--- Log opened Sat Jul 07 00:00:17 2012 | ||
wiking | it wasn't like this | 00:00 |
---|---|---|
wiking | but now after my rebase suddenly it is | 00:00 |
@sonney2k | wiking, actually there are two methods to get the size: one is the real but *internal* size | 00:00 |
wiking | i'm strictly talking about get_num_elements | 00:01 |
@sonney2k | and the other is the currently used size (which is to be used) | 00:01 |
wiking | i've seen that there's another api for it | 00:01 |
wiking | i'm guessing get_num_elements is the currently used size? | 00:01 |
@sonney2k | wiking, I think it is correct *now* | 00:01 |
wiking | ok | 00:02 |
@sonney2k | hmmhh | 00:02 |
@sonney2k | not 100% sure | 00:02 |
gsomix | there are get_num_elements and get_array_size | 00:03 |
@sonney2k | gsomix, yeah but what he does is (I guess) call CDynamicObjectArray(10) | 00:03 |
wiking | yep | 00:03 |
@sonney2k | taht should create an array of size 10 | 00:03 |
@sonney2k | granularity doesn't matter here | 00:04 |
wiking | and if i do: CDynamicObjectArray w (10); w->get_num_elements (); | 00:04 |
wiking | i'll get 0 | 00:04 |
@sonney2k | so I think get_num_elements should return 10 ? | 00:04 |
wiking | atm it works like this | 00:04 |
gsomix | try get_array_size | 00:04 |
@sonney2k | gsomix, but that is the wrong function | 00:04 |
gsomix | because CDynamicObjectArray(10) set resize_granularity | 00:04 |
@sonney2k | but maybe it returns 0 too | 00:04 |
@sonney2k | gsomix, where does it do this? | 00:05 |
@sonney2k | (I cannot see it) | 00:05 |
wiking | ok i'll check it now | 00:05 |
@sonney2k | gsomix, ahh in the m_array call | 00:06 |
wiking | [DEBUG] size of obj array: 77 0^C | 00:07 |
@sonney2k | gsomix, I just checked - indeed it just sets resize granularity | 00:07 |
wiking | SG_DEBUG ("size of obj array: %d %d", m_samples->get_array_size (), m_samples->get_num_elements ()); | 00:07 |
wiking | for this | 00:07 |
@sonney2k | not the size of the array | 00:07 |
wiking | gsomix: ^ so get_array_size works | 00:08 |
gsomix | sonney2k, hey no | 00:08 |
gsomix | num_elements=p_resize_granularity; | 00:08 |
gsomix | last_element_idx=-1; | 00:08 |
@sonney2k | gsomix, I meant only granularity | 00:08 |
@sonney2k | it sets only granularity | 00:08 |
@sonney2k | never the last element | 00:08 |
@sonney2k | wiking, so it is not what you want here! | 00:08 |
gsomix | sonney2k, yes. for getting full size I add get_array_size | 00:09 |
wiking | ok | 00:09 |
wiking | i see now i've changed everything to get_array_size | 00:09 |
gsomix | btw it's documented | 00:09 |
@sonney2k | gsomix, shouldn't wiking call get_dim1() ? | 00:09 |
blackburn | sonney2k: do you know any convenient way to get randomly sample training and test indices in python? | 00:10 |
@sonney2k | I think that is what he wants if this array is not resized | 00:10 |
gsomix | sonney2k, mmm, it works too, because ndims is stored | 00:10 |
@sonney2k | blackburn, randperm? | 00:10 |
blackburn | sonney2k: no I need a set of training indexes and test ones too | 00:10 |
gsomix | *are stored | 00:11 |
@sonney2k | permutation(10) | 00:11 |
@sonney2k | a=permutation(10) | 00:11 |
blackburn | ah | 00:11 |
@sonney2k | tr_idx=a[:5] | 00:11 |
blackburn | permute and cut? | 00:11 |
@sonney2k | te_idx=a[5:] | 00:11 |
blackburn | cool | 00:11 |
blackburn | thanks | 00:11 |
@sonney2k | shogun-buildbot, status | 00:11 |
shogun-buildbot | cyg1 - libshogun: idle | 00:11 |
shogun-buildbot | deb1 - libshogun: idle, last build 2h02m41s ago: build successful | 00:11 |
shogun-buildbot | deb2 - static_interfaces: idle, last build 1h19m13s ago: build successful | 00:11 |
shogun-buildbot | deb3 - modular_interfaces: idle, last build 54m42s ago: failed test python_modular | 00:11 |
shogun-buildbot | nightly_all: idle, last build 20h19m23s ago: build successful | 00:11 |
shogun-buildbot | nightly_default: idle, last build 20h33m14s ago: build successful | 00:11 |
shogun-buildbot | nightly_none: idle, last build 20h21m14s ago: build successful | 00:11 |
@sonney2k | shogun-buildbot, force build "cyg1 - libshogun" | 00:12 |
shogun-buildbot | build #0 of cyg1 - libshogun started, including [] | 00:12 |
shogun-buildbot | build #0 forced | 00:12 |
shogun-buildbot | I'll give a shout when the build finishes | 00:12 |
@sonney2k | karamba | 00:12 |
gsomix | sonney2k, dynarray and array1/2/3 had different interfaces... I got the smallest patches. :( | 00:13 |
gsomix | *changes | 00:13 |
@sonney2k | gsomix, yeah you are right - one was basically fixed size | 00:14 |
@sonney2k | dynarray's size can change | 00:14 |
@sonney2k | and then things like dim1 ... dim3 don't make sense | 00:14 |
@sonney2k | ohh damn personal firewalls... | 00:15 |
@sonney2k | How is one supposed to add exceptions on a *buildslave* | 00:16 |
gsomix | good night guys | 00:18 |
@sonney2k | 180M and cloning... | 00:19 |
@sonney2k | gsomix, nite and sweet python buffer dreams :D | 00:19 |
gsomix | huh, dreams w/o copying | 00:19 |
@sonney2k | hehe | 00:20 |
@sonney2k | no overhead in dreams please! | 00:20 |
@sonney2k | blackburn, look at http://www.shogun-toolbox.org/buildbot/waterfall | 00:20 |
@sonney2k | cyg1 is progressing... | 00:21 |
blackburn | heh | 00:21 |
@sonney2k | wiking, you don't happen to have some spare osx machines? | 00:21 |
@sonney2k | maybe I should ask gunnar... | 00:21 |
wiking | sonney2k: here | 00:22 |
wiking | sitting on it | 00:22 |
wiking | u need for buildbot? | 00:22 |
shogun-buildbot | build #0 of cyg1 - libshogun is complete: Failure [failed compile] Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/0 | 00:23 |
@sonney2k | bahh | 00:23 |
@sonney2k | fail you! | 00:23 |
@sonney2k | the usual path problem | 00:24 |
-!- gsomix [~gsomix@109.169.225.237] has quit [Ping timeout: 252 seconds] | 00:24 | |
@sonney2k | shogun-buildbot, force build "cyg1 - libshogun" | 00:26 |
shogun-buildbot | build #1 of cyg1 - libshogun started, including [] | 00:26 |
shogun-buildbot | build #1 forced | 00:26 |
shogun-buildbot | I'll give a shout when the build finishes | 00:26 |
wiking | sonney2k: ? | 00:26 |
@sonney2k | wiking, yeah otherwise we will never realize that we cause osx failures | 00:27 |
@sonney2k | wiking, hardcore admin action going on behind the curtains :D | 00:27 |
wiking | ok | 00:27 |
wiking | what do u need | 00:28 |
wiking | and i'll get it installed on that machine | 00:28 |
@sonney2k | wiking, shogun build dependencies and buildslave from buildbot.net | 00:28 |
wiking | ok | 00:29 |
@sonney2k | but it will build often... | 00:29 |
wiking | no worries | 00:29 |
@sonney2k | ok | 00:29 |
wiking | it's a quite good mac | 00:29 |
wiking | so it can take it | 00:29 |
@sonney2k | thats what you say now | 00:29 |
@sonney2k | it will build every commit... | 00:29 |
wiking | heheh yeah | 00:30 |
wiking | but that's ok | 00:30 |
@sonney2k | for all interfaces... | 00:30 |
@sonney2k | ok then :D | 00:30 |
wiking | just let me find the person who can give me sudo on the machine | 00:30 |
@sonney2k | good 100% load on the windows machine | 00:30 |
wiking | there's no parallel build or? | 00:30 |
wiking | so afais it'll just use one core | 00:31 |
@sonney2k | I can disable the parallel build... | 00:31 |
@sonney2k | (the linux machine uses 4 cores) | 00:31 |
wiking | ok | 00:31 |
wiking | any objections for adding virtual float64_t compute_primal_objective () const function to CLinearMachine? with default SG_NOTIMPLEMENTED by default? | 00:33 |
shogun-buildbot | build #1 of cyg1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/1 | 00:36 |
@sonney2k | feel free | 00:41 |
wiking | PR | 00:42 |
wiking | ed | 00:42 |
blackburn | HOG_SIZE 1488 :D | 00:43 |
wiking | blackburn: came from the example... | 00:43 |
wiking | but yeah start flaming | 00:43 |
blackburn | nice we have something from your project now | 00:44 |
wiking | well if u pull this | 00:45 |
wiking | then i can push the rest | 00:45 |
