--- Log opened Mon Feb 19 00:00:08 2018 | ||
-!- HeikoS [~heiko@p4FED48C7.dip0.t-ipconnect.de] has quit [Ping timeout: 240 seconds] | 00:16 | |
-!- travis-ci [~travis-ci@ec2-54-226-190-62.compute-1.amazonaws.com] has joined #shogun | 01:18 | |
travis-ci | it's Heiko Strathmann's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/343149166 | 01:18 |
---|---|---|
-!- travis-ci [~travis-ci@ec2-54-226-190-62.compute-1.amazonaws.com] has left #shogun [] | 01:18 | |
-!- travis-ci [~travis-ci@ec2-54-226-190-62.compute-1.amazonaws.com] has joined #shogun | 01:40 | |
travis-ci | it's Heiko Strathmann's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/343149166 | 01:40 |
-!- travis-ci [~travis-ci@ec2-54-226-190-62.compute-1.amazonaws.com] has left #shogun [] | 01:40 | |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4180 opened by luisffranca | 07:23 |
-!- HeikoS [~heiko@p2003005BE30F56007477AD964DEAB239.dip0.t-ipconnect.de] has joined #shogun | 08:41 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 08:41 | |
-!- HeikoS [~heiko@p2003005BE30F56007477AD964DEAB239.dip0.t-ipconnect.de] has quit [Ping timeout: 265 seconds] | 08:46 | |
-!- HeikoS [~heiko@p4FED4CD2.dip0.t-ipconnect.de] has joined #shogun | 09:19 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 09:19 | |
@HeikoS | wiking: jojo | 10:01 |
@HeikoS | wiking: so I fixed that template in typedo | 10:01 |
@HeikoS | wiking: I tried to fix the put_scalar, via putting a third empty template parameter, but swig then started complaining | 10:02 |
@HeikoS | so not sure how to rename that | 10:02 |
@HeikoS | wiking: if you could check that, would be cool! | 10:02 |
@HeikoS | wiking: also putting the %extend business if you have some time | 10:02 |
@wiking | HeikoS, hi sorry was out in an mri till now | 10:21 |
@wiking | will check on it sap | 10:21 |
@wiking | *asap | 10:21 |
@HeikoS | wiking: cool thx | 10:22 |
@HeikoS | hows the knee? | 10:23 |
durovo | HeikoS: hello, you had suggested that I use shorter BSD headers in this pull request <https://github.com/shogun-toolbox/shogun/pull/4166>, | 10:25 |
@HeikoS | durovo: hi yes | 10:25 |
durovo | Is this the one that you were talking about?: This software is distributed under BSD 3-clause license (see LICENSE file) | 10:25 |
@HeikoS | yep that one :) | 10:25 |
durovo | Got it, thanks | 10:25 |
@wiking | HeikoS, dunno i guess the mri will tell :) | 10:26 |
@wiking | but this was for some other stuff | 10:27 |
@HeikoS | ah I see | 10:27 |
@HeikoS | ok fingers crossed :) | 10:27 |
@wiking | HeikoS, btw we have an error | 10:30 |
@wiking | do you know somebody that is an R guru-ish | 10:31 |
@wiking | in R we have an 'interface definition' file in R autogenerated by swig | 10:31 |
@wiking | which is all good... | 10:31 |
@wiking | R has this option that you can make a binary format of it, so that way the whole function-lookup will be faster | 10:31 |
@wiking | now the problem is that it seems that there's some sort of a bug when the native R code is being translated into this binary RData format | 10:32 |
@wiking | this is all done in RE | 10:32 |
@wiking | *R so this is out of our control, this is some R magic | 10:32 |
@wiking | but idk how one would be able to at least report/debug this | 10:33 |
@wiking | so if you know any R hackermaster | 10:33 |
@wiking | lemme know | 10:33 |
@wiking | this would fix almost all the failing unit test in that branch | 10:33 |
-!- surya [~surya@106.77.179.69] has joined #shogun | 10:42 | |
