann@gmail.com | 22:58 |
gsomix | wow | 22:58 |
n4nd0 | wiking: we are not in the blamelist! :D | 22:59 |
-!- wiking_ [~wiking@208.76.55.195] has joined #shogun | 23:03 | |
-!- wiking_ [~wiking@208.76.55.195] has quit [Changing host] | 23:03 | |
-!- wiking_ [~wiking@huwico/staff/wiking] has joined #shogun | 23:03 | |
-!- wiking [~wiking@huwico/staff/wiking] has quit [Ping timeout: 265 seconds] | 23:04 | |
-!- wiking_ is now known as wiking | 23:04 | |
shogun-buildbot_ | build #588 of ruby_modular is complete: Failure [failed test_1] Build details are at http://www.shogun-toolbox.org/buildbot/builders/ruby_modular/builds/588 blamelist: blackburn91@gmail.com, pluskid@gmail.com, sonne@debian.org, gsomix@gmail.com, heiko.strathmann@gmail.com | 23:34 |
-!- vikram360 [~vikram360@117.192.170.52] has quit [Ping timeout: 265 seconds] | 23:55 | |
--- Log closed Wed Jun 13 00:00:41 2012 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!