blackburn | sonney2k: I'll let you take care of this ;) | 00:45 |
blackburn | wiking: do you use precomputed hogs out there? | 00:46 |
wiking | blackburn: yep | 00:46 |
@sonney2k | wiking, well I can only check this superficially not whether it works | 01:11 |
@sonney2k | so huge patch | 01:11 |
@sonney2k | what is really missing is some kind of standalone example | 01:11 |
@sonney2k | wiking, anyways from my side mostly only some minor reformatting comments | 01:12 |
@sonney2k | so alexlovesdata, blackburn have a look too :) | 01:12 |
@sonney2k | wiking, my suggestion is you fix these issues and then I merge :D | 01:12 |
@sonney2k | (ahh and you told me about the function ptrs so I didn't complain here waiting for the updated patch :) | 01:13 |
blackburn | sonney2k: sure just after calendas graecas ;) | 01:13 |
wiking | sonney2k: | 01:14 |
wiking | i always did foo () instead of foo() | 01:14 |
wiking | see my prev patches | 01:14 |
@sonney2k | well please do foo() | 01:14 |
wiking | sonney2k: onces this pulled i can poor in the rest of the patches | 01:14 |
wiking | i just thought i dont want it in on pull | 01:14 |
wiking | i mean you dont want that | 01:15 |
wiking | :P | 01:15 |
@sonney2k | yeah it is big enough already | 01:15 |
wiking | btw | 01:15 |
wiking | is there a reason why u dont want dynamic_cast ? | 01:15 |
wiking | i mean can i at least do that in obtain_from_generic ? | 01:15 |
@sonney2k | I guess grepping for '[a-zA-Z]* (' will find it | 01:15 |
@sonney2k | wiking, you know the type so no need to use it | 01:16 |
wiking | okey | 01:16 |
@sonney2k | (there is an enum already) | 01:16 |
@sonney2k | my comment was only concerned that you don't need to do type checking all the time if it is in some extra function - though in this case no dynamic cast is needed | 01:17 |
@sonney2k | hmmhh thunderstorm arrives in berlin in the morning http://www.wetteronline.de/radarprognose.htm | 01:17 |
@sonney2k | let me better sleep now | 01:17 |
-!- nietpiet [92329056@gateway/web/freenode/ip.146.50.144.86] has quit [Quit: Page closed] | 01:19 | |
* sonney2k ZZZZzzzz.... | 01:21 | |
* wiking away now... pushed the commits for the requested changes | 02:10 | |
alexlovesdata | I will check this on monday | 02:22 |
alexlovesdata | ok ? | 02:23 |
alexlovesdata | need to sleep now ... | 02:23 |
-!- blackburn [~blackburn@81.28.187.0] has quit [Quit: Leaving.] | 02:25 | |
shogun-buildbot | build #10 of nightly_all started, including [] | 03:00 |
shogun-buildbot | build #15 of nightly_default started, including [] | 03:00 |
-!- alexlovesdata [55b2212c@gateway/web/freenode/ip.85.178.33.44] has quit [Quit: Page closed] | 03:13 | |
shogun-buildbot | build #10 of nightly_all is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/nightly_all/builds/10 | 03:23 |
shogun-buildbot | build #13 of nightly_none started, including [] | 03:23 |
shogun-buildbot | build #15 of nightly_default is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/nightly_default/builds/15 | 03:31 |
shogun-buildbot | build #13 of nightly_none is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/nightly_none/builds/13 | 03:36 |
-!- puffin444 [62e3926e@gateway/web/freenode/ip.98.227.146.110] has joined #shogun | 04:58 | |
puffin444 | Hey sonney2k, I know you're probably not awake, but I fixed those memory errors. | 05:00 |
-!- puffin444 [62e3926e@gateway/web/freenode/ip.98.227.146.110] has quit [Ping timeout: 245 seconds] | 05:59 | |
-!- pluskid [~pluskid@111.120.71.247] has joined #shogun | 08:23 | |
-!- gsomix [~gsomix@109.169.227.24] has joined #shogun | 09:26 | |
gsomix | good morning | 09:26 |
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 10:52 | |
gsomix | n4nd0, hey, how are you? | 11:08 |
n4nd0 | hey gsomix! I am fine, what about you? | 11:08 |
-!- gsomix [~gsomix@109.169.227.24] has quit [Ping timeout: 244 seconds] | 11:16 | |
-!- gsomix [~gsomix@109.169.155.213] has joined #shogun | 11:16 | |
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Ping timeout: 248 seconds] | 11:24 | |
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 11:26 | |
n4nd0 | gsomix: sorry my internet died | 11:26 |
gsomix | n4nd0, and my :) | 11:28 |
gsomix | I'm fine. doing some experiments with python api | 11:29 |
-!- pluskid [~pluskid@111.120.71.247] has quit [Ping timeout: 245 seconds] | 11:30 | |
-!- pluskid [~pluskid@202.130.113.141] has joined #shogun | 11:30 | |
n4nd0 | gsomix: cool, what are you preparing now? | 11:32 |
-!- gsomix [~gsomix@109.169.155.213] has quit [Read error: Connection reset by peer] | 11:36 | |
-!- gsomix [~gsomix@80.234.24.93] has joined #shogun | 11:38 | |
gsomix | n4nd0, python buffer and sequence protocols. e.g. for features+numpy.array, sliceing and working with numpy w/o memory copying | 11:39 |
n4nd0 | gsomix: ok, have fun with it then :) | 11:40 |
gsomix | huh :) | 11:41 |
n4nd0 | gsomix: are you working with more director classes? | 11:41 |
gsomix | n4nd0, later | 11:45 |
wiking | n4nd0: can we talk in 30 mins? | 11:47 |
n4nd0 | wiking: yeah | 11:48 |
wiking | cooll thnx brb | 11:49 |
-!- gsomix_ [~gsomix@109.169.134.39] has joined #shogun | 11:53 | |
-!- gsomix [~gsomix@80.234.24.93] has quit [Ping timeout: 264 seconds] | 11:55 | |
-!- gsomix_ [~gsomix@109.169.134.39] has quit [Ping timeout: 249 seconds] | 12:04 | |
-!- bern4rd [5331ddf1@gateway/web/freenode/ip.83.49.221.241] has joined #shogun | 12:20 | |
bern4rd | hi guys | 12:21 |
n4nd0 | what's up bern4rd | 12:23 |
-!- gsomix_ [~gsomix@109.169.134.39] has joined #shogun | 12:23 | |
bern4rd | hey n4nd0 | 12:27 |
@sonney2k | thunderstorm time again! | 12:33 |
n4nd0 | hey sonney2k, I think bern4rd wants to have more info about the news parser | 12:33 |
n4nd0 | bern4rd: right? | 12:33 |
bern4rd | yes yes | 12:33 |
@sonney2k | bern4rd, I will put all the NEWS files into a folder hierarchy | 12:34 |
@sonney2k | so we could render the latest releases from that by recursively traversing directory trees | 12:34 |
@sonney2k | and accessing all the NEWS files from there | 12:35 |
n4nd0 | recursion is cool | 12:35 |
@sonney2k | ? | 12:35 |
@sonney2k | bern4rd, actually I have some code for that (the NEWS on the website is automatically created from the latest NEWS file) | 12:36 |
@sonney2k | bern4rd, if you have an email address I will send it to you | 12:36 |
bern4rd | ok, i think i understand, but the old news dont change, so I can store them in the database and there is no need to parse them again right? | 12:36 |
@sonney2k | true | 12:36 |
bern4rd | bernard.hernandez.perez@gmail.com | 12:36 |
@sonney2k | bern4rd, but how do you figure out when a new NEWS file is uploaded? | 12:37 |
wiking | sonney2k: what about for and while is it for () and while () or the normal function call foo()? | 12:40 |
@sonney2k | for () | 12:40 |
@sonney2k | while () | 12:40 |
wiking | ok | 12:41 |
@sonney2k | only function() | 12:41 |
wiking | then i'll make the regexp now | 12:41 |
@sonney2k | wiking, that is how one can easily distinguish a function from a if/for/while clause | 12:41 |
bern4rd | I suppose as they are files the have a field "last date modified", so storing in the database together with the new the date of parsing you could check it. | 12:45 |
-!- blackburn [~blackburn@81.28.187.0] has joined #shogun | 12:47 | |
bern4rd | but you need recursivity anyway | 12:47 |
@sonney2k | blackburn, please merge wikings' patch once he has resolved the whitspace/indention issues | 12:48 |
@sonney2k | pluskid, have you seen bern4rd's new shogun website? Since you have strong artistic skills - any ideas for a logo? | 12:49 |
wiking | n4nd0: http://yuml.me/9c36be1b | 12:51 |
bern4rd | in fact I havent modify the home page because the most of the things are pictures. | 12:51 |
wiking | n4nd0: i want to talk about this first... ;) | 12:51 |
wiking | oh actually even better: http://yuml.me/56559868 | 12:53 |
n4nd0 | wiking: sorry, I was afk, let me see now | 12:58 |