@HeikoS | wiking: ok, if I meet one Ill forwrad him | 10:53 |
@HeikoS | but I dont know one :( | 10:53 |
@wiking | seems really that's the root cause of all the remaining bug :( | 10:54 |
@wiking | which is weird though as this is an R feature | 10:54 |
@wiking | which then would mean that R has a bug | 10:54 |
@wiking | but i've just realised that our insrtallation magic for R would only ever work | 10:54 |
@wiking | on linux | 10:54 |
@wiking | "for f in *.so" | 10:55 |
@HeikoS | aye | 10:58 |
@HeikoS | weird | 10:58 |
@HeikoS | wiking: I just tried putting a void put_vector_or_matrix(const std::string& name, SGMatrix<T> value) into shogun.i, but typemap is not matched :( | 10:59 |
@wiking | typemap needs fixing | 11:01 |
@wiking | that's for sure | 11:01 |
@wiking | there's many of these hiccups | 11:01 |
@wiking | https://www.facebook.com/EurosportHU/videos/1190267511107520/ | 11:04 |
@wiking | \o/ | 11:04 |
@wiking | hhahahahah | 11:04 |
@wiking | this person is a national treasure :D | 11:04 |
@HeikoS | cant see it, facebook wants me to log in | 11:04 |
@wiking | just search for Swaney | 11:06 |
@wiking | :) | 11:06 |
@wiking | it's like *EVERYWHER* in the news :P | 11:06 |
@wiking | basiscally trolled halfpipe on the olympics | 11:07 |
@wiking | HeikoS, https://twitter.com/Eurosport_IT/status/965433099248021505?ref_src=twsrc%5Etfw&ref_url=https%3A%2F%2Findex.hu%2Fsport%2F2018%2Fteliolimpia%2F2018%2F02%2F19%2Fime_a_magyar_swaney_csuszasa_phjongcshangban%2F | 11:10 |
@HeikoS | lol :) | 11:11 |
@wiking | sorry not this one:( | 11:11 |
@wiking | this is waaaay too good | 11:11 |
@wiking | the halfpipe that swaney did | 11:11 |
@wiking | was like if you would just go to the olympics | 11:11 |
@wiking | and they give you a pair of skies :) | 11:11 |
@wiking | and tell you to go down on that halfpipe | 11:11 |
@sukey | [https://github.com/shogun-toolbox/shogun] Wiki page: GSoC-follow-up-blog-posts edited by karlnapf | 11:15 |
@wiking | https://shitpost.plover.com/m/math.jobs-you-dont-want.html | 11:15 |
@wiking | lol | 11:15 |
@HeikoS | haha! | 11:24 |
@sukey | [https://github.com/shogun-toolbox/shogun] New commit https://github.com/shogun-toolbox/shogun/commit/a8be9e2167ddc2bf2091b04f480f89104abb33a8 by vigsterkr | 11:31 |
@wiking | HeikoS, btw we've talked with lisitsyn regarding java; we are probably gonna drop both jblas and ujmp and use commons math | 11:41 |
@wiking | http://commons.apache.org/proper/commons-math/ | 11:42 |
-!- travis-ci [~travis-ci@ec2-54-226-190-62.compute-1.amazonaws.com] has joined #shogun | 11:42 | |
travis-ci | it's Heiko Strathmann's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/343149166 | 11:42 |
-!- travis-ci [~travis-ci@ec2-54-226-190-62.compute-1.amazonaws.com] has left #shogun [] | 11:42 | |
@HeikoS | wiking: yep that would be good | 11:45 |
@wiking | btw i have forgotten to talk about tht with you | 11:52 |
@wiking | there's this project | 11:52 |
@wiking | https://openmrs.org/ | 11:52 |
@wiking | i've met some of the devs here in zurich at the gsoc meetup | 11:52 |
@wiking | they were open to have a joint gsoc project - if we find somebody | 11:53 |
@wiking | to add an ML feature to the system | 11:53 |
-!- travis-ci [~travis-ci@ec2-54-161-62-69.compute-1.amazonaws.com] has joined #shogun | 11:56 | |
travis-ci | it's Heiko Strathmann's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/343149166 | 11:56 |
-!- travis-ci [~travis-ci@ec2-54-161-62-69.compute-1.amazonaws.com] has left #shogun [] | 11:56 | |
