00:18 < zoq> PrinceGuptaGitte: If you store the result as json or xml it should provide some more output (meta data).

01:35 < rcurtin> vcpkg is updated with mlpack 3.2.2: https://github.com/microsoft/vcpkg/issues/10273#event-3117005449

01:35 < rcurtin> some really nice packaging efforts going on there, when I mentioned that the package was out of date, they already had a PR open

04:30 < JoelJosephGitter> how can i make "Log::Debug" print its output into the command line,

05:48 < SriramSKGitter[m> I think compiling with the debug switch on should do it, @joeljosephjin `cmake -DDEBUG=ON`

07:08 < jenkins-mlpack2> Project docker mlpack nightly build build #637: STILL FAILING in 2 hr 54 min: http://ci.mlpack.org/job/docker%20mlpack%20nightly%20build/637/

07:14 < JoelJosephGitter> @sriramsk1999 i was trying to do like "g++ -DDEBUG", but it didnt work... I tried adding "#define DEBUG" at the beginnning of the .cpp file, but it didn't work either. Anyhow, "cout" fits my purpose, for now. Thanks anyways.

10:04 < PrinceGuptaGitte> I tried altering serialize function of class to use `BOOST_SERIALIZATION_BASE_OBJECT_NVP` instead of `BOOST_SERIALIZATION_NVP` but that also doesn't help

11:26 < rcurtin> JoelJosephGitter: compile your program with -DDEBUG, and make sure that mlpack is compiled with debugging symbols also, and then you should get debugging output :)

13:52 < AbishaiEbenezerG> Hi mlpack! Apparently, there is an issue in the q-learning test module. Shal i go forward with it? I am particularly interested in the rl module, so i would really like to study this...

13:57 < Param-29Gitter[m> I tired using perf for performance analysis but when i run perf record and try to see perf report file comes out to be empty. ( I compiled using g++ test_decision_tree.cpp -fopenmp -lmlpack -g). I also tried with a c program that does not use mlpack and it works perfercly fine with it. How do I solve this?

14:10 < pickle-rick[m]> how could I make the markdown for just gmm and hmm locally?

14:45 < zoq> AbishaiEbenezerG: If you find an issue would be great if you open an issue on GitHUb oder a PR with a fix.

14:52 < pickle-rick[m]> hi, I'm in the process of making a pull request after pushing my work from a certain branch. When i opened github to "compare & pull request", my commit from another branch (I pushed the work on that branch in a previous pull request) also shows up. Is that ok?

15:20 < AbishaiEbenezerG> cool @zoq. I'll keep you updated on it...

15:47 < pickle-rick[m]> > hi, I'm in the process of making a pull request after pushing my work from a certain branch. When i opened github to "compare & pull request", my commit from another branch (I pushed the work on that branch in a previous pull request) also shows up. Is that ok?

15:47 < pickle-rick[m]> Sorry, but I got flustered and closed that pull request. Opened a new one with just the new changes (markdown pipe symbol fix) .

15:47 < pickle-rick[m]> > hi, I'm in the process of making a pull request after pushing my work from a certain branch. When i opened github to "compare & pull request", my commit from another branch (I pushed the work on that branch in a previous pull request) also shows up. Is that ok?

15:47 < pickle-rick[m]> Sorry, but I got flustered and closed that pull request. Opened a new one with just the new changes (markdown pipe symbol fix) .

15:48 < rcurtin> pickle-rick: no need to get flustered :) looks like you got it worked out though?

15:48 < pickle-rick[m]> Yeah seems like it. :)

15:48 < rcurtin> also, sorry I missed your message earlier, I don't think you can make the markdown for *just* gmm/hmm... I think you'd have to build it for all of them

15:50 < pickle-rick[m]> No worries. I totally understand you guys have a lot on your plates, especially right now :)

15:52 < favre49> Running make on mlpack requires me to have superuser access due to moving mlpack_export.hpp

15:53 < favre49> Am I doing something wrong?

15:56 < rcurtin> favre49: are you running just 'make' or 'make install'?

15:56 < rcurtin> definitely 'make' shouldn't need superuser, unless root owns some file in the build tree or something?

16:21 < favre49> Running make

16:21 < favre49> This is on my personal machine, so the build tree shouldn't be a problem

16:29 < favre49> Should I open an issue for this?

16:29 < favre49> Unless birm[m] knows something about this, I think there have been recent CMake changes

16:32 < favre49> Also, is there a difference in armadillo between setting seed to `std::time(NULL)` or setting it with `arma_rng::set_seed_random()`

17:41 < rcurtin> favre49: sure, maybe opening an issue makes it a bit easier to debug

17:42 < rcurtin> either that or pastebin output here or something, I'm still not sure I understand the issue

17:42 < rcurtin> I think that those two lines in Armadillo are equivalent, also

18:13 < pickle-rick[m]> Anyone know why the "julia_binding_test" keeps failing?

18:39 < rcurtin> pickle-rick[m]: it's on me, bad merge... I should have it solved later today

18:39 < rcurtin> not sure why the tests didn't pick it up in the PR?

18:40 < rcurtin> wait... it looks like the azure pipeline jobs never ran on that PR? anyway, I've reproduced it locally, shouldn't be too hard to fix

18:41 < pickle-rick[m]> rcurtin: oh alright.

18:44 < himanshu_pathak[> I waste most of the time in debugging do anyone have better idea instead of using gdb efficiently in less time span

18:52 < favre49> rcurtin You can see the error here - http://p.ip.fi/KGgH

18:57 < rcurtin> favre49: what are the permissions on build/include/ and build/include/mlpack?

18:57 < rcurtin> i.e. `ls -lh build/include/`

18:57 < rcurtin> that should tell us

18:57 < rcurtin> himanshu_pathak[: wish I had a better response, but at least I just got used to gdb and got a bit better with it over time. maybe there is another debugger out there you could use? I'm not familiar with anything but gdb though :(

19:01 < sreenik[m]> himanshu_pathak: There is a graphical gdb also available, might be somewhat less of a hassle if that is what you mean :)

19:07 < himanshu_pathak[> rcurtin: Is there any way that I can skip boost visitor implementation if I am running the program step by step because every time I try to run a program step by step every time boost::applyvisitor part also shows I want to skip that is there any way to do that. sreenik[m] Sure I will try it thanks for suggesting

19:13 < rcurtin> himanshu_pathak[: maybe use breakpoints to try and only stop at the part that you're interested in?

19:20 < himanshu_pathak[> rcurtin: Yes I use breakpoints but in some cases when you don't know where the problem might be it is quite overwhelming to see boost::applyvisitor output again and again.