wiking | okok | 12:58 |
n4nd0 | ok, I see | 12:59 |
n4nd0 | wiking: there might be an issue, if CStructuredData and CLatentData are different things, what about the laten so-svm? | 13:01 |
n4nd0 | the same applies to CStructuredLabels and CLatentLabels | 13:01 |
wiking | n4nd0: as far as i can see u use CStructuredData as a dummy object to be able to store any sorts of user defined structured label, right? | 13:03 |
n4nd0 | yes | 13:03 |
wiking | yeah well same applies to CLatentData | 13:04 |
n4nd0 | ok | 13:05 |
wiking | so that's why i'm saying here | 13:07 |
wiking | that maybe we could just merge the two | 13:08 |
wiking | as it's for the same purpose | 13:08 |
n4nd0 | wiking: how do you suggest to merge them? | 13:08 |
wiking | well it's just a naming convention and that's it ;) | 13:09 |
n4nd0 | ok | 13:09 |
n4nd0 | so what do CLatentLabels have? | 13:09 |
wiking | CLatentData | 13:09 |
wiking | atm | 13:09 |
wiking | since it's a free form/user defined structure | 13:10 |
n4nd0 | could be that there is no need for them and you can just use CStructuredLabels? | 13:10 |
wiking | well the naming is very misleading imo | 13:10 |
n4nd0 | why? what would be better? | 13:10 |
wiking | n4nd0: and actually no... | 13:10 |
wiking | since CLatentLabels->CBinaryLabels | 13:10 |
wiking | and extended a bit | 13:11 |
n4nd0 | hhm ok | 13:11 |
n4nd0 | then | 13:11 |
n4nd0 | what about CLatentData? | 13:12 |
n4nd0 | what does it have that makes it different? | 13:12 |
wiking | that | 13:12 |
wiking | nothing | 13:12 |
wiking | so i could use directly CStructuredData | 13:13 |
wiking | just that for me the name is misleading | 13:13 |
wiking | that's all | 13:13 |
n4nd0 | I was thinking of that yes | 13:13 |
n4nd0 | well ok ... but tell me why do you think so and maybe we can fix it | 13:13 |
wiking | well since it's used both in CLatentLabels and CLatentFeatures | 13:13 |
wiking | and actually it has nothing to do with structured output ;) | 13:14 |
n4nd0 | but the name sayd nothing about structured output :) | 13:14 |
wiking | yes but it's in it's name | 13:15 |
wiking | that it's Structured | 13:15 |
wiking | ;) | 13:15 |
n4nd0 | but not output ;) | 13:15 |
wiking | lol | 13:15 |
wiking | but you get what i mean no/ | 13:15 |
wiking | ? | 13:15 |
n4nd0 | you know CData would be a bit too general | 13:15 |
wiking | well i'd like that actually | 13:16 |
wiking | ;) | 13:16 |
wiking | brb | 13:16 |
wiking | gotta get out for 5 mins sorry :( | 13:16 |
n4nd0 | I would be ok with CData too | 13:16 |
n4nd0 | so feel free to change it if you want and I will adapt my branch to it | 13:17 |
wiking | ok | 13:17 |
wiking | thnx | 13:17 |
wiking | that sounds great | 13:17 |
wiking | i'll do the changes.... and let u know | 13:17 |
n4nd0 | ok :) | 13:17 |
wiking | but a bit away now | 13:17 |
pluskid | sonney2k: no, what's the url for the new site? | 13:18 |
n4nd0 | pluskid: I think it is not already up on the internet, but in bern4rd's github repo | 13:23 |
n4nd0 | maybe you can clone the repo and run it locally if you want to check it | 13:24 |
pluskid | n4nd0: url to his repo? | 13:24 |
n4nd0 | bern4rd: ^? | 13:24 |
n4nd0 | I think it should be this on | 13:25 |
n4nd0 | github.com/bahp/shogun-web | 13:25 |
pluskid | n4nd0: thanks! | 13:25 |
bern4rd | https://github.com/bahp/shogun-web.git | 13:26 |
n4nd0 | pluskid: you are welcome | 13:26 |
pluskid | bern4rd: how to run this? django? | 13:26 |
pluskid | or any screenshot? | 13:26 |
@sonney2k | pluskid, python manage.py runserver | 13:27 |
@sonney2k | pluskid, then it will show the url on the cmdline | 13:27 |
pluskid | sonney2k: OK, let me try | 13:28 |
pluskid | sonney2k: why I just get something like a default template? | 13:30 |
bern4rd | maybe i have change something | 13:31 |
bern4rd | one minute | 13:31 |
bern4rd | did you get see anything? | 13:35 |
pluskid | bern4rd: lots of | 13:35 |
pluskid | banner says "Say hello, new version is available" | 13:35 |
pluskid | but nothings seems to be related to shogun? | 13:35 |
pluskid | Lorem ipsum dolor sit amet consectetur elite. Nulla adipising telus non massa vestibulum est ultrices ege ta este commodo cons. | 13:36 |
bern4rd | this is the home view, and all the things are pictures an i am not good with photoshop or similar software | 13:36 |
bern4rd | but go to about | 13:36 |
bern4rd | or documentation | 13:36 |
bern4rd | to see some contenct | 13:36 |
pluskid | bern4rd: ah, got it | 13:36 |
bern4rd | anyway there are only a few articles in the webpage, you can put localhost:port/admin/ and add/edit/remove articles | 13:37 |
pluskid | ok, thanks! | 13:38 |
pluskid | I just want to look at how those pages looks like, in order to think about the logo | 13:38 |
bern4rd | sure :) | 13:38 |
pluskid | though I haven't got any good idea bout the logo yet :-/ | 13:39 |
-!- blackburn [~blackburn@81.28.187.0] has quit [Ping timeout: 248 seconds] | 13:46 | |
@sonney2k | pluskid, well we just started so no rush here | 13:48 |
@sonney2k | pluskid, actually bern4rd did all the work... | 13:48 |
pluskid | sonney2k: OK | 13:48 |
pluskid | cool work | 13:48 |
pluskid | byw, do you prefer complicated logo or more simple design? | 13:49 |
@sonney2k | no idea | 13:49 |
pluskid | anyway, let me keep this in my mind | 13:49 |
@sonney2k | pluskid, if all else fails we use the kanjis again... | 13:50 |
pluskid | sonney2k: haha, kanji looks cool actually | 13:50 |
-!- blackburn [~blackburn@81.28.187.0] has joined #shogun | 13:51 | |
pluskid | but at least we might be able to choose some other font | 13:51 |
@sonney2k | pluskid, well you could probably I cannot - a Japanese colleague was helping me here | 13:53 |
-!- bern4rd [5331ddf1@gateway/web/freenode/ip.83.49.221.241] has quit [Quit: Page closed] | 13:53 | |
pluskid | sonney2k: that's what I mean :), I might be able to find some good looking fonts to try out | 13:54 |
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: leaving] | 13:57 | |
-!- blackburn [~blackburn@81.28.187.0] has quit [Quit: Leaving.] | 14:41 | |
-!- heiko [~heiko@host86-177-117-30.range86-177.btcentralplus.com] has joined #shogun | 14:43 | |
wiking | sonney2k: around? | 15:29 |
wiking | is there anything that has been ported from COFFIN into shogun? | 15:29 |
wiking | ah yeah | 15:31 |
wiking | i've found | 15:31 |
-!- gsomix_ [~gsomix@109.169.134.39] has quit [Ping timeout: 264 seconds] | 15:45 | |
-!- gsomix_ [~gsomix@109.169.134.39] has joined #shogun | 15:45 | |
-!- pluskid [~pluskid@202.130.113.141] has quit [Ping timeout: 264 seconds] | 15:50 | |
-!- pluskid [~pluskid@202.130.113.141] has joined #shogun | 15:51 | |
CIA-18 | shogun: Heiko Strathmann master * raa97093 / src/shogun/kernel/Kernel.cpp : added proper errors message if kernel has no features - http://git.io/T_Gn4w | 16:02 |
CIA-18 | shogun: Heiko Strathmann master * r290672e / (3 files): loads of fixes and debug messages - http://git.io/eE0agw | 16:02 |
CIA-18 | shogun: Heiko Strathmann master * r7051cc8 / src/shogun/statistics/IndependenceTestStatistic.h : removed copy paste error which caused a strange bug - http://git.io/tM9FvA | 16:02 |
CIA-18 | shogun: Heiko Strathmann master * r377658c / examples/undocumented/libshogun/statistics_hsic.cpp : added test for hsic_gamma - http://git.io/e6OQqA | 16:02 |
CIA-18 | shogun: Heiko Strathmann master * r05e12d2 / (6 files in 3 dirs): Merge pull request #627 from karlnapf/master - http://git.io/e4hXMA | 16:02 |
shogun-buildbot | build #71 of deb1 - libshogun started, including [aa97093a161aecfbe08b9318b8ad3a8ccc363e00] | 16:02 |
shogun-buildbot | build #72 of deb1 - libshogun started, including [290672e263c6ee2f71a7a1c369330e55f4d67673, 7051cc84c50f7e88e4fb39ab5b07b74f2c736d25, 377658c556b16359f7901e3bbdf89ce369da3773] | 16:02 |
shogun-buildbot | build #72 of deb1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/72 | 16:03 |