-!- travis-ci [~travis-ci@54.205.26.29] has joined #shogun | 11:57 | |
travis-ci | it's Viktor Gal's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/343302905 | 11:57 |
-!- travis-ci [~travis-ci@54.205.26.29] has left #shogun [] | 11:57 | |
lisitsyn | HeikoS: jo | 12:19 |
lisitsyn | wiking: INTEGRATE!!! | 12:20 |
@HeikoS | lisitsyn: jojo | 12:29 |
lisitsyn | HeikoS: can I help you somehow? | 12:31 |
@HeikoS | lisitsyn: yes | 12:31 |
@HeikoS | there are two problems I have righ tnow | 12:32 |
@HeikoS | check SGObject.h in typedo branch | 12:32 |
@HeikoS | lisitsyn: there you see put_vector_or_matrix | 12:35 |
@HeikoS | I would like to move that to shogun.i behind an %extend CSGObject | 12:35 |
@HeikoS | but the problem is that the typemap only matches if SGMatrix<float64_t> itself is the parameter | 12:35 |
lisitsyn | HeikoS: why %extend? | 12:35 |
@HeikoS | because it is only relevant for java atm | 12:36 |
@HeikoS | should not be part of the cpp lib | 12:36 |
@HeikoS | but you see | 12:38 |
@HeikoS | check the implementation in the SGObject.cpp | 12:38 |
@HeikoS | I specialise it for SGMatrix<float64_t> (since I need to access the passed value) | 12:38 |
@HeikoS | so would like to put all that into swig shogun.i | 12:39 |
-!- travis-ci [~travis-ci@ec2-54-161-62-69.compute-1.amazonaws.com] has joined #shogun | 12:40 | |
travis-ci | it's Viktor Gal's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/343302905 | 12:40 |
-!- travis-ci [~travis-ci@ec2-54-161-62-69.compute-1.amazonaws.com] has left #shogun [] | 12:40 | |
lisitsyn | HeikoS: put_scalar? | 12:43 |
lisitsyn | should be a part of cpp? | 12:43 |
@HeikoS | lisitsyn: yes, but renamed to "put" | 12:43 |
@HeikoS | I tried: | 12:43 |
@HeikoS | just renaming it -> compile error | 12:43 |
@HeikoS | adding a third template argument whcih defaults to void -> compiles, but swig says "undefined template" in shogun.i | 12:44 |
@HeikoS | that is issue number 2 :) | 12:44 |
@HeikoS | fixing any of those (without breaking the build, which is green) would be appreciated :) | 12:45 |
@HeikoS | lisitsyn: thing for the put_scalar is | 12:45 |
@HeikoS | atm, swig renamed this to "put" and the swig interfaces use this put method | 12:45 |
lisitsyn | HeikoS: sooo you rename put_scalar to put | 12:45 |
lisitsyn | and all works? | 12:45 |
@HeikoS | compile error | 12:46 |
@HeikoS | it conflicts with put from above | 12:46 |
@HeikoS | ah wait | 12:46 |
@HeikoS | I guess need to remove arithmetic template arugments from the put above? | 12:48 |
@HeikoS | so change std::enable_if<!std::is_base_of<CSGObject, typename std::remove_pointer<T>::type>::value, T>::type> | 12:48 |
@HeikoS | to something like | 12:48 |
@HeikoS | typename std::enable_if<std::is_base_of<SGReferencedData, T>::value, T>::type> | 12:48 |
@HeikoS | that makes sense? | 12:48 |
@HeikoS | still clashed -> compile error | 12:49 |
@HeikoS | only compiles if I add another template arugment to the put_scalar | 12:50 |
@HeikoS | but then | 12:50 |
@HeikoS | %template(put) CSGObject::put<int32_t, int32_t, void>; | 12:50 |
@HeikoS | doesn't work anymore :( | 12:50 |
@HeikoS | lisitsyn: so no idea | 12:57 |
lisitsyn | HeikoS: I am a bit lost | 12:57 |
lisitsyn | HeikoS: let me check the code again a bit later today in a quiet | 12:58 |
@HeikoS | lisitsyn: you can easily get unlost via renaming put_scalar to "put" and then try to compile :) | 12:58 |
lisitsyn | :) | 12:58 |
@HeikoS | ok sure | 12:58 |
lisitsyn | I am basically not into the problem you're solving right now | 12:58 |
lisitsyn | I guess it is swig and typemaps | 12:58 |
@HeikoS | I think this is c++ | 12:58 |
@HeikoS | actually no it is swig | 12:59 |
@HeikoS | but whatever it is | 12:59 |
@HeikoS | what is in the branch at least works with the basic example | 12:59 |
@HeikoS | and all targets | 12:59 |
@HeikoS | but it is not uniform | 12:59 |
@HeikoS | as swig uses put_scalar type dispatcher and cpp doesnt | 13:00 |
@HeikoS | Ill hav elunch now | 13:00 |
@wiking | yo | 13:04 |
@wiking | sorry was out for lunch | 13:04 |
@HeikoS | wiking: | 13:06 |
@HeikoS | /home/heiko/git/shogun/build/src/interfaces/java/shogun.i:125: Error: Template 'put_scalar' undefined. | 13:06 |
@HeikoS | any idea? | 13:06 |
@HeikoS | %template(put) CSGObject::put<SGVector<float64_t>, SGVector<float64_t>, void>; | 13:07 |
@HeikoS | template <typename T, typename T2 = typename std::enable_if<std::is_base_of<SGReferencedData, T>::value, T>::type, typename T3=void> | 13:07 |
@HeikoS | void put(const std::string& name, T value) | 13:07 |
@wiking | why 3 template arg? | 13:07 |
@HeikoS | to make compiler happy | 13:07 |
@wiking | oic | 13:07 |
@HeikoS | to rename put_scalar | 13:07 |
@HeikoS | and avoid clash | 13:07 |
@HeikoS | compiles and cpp works | 13:07 |
@HeikoS | but I dont get the template undefined by swig | 13:07 |
@wiking | Error: Template 'put_scalar' undefined | 13:08 |
@wiking | but it complains that it doesn't have put_scalar :) | 13:08 |
@wiking | i mean the given error is that there's no put_scalar in .h | 13:08 |
@wiking | can you copy the *latest* error? | 13:08 |
@HeikoS | ah | 13:09 |
@HeikoS | screw | 13:09 |
@HeikoS | yes I had a copy paste error | 13:09 |
@HeikoS | 1 sec | 13:09 |
@wiking | :) | 13:09 |
@HeikoS | wiking: shogun.i:125: Error: Template 'put' undefined. | 13:12 |
@HeikoS | thats the error | 13:12 |
@HeikoS | with the methods as above | 13:12 |
@HeikoS | and actually | 13:12 |
@HeikoS | /home/heiko/git/shogun/src/shogun/base/SGObject.h:454: Warning 302: Identifier 'put' redefined (ignored), | 13:12 |
@wiking | ah | 13:12 |
@HeikoS | so I guess swig ignores one of the put | 13:12 |
@wiking | that is shito | 13:12 |
@HeikoS | yes | 13:12 |
@HeikoS | so swig in fact either needs one put, or a rename | 13:13 |
@HeikoS | maybe we have to split the interface then | 13:14 |
@HeikoS | I mean between cpp and swig | 13:14 |
@HeikoS | so that cpp can be strongly typed, no dispatching, and swig uses the put_scalar, but renamed | 13:14 |
@HeikoS | wiking: thoughts? | 13:14 |
@wiking | just trying to read the swig docs | 13:15 |
@HeikoS | kk | 13:15 |
@wiking | about vargs | 13:15 |
@wiking | for template | 13:15 |
@HeikoS | Im going to lunch | 13:15 |
@HeikoS | let me know what you find :) | 13:15 |
@wiking | just trying to find a cached site | 13:15 |
@wiking | as it seems the one currently is still down | 13:15 |
-!- HeikoS [~heiko@p4FED4CD2.dip0.t-ipconnect.de] has quit [Ping timeout: 276 seconds] | 13:20 | |
-!- travis-ci [~travis-ci@ec2-54-226-190-62.compute-1.amazonaws.com] has joined #shogun | 13:24 | |
travis-ci | it's Heiko Strathmann's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/343149166 | 13:24 |
-!- travis-ci [~travis-ci@ec2-54-226-190-62.compute-1.amazonaws.com] has left #shogun [] | 13:24 | |
-!- surya1 [~surya@106.76.22.176] has joined #shogun | 13:37 | |
-!- surya [~surya@106.77.179.69] has quit [Ping timeout: 240 seconds] | 13:39 | |