shogun-buildbot | build #73 of deb1 - libshogun started, including [05e12d24ac0c75120f8a19f86c5edd9d8f1b9198] | 16:03 |
shogun-buildbot | build #2 of cyg1 - libshogun started, including [290672e263c6ee2f71a7a1c369330e55f4d67673, 7051cc84c50f7e88e4fb39ab5b07b74f2c736d25, 377658c556b16359f7901e3bbdf89ce369da3773] | 16:03 |
shogun-buildbot | build #73 of deb1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/73 | 16:03 |
shogun-buildbot | build #81 of deb3 - modular_interfaces started, including [290672e263c6ee2f71a7a1c369330e55f4d67673, 7051cc84c50f7e88e4fb39ab5b07b74f2c736d25, 377658c556b16359f7901e3bbdf89ce369da3773] | 16:03 |
shogun-buildbot | build #71 of deb1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/71 | 16:06 |
shogun-buildbot | build #80 of deb2 - static_interfaces started, including [290672e263c6ee2f71a7a1c369330e55f4d67673, 7051cc84c50f7e88e4fb39ab5b07b74f2c736d25, 377658c556b16359f7901e3bbdf89ce369da3773, 05e12d24ac0c75120f8a19f86c5edd9d8f1b9198] | 16:06 |
-!- heiko [~heiko@host86-177-117-30.range86-177.btcentralplus.com] has quit [Quit: Leaving.] | 16:07 | |
shogun-buildbot | build #2 of cyg1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/2 | 16:14 |
shogun-buildbot | build #3 of cyg1 - libshogun started, including [05e12d24ac0c75120f8a19f86c5edd9d8f1b9198, aa97093a161aecfbe08b9318b8ad3a8ccc363e00] | 16:14 |
shogun-buildbot | build #80 of deb2 - static_interfaces is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/80 | 16:15 |
shogun-buildbot | build #82 of deb3 - modular_interfaces started, including [05e12d24ac0c75120f8a19f86c5edd9d8f1b9198, aa97093a161aecfbe08b9318b8ad3a8ccc363e00] | 16:15 |
-!- pluskid [~pluskid@202.130.113.141] has quit [Quit: Leaving] | 16:26 | |
shogun-buildbot | build #3 of cyg1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/3 | 16:26 |
shogun-buildbot | build #81 of deb3 - modular_interfaces is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/81 | 16:41 |
shogun-buildbot | build #81 of deb2 - static_interfaces started, including [aa97093a161aecfbe08b9318b8ad3a8ccc363e00] | 16:41 |
shogun-buildbot | build #81 of deb2 - static_interfaces is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/81 | 16:47 |
shogun-buildbot | build #82 of deb3 - modular_interfaces is complete: Failure [failed test python_modular] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/82 blamelist: Heiko Strathmann <heiko.strathmann@gmail.com> | 17:04 |
-!- puffin444 [62e3926e@gateway/web/freenode/ip.98.227.146.110] has joined #shogun | 17:18 | |
-!- gsomix_ [~gsomix@109.169.134.39] has quit [Quit: Ex-Chat] | 17:18 | |
puffin444 | Hey soeren, about the memory read errors on Pull Request 615 | 17:22 |
puffin444 | *sonney2k | 17:22 |
-!- os253 [~oliverste@82.113.121.73] has joined #shogun | 17:45 | |
puffin444 | hello os253 | 17:45 |
os253 | Hi puffin444 | 17:45 |
os253 | Had a hard time connecting for some strange reason. Inventing yet another nickname helped :-) | 17:46 |
puffin444 | I see | 17:46 |
os253 | How are things? | 17:46 |
os253 | Hope no more encounters with unpleasant animals... | 17:47 |
puffin444 | They are progressing.The hash patch should be merged very soon making way for the gradient code. | 17:47 |
puffin444 | Yes. I sure hope so. | 17:47 |
os253 | Ok, sounds very good. | 17:48 |
puffin444 | I am a little disappointed by the amount of time needed to implement these kernel derivatives. | 17:48 |
puffin444 | I might have to double up to get back on schedule. | 17:48 |
os253 | Yes, it is cumbersome. Do you test the gradients standalone? | 17:48 |
puffin444 | No not yet. | 17:48 |
os253 | I'd recommend to use standalone grad checks for these. | 17:49 |
puffin444 | Okay, like tests in the examples folder? | 17:49 |
os253 | You can also "borrow" substantial proportions of code. The gplm stuff is efficiently done in principle | 17:49 |
os253 | But suffers from matlab constraints. | 17:49 |
puffin444 | Sure. The matlab code so far has been extremely helpful. | 17:50 |
os253 | The pygp thing I pointed you to earlier has reasonable readable gradients for most covariances needed, including sum/product. | 17:50 |
os253 | Ok...nit you are happy with thats also fine. | 17:50 |
os253 | *if | 17:50 |
os253 | Yes, I meant just checking the gradients for each kernel, including some unit tests. | 17:51 |
os253 | You can just calc (K(theta+delta)-K(theta-delta))/2 delta to estimate the gradient matrix. | 17:51 |
os253 | (assuming delta is a slight offset for *one* of the kernel parameters in theta. | 17:52 |
os253 | That matrix should be close to your analytical expressions. In my experience it is way easier to debug and test the ingredients separately. | 17:53 |
puffin444 | What about just comparing the results to GPML? | 17:53 |
puffin444 | Using the exact same inputs for the same kernels? | 17:53 |
puffin444 | For an examples/ test though it would make sense for an approximation comparision. | 17:54 |
os253 | That solution works of course, but will make it hard to ever invent new kernels... | 17:55 |
os253 | I think unit tests for the gradients are very useful and will safe you pain down the line. | 17:56 |
puffin444 | Good point | 17:56 |
os253 | Regarding the combination of kernels: I can see how that is difficult in the current setup with names. | 17:56 |
os253 | You could just create a mapping adding #kernelnumber_ as prefix to the respective parameter names. | 17:57 |
os253 | Is that your intention with the hash? | 17:57 |
os253 | Also: how does the existing multiple kernel code in shogun deals with this; is the combinator not used to update the parameters and this is done ma | 17:58 |
os253 | manually? | 17:58 |
puffin444 | The hash is actually used to check if any of the parameters have been changed. This way the GP/Inference Method needs only to update their matrices when something changes. | 17:58 |
os253 | Ok, so that's for caching - good. | 17:59 |
puffin444 | What I have right now (on my hard drive) for kernel derivatives is an extra dictionary that uses a parameter pointer as a key and the kernel pointer as the data | 17:59 |
puffin444 | So that I can automatically look up which parameter belongs to which kernel. | 18:00 |
os253 | Yes, that seems suitable. And what names are generating for them, in case you have two copies of the same kernel? | 18:00 |
puffin444 | They may have the same name, its the pointer to the kernel which differentiates them. | 18:02 |
os253 | And outside?the current interface for kernel parameters is make-based. | 18:03 |
os253 | *name-based | 18:03 |
puffin444 | The current interface is name-based. My new interface will be pointer based. | 18:03 |
os253 | You do this in addition or instead of names? | 18:04 |
puffin444 | The problem is that I can't manipulate parameter names much because the rest of the model selection depends on them for other things. | 18:04 |
puffin444 | In addition to. The pointer is used to resolve duplicates. | 18:04 |
os253 | Hm... The downside is that it will be hard for a user to manually inspect the current parameter settings. | 18:05 |
os253 | But perhaps that is ok, as on can always query the sub kernels which offer named access again. | 18:05 |
os253 | I guess it's fine actually. | 18:06 |
puffin444 | Keep in mind too that at least in GradientModelSelection there is a tree which contains all the parameters | 18:06 |
puffin444 | Printing out the tree reveals the values and what kernels they belong to. | 18:07 |
os253 | Yes, good point. | 18:08 |
os253 | Well, it sounds like it is pretty clear what to work on. Fingers crossed that the logistic changes can be resolved reasonably quickly. | 18:09 |
os253 | To come back to the points above: I ink grad check should be testable for each kernel. | 18:11 |
os253 | *think | 18:11 |
puffin444 | Yes. If all things go to plan I can catch up next week, | 18:11 |
os253 | If products and sums work the worst should be over. | 18:11 |
os253 | Perhaps we can catch up mid week or so? | 18:12 |
puffin444 | Yes | 18:12 |
puffin444 | What day/time would be good for you? | 18:13 |
os253 | I have virtually no constraints. We, same time as today for example? | 18:13 |
os253 | But we can setup sth. Short notice as well... I am generally around. | 18:14 |
puffin444 | Perhaps Tuesday/Wednesday? | 18:15 |
os253 | Both good. Perhaps just email what will be better depending in progress. | 18:17 |
os253 | I'll mark 3:30 UCT both days. | 18:17 |
puffin444 | ok | 18:18 |
os253 | Talk next week then. | 18:19 |
puffin444 | ok | 18:19 |
os253 | Cu | 18:20 |
puffin444 | See you later | 18:20 |
-!- os253 [~oliverste@82.113.121.73] has quit [Quit: Colloquy for iPad - http://colloquy.mobi] | 18:20 | |
-!- os253 [~os@82.113.121.73] has joined #shogun | 18:20 | |
-!- os253 [~os@82.113.121.73] has quit [Client Quit] | 18:21 | |
puffin444 | sonney2k, about the question on the pull request | 18:21 |
-!- blackburn [~blackburn@81.28.187.0] has joined #shogun | 18:55 | |
puffin444 | hey blackburn | 18:55 |
blackburn | hey | 18:55 |
blackburn | what's up? | 18:56 |
puffin444 | I have updated my pull request #615. Do you have the power to merge it? | 18:56 |
blackburn | I am a superman ya ya ya | 18:57 |
puffin444 | blackburn, you are the one of the shogun deities | 18:58 |
blackburn | puffin444: here you goo | 19:01 |
CIA-18 | shogun: Sergey Lisitsyn master * r9d82c0b / (30 files in 10 dirs): Merge branch 'master' of git://github.com/puffin444/shogun (+8 more commits...) - http://git.io/ZstjGQ | 19:01 |
puffin444 | thanks blackburn | 19:01 |
shogun-buildbot | build #74 of deb1 - libshogun started, including [93caa1cd49b29c1d9e027f079d64455e3bdb28fd] | 19:02 |
shogun-buildbot | build #75 of deb1 - libshogun started, including [323dd6117f8907b0c3a1c2ecdbe1478fa7fdf1d2, 2b6ea8ffa360a14902736d73f7d607562a082d48] | 19:02 |
shogun-buildbot | build #75 of deb1 - libshogun is complete: Failure [failed compile] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/75 blamelist: puffin444 <walke434@msu.edu> | 19:04 |
shogun-buildbot | build #76 of deb1 - libshogun started, including [06285a899ac103df166be88d780a3e1def412494, 1e4e7317838682cc0615212a71ea4133a11a87cf, 41bc0efc7cd01c385232bb45acca8a8fcb571799, 009faf212d641da1e684520fe9b863523b7d4ed6, 3711463a7e8d7ed6ed4fc037e1be5e536bdf5532, 9d82c0b21e2d7cd6ff4624a8cb684e3f5ecd6c76] | 19:04 |
blackburn | MERGE-O-RAMA | 19:04 |
puffin444 | Noooo! doggone it | 19:04 |
CIA-18 | shogun: Chiyuan Zhang master * rd66fc4a / (3 files in 2 dirs): measure running time of each round in ShareBoost. - http://git.io/IY3NIQ | 19:04 |
CIA-18 | shogun: Chiyuan Zhang master * r132f7bd / src/shogun/multiclass/ShareBoost.cpp : use shogun CTime to measure run-time. - http://git.io/Thmd7Q | 19:04 |
CIA-18 | shogun: Chiyuan Zhang master * rf3e3c70 / src/shogun/multiclass/ShareBoost.cpp : better use of CTime - http://git.io/a9we5Q | 19:04 |
CIA-18 | shogun: Chiyuan Zhang master * r658757d / src/shogun/multiclass/ShareBoost.cpp : remove unnecessary type casts. - http://git.io/cQiS_Q | 19:04 |
CIA-18 | shogun: Sergey Lisitsyn master * r6af57ce / (3 files in 2 dirs): Merge branch 'multiclass' of git://github.com/pluskid/shogun - http://git.io/Flzqpw | 19:04 |
blackburn | mergeality! | 19:04 |
shogun-buildbot | build #74 of deb1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/74 | 19:04 |
shogun-buildbot | build #83 of deb3 - modular_interfaces started, including [93caa1cd49b29c1d9e027f079d64455e3bdb28fd] | 19:04 |
shogun-buildbot | build #4 of cyg1 - libshogun started, including [93caa1cd49b29c1d9e027f079d64455e3bdb28fd] | 19:04 |
shogun-buildbot | build #83 of deb3 - modular_interfaces is complete: Failure [failed configure] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/83 blamelist: puffin444 <walke434@msu.edu> | 19:05 |
shogun-buildbot | build #82 of deb2 - static_interfaces started, including [93caa1cd49b29c1d9e027f079d64455e3bdb28fd] | 19:05 |
wiking | wooooooooooooooooooooh | 19:05 |
wiking | i've woken up ;) | 19:05 |
shogun-buildbot | build #82 of deb2 - static_interfaces is complete: Failure [failed configure] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/82 blamelist: puffin444 <walke434@msu.edu> | 19:05 |
puffin444 | Hey blackburn, what in the world would be causing these errors? It looks like it can't find certain files that I did not touch. | 19:06 |
CIA-18 | shogun: Sergey Lisitsyn master * r25dc5ad / (6 files in 4 dirs): Updated feature block logit regression - http://git.io/Wmw10Q | 19:06 |
CIA-18 | shogun: Sergey Lisitsyn master * r3717f65 / (6 files in 4 dirs): Merge branch 'slep' of git://github.com/lisitsyn/shogun - http://git.io/nWJOFg | 19:06 |
shogun-buildbot | build #77 of deb1 - libshogun started, including [d66fc4accdce3dbe989590570c07f7d8d252032e] | 19:07 |
blackburn | puffin444: nevermind, something related to sonney2k dirty deals | 19:07 |
puffin444 | oh okay :) | 19:07 |
blackburn | ok finally MERGEALITY | 19:07 |
CIA-18 | shogun: Sergey Lisitsyn master * rafbc2bc / (12 files in 5 dirs): Merge branch 'latent' of git://github.com/vigsterkr/shogun (+29 more commits...) - http://git.io/EcDWPg | 19:07 |
shogun-buildbot | build #77 of deb1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/77 | 19:07 |
wiking | so is set ts=4 the default setting for everybody here? :) | 19:07 |
shogun-buildbot | build #78 of deb1 - libshogun started, including [132f7bd9abc69738cc62455fcf550c30c01b2236, f3e3c703d12cfc79f6948cf48d1aedaa84c29ac8, 658757de4fecf0e5435f5e45deab88cbcf2a3eec, 6af57ce70c5c1eeb636fb7b6bee7dbf3cced42da, 25dc5ade28a0e69fd6b827d7fe1856c370cd9e37, 3717f6576a8bbdbe2b09731494d2005c66311b0f] | 19:07 |
blackburn | wiking: yeah for me it is | 19:08 |
wiking | puffin444: ? | 19:08 |
shogun-buildbot | build #76 of deb1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/76 | 19:08 |
shogun-buildbot | build #84 of deb3 - modular_interfaces started, including [d66fc4accdce3dbe989590570c07f7d8d252032e] | 19:08 |
puffin444 | what's ts? | 19:08 |
wiking | is there a way i can have ts=4 but displayed it as ts=2 ?:) | 19:09 |
blackburn | hmm | 19:09 |
wiking | as currently i'm tabstop=2 | 19:09 |
wiking | softtabstop=0 | 19:09 |
wiking | shiftwidth=2 | 19:09 |
wiking | and would like to keep it like this | 19:09 |
shogun-buildbot | build #78 of deb1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/78 | 19:09 |
shogun-buildbot | build #83 of deb2 - static_interfaces started, including [d66fc4accdce3dbe989590570c07f7d8d252032e, 06285a899ac103df166be88d780a3e1def412494, 1e4e7317838682cc0615212a71ea4133a11a87cf, 41bc0efc7cd01c385232bb45acca8a8fcb571799, 009faf212d641da1e684520fe9b863523b7d4ed6, 3711463a7e8d7ed6ed4fc037e1be5e536bdf5532, 9d82c0b21e2d7cd6ff4624a8cb684e3f5ecd6c76] | 19:09 |
blackburn | 14 set tabstop=4 | 19:10 |
blackburn | 15 set shiftwidth=4 | 19:10 |
blackburn | 16 set listchars=tab:?? | 19:10 |
blackburn | I use that | 19:10 |
wiking | mmhm | 19:10 |
wiking | so yeah u use 4 instead of 2 | 19:10 |
wiking | ok i'll check vim wiki | 19:10 |
wiking | maybe there's a way around here | 19:11 |
puffin444 | hmm.... I better make sure that my tabstops are consistent. | 19:11 |
wiking | btw: my 1tb external drive just died today.. i've bought it 2.5 years ago BUT since it has 3 years warranty by the manufacturer, they'll replace it \o/ pretty cool | 19:12 |
blackburn | cool | 19:12 |
blackburn | but did you lost much data? | 19:12 |
shogun-buildbot | build #83 of deb2 - static_interfaces is complete: Failure [failed test libshogun] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/83 blamelist: puffin444 <walke434@msu.edu>, Sergey Lisitsyn <lisitsyn.s.o@gmail.com>, Chiyuan Zhang <pluskid@gmail.com> | 19:13 |