-!- HeikoS [~heiko@p4FED4CD2.dip0.t-ipconnect.de] has joined #shogun | 13:44 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 13:44 | |
@HeikoS | wiking: jo | 13:45 |
@HeikoS | any insights? | 13:45 |
@HeikoS | wiking: lisitsyn do you see this https://www.numfocus.org/numfocus-awards-small-development-grants-to-projects/ | 13:46 |
@HeikoS | you have some ideas on that? | 13:46 |
@wiking | HeikoS, none yet | 13:46 |
@wiking | HeikoS, regarding numfocus award | 13:46 |
@wiking | any of the interesting/hard gsoc project could go there | 13:47 |
@wiking | see for example based on pytables, we could say shogun + arrow | 13:47 |
@wiking | that's something similar | 13:47 |
@HeikoS | so that's the topic | 13:47 |
@HeikoS | big question is how we would spend the money | 13:47 |
@wiking | :D | 13:48 |
@wiking | https://www.youtube.com/watch?v=5LLS9ZwRo3k | 13:48 |
@HeikoS | some other projects had the "we spend 2-3 weeks together" idea | 13:48 |
@HeikoS | we could do something like that | 13:48 |
@HeikoS | like another hackathon | 13:49 |
@HeikoS | but longer | 13:49 |
@HeikoS | fewer peeps and longer | 13:49 |
@wiking | oh good that you are saying | 13:49 |
@wiking | :D | 13:49 |
@HeikoS | "To do this, two core developers (Richard Gowers (RG) & Tyler Reddy (TR)) will visit Arizona State University to work on the issue full-time for 2 weeks. " | 13:50 |
@wiking | :> | 13:50 |
@HeikoS | so maybe rather that phrasing it hackathon | 13:51 |
@HeikoS | we would call it developer sprint :D | 13:51 |
@HeikoS | one other project hosted a public open workshop | 13:51 |
@wiking | HeikoS, just sent out the email we discussed the other day | 13:52 |
@HeikoS | sympy just had a "fulltime work" thing | 13:52 |
@HeikoS | cool will reply | 13:53 |
@HeikoS | wiking: I think maybe the tags integration might be something | 13:54 |
@HeikoS | for the nf | 13:55 |
@wiking | yeah | 13:55 |
@HeikoS | there is so many details that require core devs to discuss | 13:55 |
@HeikoS | like the typedo branch | 13:55 |
@wiking | mmm i mean | 13:55 |
@wiking | rather do | 13:55 |
@wiking | than discuss :) | 13:55 |
@wiking | but yeah | 13:55 |
@HeikoS | sry yes thats what I mean | 13:55 |
@HeikoS | like: we are sitting together and trying to make it work | 13:55 |
@HeikoS | rather than discussing | 13:55 |
@HeikoS | or in addition rather :) | 13:56 |
@wiking | k | 13:56 |
@wiking | i mean as said | 13:56 |
@wiking | any of the 'really good to have' | 13:56 |
@wiking | gsoc project | 13:56 |
@wiking | would be good topic imo | 13:56 |
@wiking | HeikoS, it's a rolling deadline? | 13:56 |
@wiking | or is there like a hardcore deadline for this | 13:56 |
@HeikoS | it is in 1-5 weeks | 13:56 |
@wiking | ? | 13:56 |
@HeikoS | 1.5 | 13:56 |
@wiking | mmmm | 13:56 |
@HeikoS | so I am thinking | 13:56 |
@wiking | ok then we should do it | 13:56 |
@wiking | i thought that if it's a rolling deadline | 13:57 |
@HeikoS | we should probably work on stuff that is really needed for some time now | 13:57 |
@HeikoS | but on the other hand | 13:57 |
@wiking | then we could wait till gsoc applications :) | 13:57 |
@HeikoS | doing sexy new stuff is more likely to get in | 13:57 |
@HeikoS | but for practical reasons, I would try to focus on things that gsoc students cant really do | 13:57 |
@HeikoS | I see | 13:57 |
@HeikoS | no , need to write it soon | 13:57 |
@HeikoS | so maybe Ill write something on: | 13:58 |