shogun-buildbot | build #85 of deb3 - modular_interfaces started, including [06285a899ac103df166be88d780a3e1def412494, 1e4e7317838682cc0615212a71ea4133a11a87cf, 41bc0efc7cd01c385232bb45acca8a8fcb571799, 009faf212d641da1e684520fe9b863523b7d4ed6, 3711463a7e8d7ed6ed4fc037e1be5e536bdf5532, 9d82c0b21e2d7cd6ff4624a8cb684e3f5ecd6c76, 132f7bd9abc69738cc62455fcf550c30c01b2236, f3e3c703d12cfc79f6948cf48d1aedaa84c29ac8, 658757de4fecf0e5435f5e45deab88cbcf2a3eec, 6af57ce70c5c1eeb636fb | 19:13 |
wiking | thanks god i've already transferred everything on my new 2tb raid-1 backup drive couple of months ago... so i've just used this one as spare external drive | 19:13 |
blackburn | puffin444: argh your thing caused error I guess | 19:13 |
wiking | sonney2k: freebsd build? | 19:13 |
blackburn | puffin444: modelselection_parameter_tree | 19:14 |
wiking | sonney2k: i have a freebsd server in hetzner.de so i could set up a buildbot there for shogun | 19:14 |
shogun-buildbot | build #4 of cyg1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/4 | 19:15 |
shogun-buildbot | build #5 of cyg1 - libshogun started, including [d66fc4accdce3dbe989590570c07f7d8d252032e, 06285a899ac103df166be88d780a3e1def412494, 1e4e7317838682cc0615212a71ea4133a11a87cf, 41bc0efc7cd01c385232bb45acca8a8fcb571799, 009faf212d641da1e684520fe9b863523b7d4ed6, 3711463a7e8d7ed6ed4fc037e1be5e536bdf5532, 9d82c0b21e2d7cd6ff4624a8cb684e3f5ecd6c76, 132f7bd9abc69738cc62455fcf550c30c01b2236, f3e3c703d12cfc79f6948cf48d1aedaa84c29ac8, 658757de4fecf0e5435f5e45deab88c | 19:15 |
puffin444 | Hmm... This may be related to HashedWDFeatures. I need to talk to sonney2k about this. | 19:17 |
wiking | mmm the succesfully built binary is not kept for download? at least 1/per day? | 19:18 |
blackburn | wiking: nope | 19:19 |
blackburn | HARDCORE | 19:19 |
wiking | well nightly build | 19:20 |
wiking | as all other sw has, would be cool | 19:20 |
blackburn | wiking: we actually wanted to set up .deb packaging here | 19:20 |
wiking | nightly? | 19:20 |
wiking | :D | 19:20 |
@sonney2k | puffin444, talk to me please :) | 19:20 |
blackburn | I have no idea how is it going under mac os but would be nice to have some package too | 19:20 |
wiking | blackburn: it's in mac.... macports | 19:21 |
blackburn | wiking: why not? | 19:21 |
wiking | yeah sounds cool | 19:21 |
@sonney2k | wiking, well you need 3G of memory and all the shogun deps installed | 19:21 |
@sonney2k | wiking, for freebsd bb | 19:21 |
wiking | sonney2k: have 8 | 19:21 |
@sonney2k | k | 19:21 |
@sonney2k | you spend quite a bit of $$ then right? | 19:22 |
wiking | blackburn: Maintainers: jameskyle@macports.org | 19:22 |
@sonney2k | eur €€€ :D | 19:22 |
wiking | sonney2k: it was the 50euro/month option | 19:22 |
puffin444 | sonney2k, the new incremental hash function works slightly differently from the previous function. HashedWDFeatures used the incremental hash function. | 19:22 |
wiking | sonney2k: but bought it like 2 years ago or so | 19:22 |
@sonney2k | sounds like a good deal then | 19:22 |
puffin444 | And I updated that code. I may have not implemented it correctly. | 19:22 |
wiking | sonney2k: i have it with a couple of friends of mine, and they are doing some hosting there, so it pays for itself | 19:22 |
@sonney2k | wiking, the shogun toolbox server is hosted at ovh (~27 €/month), 4G, and some core 2 quad | 19:23 |
wiking | sonney2k: oh that's nice! | 19:23 |
wiking | sonney2k: so nightly builds for download? : | 19:23 |
@sonney2k | was also some offer | 19:23 |
wiking | :) | 19:23 |
wiking | blackburn: that's the guy who maintains shogun package for mac | 19:24 |
wiking | blackburn: shogun @0.9.3 | 19:24 |
wiking | blackburn: i might need to email him about not being up-to-date | 19:24 |
@sonney2k | wiking, did you manage to adjust the tabs to ts=4 ? | 19:24 |
@sonney2k | I mean with ts=2 it all must look weird? | 19:24 |
@sonney2k | wiking, well... | 19:25 |
wiking | sonney2k: i like ts=2... but i'm just checking if it's possible to visualize as ts=2 but actually write it as ts=4 | 19:25 |
wiking | sonney2k: and then i'll send the patch | 19:25 |
@sonney2k | wiking, the new one you mean? | 19:26 |
@sonney2k | blackburn, already merged everything :D | 19:26 |
shogun-buildbot | build #5 of cyg1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/5 | 19:26 |
wiking | oh fuck | 19:27 |
wiking | it's in? :))) | 19:27 |
wiking | rotfl | 19:27 |
wiking | ok then i'll send a new PR soon | 19:27 |
* sonney2k check how it looks like | 19:27 | |
@sonney2k | wiking, you actually use spaces instead of tabs :/ | 19:27 |
wiking | sonney2k: yes | 19:28 |
@sonney2k | wiking, so please correct to true tabs with ts=4 | 19:29 |
wiking | ok | 19:29 |
* wiking hates tabs :P | 19:29 | |
@sonney2k | wiking, ACTION? | 19:29 |
@sonney2k | puffin444, was the new incremental murmur hash giving same results as murmur? | 19:30 |
puffin444 | Yes. As long as the finalization function is called. I get the same final results in lib_hashing | 19:31 |
puffin444 | *llbrary_hash | 19:31 |
puffin444 | *library_hash | 19:31 |
puffin444 | can't type :) | 19:31 |
puffin444 | Do you know what the role of incremental hashing was in HashedWDFeatures? | 19:32 |
shogun-buildbot | build #85 of deb3 - modular_interfaces is complete: Failure [failed compile csharp_modular] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/85 blamelist: puffin444 <walke434@msu.edu>, Sergey Lisitsyn <lisitsyn.s.o@gmail.com>, Chiyuan Zhang <pluskid@gmail.com> | 19:33 |
shogun-buildbot | build #84 of deb2 - static_interfaces started, including [132f7bd9abc69738cc62455fcf550c30c01b2236, f3e3c703d12cfc79f6948cf48d1aedaa84c29ac8, 658757de4fecf0e5435f5e45deab88cbcf2a3eec, 6af57ce70c5c1eeb636fb7b6bee7dbf3cced42da, 25dc5ade28a0e69fd6b827d7fe1856c370cd9e37, 3717f6576a8bbdbe2b09731494d2005c66311b0f] | 19:33 |
@sonney2k | puffin444, yeah I wrote that code :D | 19:33 |
@sonney2k | it just determines some index in a weight vector | 19:33 |
@sonney2k | puffin444, so when you now call the new hash function - it should be fine | 19:34 |
wiking | sonney2k: action? | 19:34 |
shogun-buildbot | build #84 of deb2 - static_interfaces is complete: Failure [failed test libshogun] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/84 blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com>, Chiyuan Zhang <pluskid@gmail.com> | 19:35 |
shogun-buildbot | build #79 of deb1 - libshogun started, including [88be1739096e3c7b4d59e3465f97b5370bc16605, e3a2ac684df06fcfadc9162ec3dcaa82a7daa89e, 68378b41163d7f15eda0f6503fc44e8745e56de2, 091285be4c44dca5fa73020688cce8cd29c36cb8, a7f18bb90648e3d02e641475a71942db79e72484, a6a715cfe9151870d6b99974568d044f25773f04, 7ebe42224b08e5bddbee8e769fa57d28b5b49b63, 7b4eefd84de098993933f77f3c79850434d8a4a2, 33342d3614e403e89ff3e9010c44eb9f9d8078b6, df087f20783819a85ee0361c145442 | 19:35 |
@sonney2k | wiking, ahh it is some color probably? | 19:36 |
wiking | ohoooo this freebsd build starts nice: -bash: ./configure: /bin/bash: bad interpreter: No such file or directory | 19:36 |
@sonney2k | haha | 19:37 |
wiking | :D | 19:37 |
shogun-buildbot | build #79 of deb1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/79 | 19:37 |
shogun-buildbot | build #86 of deb3 - modular_interfaces started, including [88be1739096e3c7b4d59e3465f97b5370bc16605, e3a2ac684df06fcfadc9162ec3dcaa82a7daa89e] | 19:37 |
shogun-buildbot | build #6 of cyg1 - libshogun started, including [88be1739096e3c7b4d59e3465f97b5370bc16605, e3a2ac684df06fcfadc9162ec3dcaa82a7daa89e, 68378b41163d7f15eda0f6503fc44e8745e56de2, 091285be4c44dca5fa73020688cce8cd29c36cb8, a7f18bb90648e3d02e641475a71942db79e72484, a6a715cfe9151870d6b99974568d044f25773f04, 7ebe42224b08e5bddbee8e769fa57d28b5b49b63] | 19:37 |
wiking | i guess there'll be some patches now from me for fbsd build | 19:37 |
@sonney2k | puffin444, so if it computes something with similar hash properties as before at least newly trained models will just be fine | 19:37 |
shogun-buildbot | build #86 of deb3 - modular_interfaces is complete: Failure [failed compile libshogun] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/86 blamelist: Viktor Gal <viktor.gal@maeth.com> | 19:37 |