@HeikoS | dev-sprint on: integrating tags, fixing issues on the side (typemaps, swig dependencies), and then making a release with the new API | 13:59 |
@HeikoS | wiking: so then we have deliverable "release" | 13:59 |
@wiking | yeah | 13:59 |
@HeikoS | and we have some initial work done already | 13:59 |
@wiking | that'd be a good thing as well | 13:59 |
@HeikoS | and then we do a SF style meeting where we chill out for a week in a flat | 14:00 |
@HeikoS | and make things happen | 14:00 |
@HeikoS | it is $3000 | 14:00 |
@HeikoS | so could invite Sergey | 14:01 |
@HeikoS | lisitsyn: you have a week some time this year? :D | 14:01 |
@wiking | HeikoS, would be the best before gsoc is actually started | 14:01 |
@HeikoS | yeah | 14:01 |
@wiking | so that way there's a clean release | 14:01 |
@HeikoS | wiking: ok cool I will cook something up | 14:01 |
@HeikoS | wiking: now on that put_scalar thing | 14:02 |
@HeikoS | I was thinking: why don't we move "put_scalar" to swig.i | 14:02 |
@HeikoS | and then just renmae | 14:02 |
@sukey | [https://github.com/shogun-toolbox/shogun] New commit https://github.com/shogun-toolbox/shogun/commit/bc6fafef27859e9a3993c439e041476bec70a48a by karlnapf | 14:03 |
@wiking | HeikoS, i'm not so sure if i follow? | 14:04 |
@wiking | or you would move that is_arithmetic | 14:05 |
@wiking | to be not see-able by swig | 14:05 |
@wiking | and do another hax0r stuff with swig? | 14:05 |
@HeikoS | thats not what I meant, but what do you mean there? | 14:05 |
@HeikoS | wiking: so my thinking was this | 14:05 |
@HeikoS | wiking: we have a set of swig-put (like put scalar), they are not part of the cpp library | 14:06 |
@HeikoS | and the cpp library just has a strongly typed standard put | 14:06 |
@HeikoS | since in cpp, we dont need the dispatching | 14:06 |
@HeikoS | wiking: which basically means moving the put_scalar (which is not used by cpp examples atm), to the swig.i | 14:07 |
@HeikoS | and then conceptually, it is clear that swig has some additional stuff for "put" | 14:07 |
-!- surya1 [~surya@106.76.22.176] has left #shogun [] | 14:24 | |
@HeikoS | wiking: lisitsyn http://www.pymvpa.org/ | 14:25 |
@HeikoS | they have shogun as a backend | 14:25 |
@sukey | [https://github.com/shogun-toolbox/shogun] New commit https://github.com/shogun-toolbox/shogun/commit/339a3a5ce09b06af5e1a64ea6737c494eca5a944 by karlnapf | 14:28 |
@wiking | HeikoS, yes if you talk about the lmnn | 14:45 |
@wiking | then indeed | 14:45 |
@wiking | they are the only ones :D | 14:45 |
@HeikoS | svm as well | 14:46 |
@wiking | or is this the other one? | 14:46 |
@HeikoS | another one | 14:46 |
@wiking | ah ok so this is a second lib | 14:46 |
@wiking | ! | 14:46 |
@wiking | :D | 14:46 |
@wiking | pybrain | 14:46 |
@wiking | or whichever medical dataset munging tool | 14:46 |
@wiking | used lmnn | 14:46 |
@wiking | i'm up for moving eerything out | 14:47 |
@wiking | that is contaminating | 14:47 |
@wiking | the clean cpp | 14:47 |
@HeikoS | cool | 14:47 |
@HeikoS | doing it | 14:47 |
@wiking | api | 14:47 |
@wiking | we could even | 14:47 |
@HeikoS | just the only thing is | 14:47 |
@wiking | do more of that :) | 14:47 |
@wiking | but yeah it's a good start | 14:48 |
@HeikoS | I cannot make this typemap in put_vector_or_matrix work | 14:48 |
@HeikoS | but trying something right now | 14:48 |
@wiking | note that you could actually have an SwigSGObject wrapper | 14:48 |
@wiking | that is a c++ implementation | 14:48 |
@wiking | of the swig interface | 14:48 |
@HeikoS | yeah that might be a good idea for later | 14:48 |
@wiking | that wraps code between | 14:48 |
@wiking | sgobject | 14:48 |
@wiking | and swig interfaces | 14:48 |
@HeikoS | yep | 14:48 |
@wiking | bascically we could start adding | 14:48 |
@wiking | .cpp to swig | 14:48 |
@wiking | so offload some of the BS stuff | 14:49 |
@wiking | from the actual cpp | 14:49 |
@wiking | HeikoS, have you seen my openmrs stuff? | 14:49 |
@HeikoS | yes | 14:49 |
@HeikoS | just need to spend some more time looking | 14:49 |
@HeikoS | too many things atm | 14:49 |
@wiking | i mean it woudl be a good project | 14:50 |
@wiking | for the java interface | 14:50 |
@HeikoS | yeah | 14:50 |
@wiking | as the whoel project is java based | 14:50 |
@wiking | the sdk is in java | 14:50 |
@HeikoS | wiking: you know how to specialize templates in swig? | 14:50 |
@HeikoS | like template <class T> bla; | 14:50 |
@HeikoS | but then there is no implementation | 14:50 |
@wiking | ? | 14:50 |
@HeikoS | and I do something like template <> bla<float64_t> { ... } in swig | 14:50 |
@wiking | i mean you could even do | 14:50 |
@wiking | %template() Bla<float> | 14:50 |
@wiking | to 'instantiate' | 14:51 |
@HeikoS | ah really | 14:51 |
@HeikoS | cool | 14:51 |
@HeikoS | let me try | 14:51 |
@wiking | that's a valid thing in swig | 14:51 |
@HeikoS | ah now wait | 14:51 |
@HeikoS | but the code | 14:51 |
@HeikoS | it is not just instantiating | 14:51 |
@HeikoS | but writing special code for one particular T | 14:51 |
@wiking | in swig? | 14:52 |
@HeikoS | template <> bla(float64_t value) { my own impl using value } | 14:52 |
@HeikoS | y | 14:52 |
@wiking | you should be able to write any c++ code in swig | 14:53 |
@HeikoS | k | 14:53 |
@wiking | check here | 14:53 |
@wiking | http://www.swig.org/Doc3.0/CPlusPlus11.html | 14:53 |
@wiking | search for "7.3.10 Uniform" | 14:53 |
@wiking | you'll see how they do it | 14:53 |
@wiking | btw it's basically the last section of that page | 14:54 |
@wiking | so just scroll to the bottom | 14:54 |
@HeikoS | cool thx will check | 14:55 |
@HeikoS | wiking: see typedo btw | 14:55 |
@HeikoS | all thats missing is to move the put_vector_or_matrix to swig.i | 14:55 |
@HeikoS | I have to run now | 14:55 |
@HeikoS | talk later! | 14:55 |
-!- HeikoS [~heiko@p4FED4CD2.dip0.t-ipconnect.de] has quit [Quit: Leaving.] | 14:55 | |
-!- travis-ci [~travis-ci@ec2-54-161-62-69.compute-1.amazonaws.com] has joined #shogun | 15:18 | |
travis-ci | it's Heiko Strathmann's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/343352432 | 15:18 |
-!- travis-ci [~travis-ci@ec2-54-161-62-69.compute-1.amazonaws.com] has left #shogun [] | 15:18 | |
-!- travis-ci [~travis-ci@ec2-54-205-26-29.compute-1.amazonaws.com] has joined #shogun | 15:44 | |
travis-ci | it's Heiko Strathmann's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/343352432 | 15:44 |
-!- travis-ci [~travis-ci@ec2-54-205-26-29.compute-1.amazonaws.com] has left #shogun [] | 15:44 | |
-!- travis-ci [~travis-ci@54.161.62.69] has joined #shogun | 17:03 | |
travis-ci | it's Heiko Strathmann's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/343361248 | 17:03 |
-!- travis-ci [~travis-ci@54.161.62.69] has left #shogun [] | 17:03 | |
@sukey | [https://github.com/shogun-toolbox/shogun] New commit https://github.com/shogun-toolbox/shogun/commit/dc3e057c1eefe45809b8d8632f327219e76a08ce by karlnapf | 17:53 |
-!- HeikoS [~heiko@p2003005BE30F5600E5652FE8C6E26A20.dip0.t-ipconnect.de] has joined #shogun | 17:53 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 17:53 | |