shogun-buildbot | build #87 of deb3 - modular_interfaces started, including [68378b41163d7f15eda0f6503fc44e8745e56de2, 091285be4c44dca5fa73020688cce8cd29c36cb8, a7f18bb90648e3d02e641475a71942db79e72484, a6a715cfe9151870d6b99974568d044f25773f04, 7ebe42224b08e5bddbee8e769fa57d28b5b49b63, 7b4eefd84de098993933f77f3c79850434d8a4a2, 33342d3614e403e89ff3e9010c44eb9f9d8078b6, df087f20783819a85ee0361c145442dd5e5825c7, a530bc94426ce4992103c6bc4dbd4615c0d00814, 20a604e44d5a472fdf134 | 19:37 |
@sonney2k | hmmhh really not good - lots of commits that don't build are in here :/ | 19:38 |
wiking | sonney2k: great! i've got an email! | 19:38 |
wiking | from the buildbot | 19:38 |
wiking | sonney2k: aaaaaah... this is commit by commit right? | 19:39 |
@sonney2k | wiking, yeah - you had some intermediate commits that just dont' compile | 19:39 |
@sonney2k | yes | 19:39 |
wiking | sonney2k: yeps.... | 19:39 |
wiking | that's the before the merger commits | 19:39 |
@sonney2k | that is why the rule is to never commit sth (and pull request) that doesn't compile | 19:39 |
@sonney2k | yeah | 19:39 |
@sonney2k | too late now anyway | 19:40 |
@sonney2k | git commit --amend will work around this in future | 19:40 |
wiking | sonney2k: mmmm let's check that command | 19:40 |
@sonney2k | woah, on cygwin libshogun builds already for 10 minutes | 19:41 |
@sonney2k | (<1 minute under linux) | 19:41 |
wiking | /usr/bin/env bash would work on linux right? | 19:42 |
@sonney2k | wiking, yes | 19:43 |
wiking | ok | 19:43 |
wiking | sonney2k: file-by-file PR or as a big one for the fbsd fixes? | 19:43 |
@sonney2k | wiking, you can send one big PR but with several small commits please | 19:43 |
wiking | sonney2k: ok | 19:43 |
@sonney2k | (or when you are done with a couple of fixes and won't work on that for some time) | 19:44 |
shogun-buildbot | build #6 of cyg1 - libshogun is complete: Failure [failed compile] Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/6 blamelist: Viktor Gal <viktor.gal@maeth.com> | 19:45 |
shogun-buildbot | build #7 of cyg1 - libshogun started, including [7b4eefd84de098993933f77f3c79850434d8a4a2, 33342d3614e403e89ff3e9010c44eb9f9d8078b6, df087f20783819a85ee0361c145442dd5e5825c7, a530bc94426ce4992103c6bc4dbd4615c0d00814, 20a604e44d5a472fdf134529c2c017a0da4a9be8, ec0ebffdea64b392c6c60c3d0ea5198c17e9f2f9, 89d1375604d87f75970aef1f6ea3bd5c06bc2178, 4dcd9db008d00d09abed04baaaa45f1f21b2f562, 6314a238ed474ee8b35630a58b05535051f172d0, 5b5ef0d83a5dc6ecead8bcd8573d25a | 19:45 |
wiking | sonney2k: buildbot-slave shoudl be enough for me i guess | 19:47 |
@sonney2k | I think we have no choice but to wait until the buildbot went to all of the builds... | 19:47 |
@sonney2k | wiking, yes | 19:47 |
wiking | atlas 3.8.4 should be good right? | 19:51 |
wiking | or should i go with 3.9.11 | 19:51 |
wiking | ? | 19:51 |
@sonney2k | wiking, any blas/atlas/lapack will work - but fastest is usually that latest... | 19:53 |
wiking | sonney2k: 3.9.11 seems like an experimental ports in fbsd so i stick with stable version of it | 19:54 |
@sonney2k | wiking, COFFIN is what we call dotfeatures in shogun | 19:54 |
wiking | sonney2k: yeah i was checking your supplementary material for that COFFIN paper | 19:55 |
wiking | sonney2k: so then here's the question | 19:55 |
wiking | sonney2k: having these very free form of features (CLatentFeatures) | 19:55 |
wiking | first i went with CDotFeatures inheritance... | 19:56 |
wiking | but then there's no way one would be able to define all the dot* functions without knowing it's structure... (i mean the features itself) | 19:56 |
wiking | lol gcc 4.6.4.20120608 | 19:57 |
wiking | that's the dependency for atlas :) | 19:57 |
wiking | this will take a while | 19:57 |
shogun-buildbot | build #87 of deb3 - modular_interfaces is complete: Failure [failed compile csharp_modular] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/87 blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com>, Viktor Gal <viktor.gal@maeth.com> | 19:57 |
shogun-buildbot | build #85 of deb2 - static_interfaces started, including [88be1739096e3c7b4d59e3465f97b5370bc16605, e3a2ac684df06fcfadc9162ec3dcaa82a7daa89e, 68378b41163d7f15eda0f6503fc44e8745e56de2, 091285be4c44dca5fa73020688cce8cd29c36cb8, a7f18bb90648e3d02e641475a71942db79e72484, a6a715cfe9151870d6b99974568d044f25773f04, 7ebe42224b08e5bddbee8e769fa57d28b5b49b63, 7b4eefd84de098993933f77f3c79850434d8a4a2, 33342d3614e403e89ff3e9010c44eb9f9d8078b6, df087f20783819a85ee036 | 19:57 |
shogun-buildbot | build #7 of cyg1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/7 | 19:57 |
shogun-buildbot | build #84 of deb3 - modular_interfaces is complete: Failure [failed test python_modular] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/84 blamelist: Chiyuan Zhang <pluskid@gmail.com> | 19:58 |
blackburn | I do not understand krr issue | 19:58 |
@sonney2k | blackburn, ? | 19:59 |
wiking | sonney2k: any insights about that? :) | 19:59 |
@sonney2k | wiking, not sure what you are asking... | 19:59 |
@sonney2k | wiking, for dot features you need to know the feature space (at least how to compute it) | 20:00 |
blackburn | sonney2k: krr example hangs.. | 20:00 |
@sonney2k | to get these add / etc stuff to work | 20:00 |
@sonney2k | blackburn, I dont' understand that either... it worked totally fine when I was running this on the cmdline | 20:00 |
shogun-buildbot | build #85 of deb2 - static_interfaces is complete: Failure [failed test libshogun] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/85 blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com>, Viktor Gal <viktor.gal@maeth.com> | 20:00 |
blackburn | sonney2k: did I tell you how I tried svm in fourier space? :D | 20:01 |
wiking | sonney2k: yeps, but i do not... thus i cannot have my features be dot features... or i force it that it is and ask the user to implement the dot features api | 20:01 |
@sonney2k | blackburn, no | 20:01 |
blackburn | sonney2k: I've got the same results :D | 20:01 |
@sonney2k | blackburn, like? | 20:02 |
blackburn | sonney2k: original R domain and fourier image of it | 20:02 |
blackburn | sonney2k: dot product is equal in both.. | 20:02 |
@sonney2k | blackburn, heh :D | 20:02 |
blackburn | sonney2k: I forgot that | 20:02 |
blackburn | but then experimentally proved :D | 20:02 |
@sonney2k | wiking, I guess real valued features are sufficient. | 20:03 |
@sonney2k | blackburn, good job :-)) | 20:03 |
@sonney2k | though numerics usually causes different results | 20:04 |
blackburn | sonney2k: parceval's theorem to be correct | 20:04 |
puffin444 | sonney2k, fixed that error with modelselection_parameter_tree | 20:10 |
@sonney2k | puffin444, is that already merged? | 20:16 |
@sonney2k | puffin444, hmmhh cannot be merged - my suggestion is to stay tuned until the buildbots managed to build all the commits | 20:18 |
@sonney2k | then start to fix things (if there are any issues) | 20:18 |
puffin444 | hmmhh? | 20:18 |
@sonney2k | puffin444, the buildbot builds *every* commit | 20:30 |
@sonney2k | so when you have one broken one in there it will complain | 20:31 |
blackburn | sonney2k: can you set up timeout? | 20:31 |
blackburn | or anything | 20:31 |
@sonney2k | that is why you should make sure everything is build | 20:31 |
blackburn | to build only latest | 20:31 |
@sonney2k | blackburn, no | 20:31 |
@sonney2k | I don't want to | 20:31 |
blackburn | pleease | 20:31 |
blackburn | :D | 20:31 |
@sonney2k | it makes a lot of sense this way | 20:31 |
blackburn | bisect? | 20:31 |
@sonney2k | yes | 20:31 |
blackburn | I don't think this actually hepls | 20:31 |
@sonney2k | you really see the commit that caused the failure | 20:32 |
puffin444 | When do I know that all the commits have been built? | 20:32 |
blackburn | puffin444: when it calmed down | 20:32 |
blackburn | :D | 20:32 |
@sonney2k | puffin444, well look at http://www.shogun-toolbox.org/buildbot/waterfall | 20:34 |
@sonney2k | 30 pending... | 20:34 |
blackburn | night of fun for buildbot | 20:35 |
blackburn | 30*5 minutes at least | 20:35 |
blackburn | or more? | 20:35 |
puffin444 | So I better wait at least a few hours | 20:35 |
blackburn | puffin444: take a book :D | 20:36 |
blackburn | 'war and peace' may be | 20:36 |
@sonney2k | blackburn, if all fail then 1 minute for each build :D | 20:51 |
@sonney2k | puffin444, blackburn its done | 21:01 |
blackburn | cool | 21:05 |
puffin444 | sonney2k, I am confused. When can I merge my fix? | 21:37 |
-!- rieck [~rieck@login.informatik.uni-goettingen.de] has joined #shogun | 21:48 | |
puffin444 | blackburn, can my fix be merged now? | 21:51 |
blackburn | rieck: welcome back ;) | 21:51 |
CIA-18 | shogun: puffin444 master * r79cfd56 / (5 files in 2 dirs): Merge remote branch 'upstream/master' - http://git.io/dYbe9A | 21:52 |
CIA-18 | shogun: puffin444 master * r864211e / (2 files): Added checking for empty features in InferenceMethod and ExactInferenceMethod. - http://git.io/iv7nSQ | 21:52 |
CIA-18 | shogun: Sergey Lisitsyn master * rd8d1b43 / (2 files): Merge branch 'master' of git://github.com/puffin444/shogun - http://git.io/i0BPCg | 21:52 |
blackburn | puffin444: for your pleasure | 21:52 |
rieck | it's been a long time. | 21:56 |
shogun-buildbot | build #80 of deb1 - libshogun started, including [79cfd562aef157e6b1859c27e0b2b80bd68d8455] | 21:57 |
shogun-buildbot | build #81 of deb1 - libshogun started, including [864211e1b28b4cb5cd91456772d2d20083a52346, d8d1b438115a43a55224b51336eaee9b9765f7fc] | 21:57 |
blackburn | rieck: how is it going? | 21:57 |
shogun-buildbot | build #81 of deb1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/81 | 21:58 |
shogun-buildbot | build #86 of deb2 - static_interfaces started, including [864211e1b28b4cb5cd91456772d2d20083a52346] | 21:58 |
shogun-buildbot | build #8 of cyg1 - libshogun started, including [864211e1b28b4cb5cd91456772d2d20083a52346, d8d1b438115a43a55224b51336eaee9b9765f7fc] | 21:58 |
puffin444 | oh please don't fail | 22:00 |
shogun-buildbot | build #80 of deb1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/80 | 22:02 |
shogun-buildbot | build #88 of deb3 - modular_interfaces started, including [864211e1b28b4cb5cd91456772d2d20083a52346, d8d1b438115a43a55224b51336eaee9b9765f7fc] | 22:02 |
shogun-buildbot | build #86 of deb2 - static_interfaces is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/86 | 22:03 |
shogun-buildbot | build #87 of deb2 - static_interfaces started, including [d8d1b438115a43a55224b51336eaee9b9765f7fc, 79cfd562aef157e6b1859c27e0b2b80bd68d8455] | 22:03 |
shogun-buildbot | build #87 of deb2 - static_interfaces is complete: Failure [failed test libshogun] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/87 blamelist: puffin444 <walke434@msu.edu>, Sergey Lisitsyn <lisitsyn.s.o@gmail.com> | 22:05 |
shogun-buildbot | build #89 of deb3 - modular_interfaces started, including [79cfd562aef157e6b1859c27e0b2b80bd68d8455] | 22:05 |
puffin444 | okay, why did build 86 succeed and 87 fail? | 22:06 |
puffin444 | blackburn, I am confused. Why did one succeed and the other fail? | 22:08 |
shogun-buildbot | build #8 of cyg1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/8 | 22:09 |
shogun-buildbot | build #9 of cyg1 - libshogun started, including [79cfd562aef157e6b1859c27e0b2b80bd68d8455] | 22:09 |
blackburn | puffin444: no idea :) | 22:11 |
puffin444 | The failed build reported the exact same error, which shouldn't happen after my fix. | 22:12 |
shogun-buildbot | build #88 of deb3 - modular_interfaces is complete: Failure [failed compile csharp_modular] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/88 blamelist: puffin444 <walke434@msu.edu>, Sergey Lisitsyn <lisitsyn.s.o@gmail.com> | 22:21 |
shogun-buildbot | build #9 of cyg1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/9 | 22:21 |
-!- puffin444 [62e3926e@gateway/web/freenode/ip.98.227.146.110] has quit [Quit: Page closed] | 22:32 | |
wiking | i has atlas! | 22:35 |
blackburn | I can haz atlaz | 22:35 |
shogun-buildbot | build #89 of deb3 - modular_interfaces is complete: Failure [failed test libshogun] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/89 blamelist: puffin444 <walke434@msu.edu> | 22:36 |
wiking | mmm i think i have to teach ./configure script to realise that i have atlas :) | 22:36 |
blackburn | isn't it detected? | 22:36 |
wiking | noup | 22:36 |
wiking | mmm nout even with --libs=/usr/local/lib --includes=/usr/local/include/ | 22:37 |
wiking | so i gotta check the configure.log | 22:37 |
wiking | doooh | 22:38 |
wiking | libalapack.so | 22:38 |
wiking | i has that but no liblapack :) | 22:38 |
blackburn | alapack?? | 22:38 |
wiking | i guess it's some stupid thingy by fbsd people... | 22:38 |
blackburn | may be they mean a la pack? | 22:39 |
-!- heiko [~heiko@host86-177-117-30.range86-177.btcentralplus.com] has joined #shogun | 22:39 | |
blackburn | a bit of french | 22:39 |
blackburn | :D | 22:39 |
-!- heiko [~heiko@host86-177-117-30.range86-177.btcentralplus.com] has quit [Client Quit] | 22:42 | |
blackburn | sonney2k: ping | 23:12 |
CIA-18 | shogun: Sergey Lisitsyn master * re510cf0 / src/shogun/machine/KernelMachine.cpp : Fixed uninitialized result vector - http://git.io/TJOtCQ | 23:34 |
shogun-buildbot | build #82 of deb1 - libshogun started, including [e510cf00af59d22dece9ce13030ccfadb1fb86d8] | 23:37 |
shogun-buildbot | build #82 of deb1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb1%20-%20libshogun/builds/82 | 23:42 |
shogun-buildbot | build #90 of deb3 - modular_interfaces started, including [e510cf00af59d22dece9ce13030ccfadb1fb86d8] | 23:42 |
shogun-buildbot | build #88 of deb2 - static_interfaces started, including [e510cf00af59d22dece9ce13030ccfadb1fb86d8] | 23:42 |
shogun-buildbot | build #10 of cyg1 - libshogun started, including [e510cf00af59d22dece9ce13030ccfadb1fb86d8] | 23:42 |
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 23:42 | |
blackburn | hey n4nd0 | 23:46 |
n4nd0 | hey blackburn, what's up? | 23:46 |
blackburn | fine, currently valgrinding examples :D | 23:46 |
blackburn | and you? | 23:46 |
n4nd0 | nice | 23:47 |
n4nd0 | I am going to check my code for data simulation for an HM-SVM example | 23:47 |
n4nd0 | a port from the HM-SVM toolbox | 23:48 |
shogun-buildbot | build #88 of deb2 - static_interfaces is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/88 | 23:50 |
n4nd0 | shogun-buildbot: good work :) | 23:51 |
n4nd0 | shogun-buildbot: dance | 23:51 |
shogun-buildbot | <(^.^<) | 23:51 |
shogun-buildbot | <(^.^)> | 23:51 |
shogun-buildbot | (>^.^)> | 23:51 |
shogun-buildbot | (7^.^)7 | 23:51 |
shogun-buildbot | (>^.^<) | 23:51 |
n4nd0 | haha I love that | 23:51 |
blackburn | making it no so depressive-causing | 23:51 |
blackburn | n4nd0: earlier bettyboo by rieck was here | 23:53 |
blackburn | we had a lof of fun :D | 23:53 |
blackburn | lot* | 23:53 |
n4nd0 | haha | 23:53 |
shogun-buildbot | build #10 of cyg1 - libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/10 | 23:53 |
n4nd0 | I have to check logs | 23:53 |
n4nd0 | today? | 23:53 |
blackburn | n4nd0: year ago | 23:53 |
n4nd0 | ah ok | 23:53 |
n4nd0 | who is rieck btw? | 23:54 |
blackburn | well he is in there :) http://jmlr.csail.mit.edu/papers/v11/sonnenburg10a.html | 23:55 |
blackburn | no? :D | 23:55 |
blackburn | sorry | 23:55 |
blackburn | well former colleague of sonney2k | 23:55 |
shogun-buildbot | build #90 of deb3 - modular_interfaces is complete: Failure [failed compile csharp_modular] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/90 blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com> | 23:56 |
blackburn | ehm | 23:59 |
blackburn | shogun-buildbot: wtf? | 23:59 |
--- Log closed Sun Jul 08 00:00:17 2012 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!