@HeikoS | wiking: I moved everything to swig now | 17:53 |
@HeikoS | and it works (on my machine at least) | 17:53 |
@HeikoS | let's see what swig says | 17:53 |
@HeikoS | I think this is much nicer niw | 17:53 |
@HeikoS | even though cpp api is slightly different, but still close enough for meta examples to work | 17:54 |
@HeikoS | and I like it much more not polluting the libshogun with that stuff | 17:54 |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4085 merged by karlnapf | 18:12 |
@sukey | [https://github.com/shogun-toolbox/shogun] New commit https://github.com/shogun-toolbox/shogun/commit/bb7242762b900f5576ebf61e1c94a5bae7025339 by karlnapf | 18:12 |
-!- HeikoS [~heiko@p2003005BE30F5600E5652FE8C6E26A20.dip0.t-ipconnect.de] has quit [Quit: Leaving.] | 18:19 | |
-!- travis-ci [~travis-ci@ec2-54-161-62-69.compute-1.amazonaws.com] has joined #shogun | 19:04 | |
travis-ci | it's Heiko Strathmann's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/343453532 | 19:04 |
-!- travis-ci [~travis-ci@ec2-54-161-62-69.compute-1.amazonaws.com] has left #shogun [] | 19:04 | |
-!- travis-ci [~travis-ci@ec2-54-226-190-62.compute-1.amazonaws.com] has joined #shogun | 19:33 | |
travis-ci | it's Heiko Strathmann's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/343453532 | 19:33 |
-!- travis-ci [~travis-ci@ec2-54-226-190-62.compute-1.amazonaws.com] has left #shogun [] | 19:33 | |
-!- HeikoS [~heiko@p4FED4CD2.dip0.t-ipconnect.de] has joined #shogun | 19:33 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 19:33 | |
@HeikoS | wiking: you around? | 19:34 |
-!- HeikoS [~heiko@p4FED4CD2.dip0.t-ipconnect.de] has quit [Ping timeout: 276 seconds] | 19:49 | |
-!- HeikoS [~heiko@p4FED4CD2.dip0.t-ipconnect.de] has joined #shogun | 21:06 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 21:06 | |
@sukey | [https://github.com/shogun-toolbox/shogun] New commit https://github.com/shogun-toolbox/shogun/commit/33b855cce8cc962e356f45d44a4752f26abfc74c by karlnapf | 21:12 |
@HeikoS | lisitsyn: jojo | 21:40 |
-!- HeikoS [~heiko@p4FED4CD2.dip0.t-ipconnect.de] has quit [Read error: No route to host] | 21:49 | |
-!- HeikoS [~heiko@p4FED4CD2.dip0.t-ipconnect.de] has joined #shogun | 21:50 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 21:50 | |
-!- travis-ci [~travis-ci@ec2-54-166-201-245.compute-1.amazonaws.com] has joined #shogun | 22:12 | |
travis-ci | it's Heiko Strathmann's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/343538333 | 22:12 |
-!- travis-ci [~travis-ci@ec2-54-166-201-245.compute-1.amazonaws.com] has left #shogun [] | 22:12 | |
-!- travis-ci [~travis-ci@ec2-54-145-158-139.compute-1.amazonaws.com] has joined #shogun | 22:22 | |
travis-ci | it's Heiko Strathmann's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/343538333 | 22:22 |
-!- travis-ci [~travis-ci@ec2-54-145-158-139.compute-1.amazonaws.com] has left #shogun [] | 22:22 | |
-!- HeikoS [~heiko@p4FED4CD2.dip0.t-ipconnect.de] has quit [Ping timeout: 248 seconds] | 22:26 | |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4181 opened by sorig | 23:46 |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4181 | 23:50 |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4181 | 23:51 |
@sukey | [https://github.com/shogun-toolbox/shogun] Pull Request https://github.com/shogun-toolbox/shogun/pull/4181 synchronized by sorig | 23:59 |
--- Log closed Tue Feb 20 00:00:10 2018 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!