00:07 < abernauer[m]> Got like four or five more lines of code to write and then I have a working draft R script, for the cf examples that's similar to zoq Python notebook. Just need to transfer it into a Jupyter notebook format, but that should be easy enough.

00:39 < shrit[m]> @rcurtin, it seems that there is nothing we can do for VS15.

00:40 < shrit[m]> There is not allocated RAM for the compilation process.

00:40 < shrit[m]> enough

01:44 < rcurtin> shrit[m]: agreed

02:10 < rcurtin> I left a comment on the PR, I guess we are just about done then

04:20 < yashwants19[m]> [abernauer](https://matrix.to/#/@abernauer:matrix.org) preprocess_describe works fine for me.

04:23 < yashwants19[m]> I am not sure, what bug you are facing.

04:23 < yashwants19[m]> Can you please provide me more detail.

10:16 < jeffin143[m]> @yashwants19:matrix.org: Don't tell me you are going to pull of java binding as well 😂

10:30 < yashwants19[m]> Yeah i am looking forward to it. :)

10:32 < yashwants19[m]> You won't believe i even tried to build octave-bindings. 🤣

10:32 < yashwants19[m]> https://github.com/Yashwants19/mlpack-r/tree/master/src/mlpack/bindings/octave

10:34 < yashwants19[m]> There I am almost complete with octave_utils.

10:34 < yashwants19[m]> Except passing model to/from octave.

11:38 < rcurtin> yashwants19[m]: awesome!

11:42 < yashwants19[m]> :)

11:46 < RishabhGarg108Gi> Hii everyone, I have been trying to build mlpack since last 3 days. I have tried everything I could but still I am getting out of memory errors again and again while building the tests. I don't know what to do now. Please help is required.

11:48 < zoq> RishabhGarg108Gi: You are building under Linux correct?

11:49 < zoq> RishabhGarg108Gi: Can you post your cmake command?

11:49 < zoq> RishabhGarg108Gi: Also what are your specs?

11:49 < RishabhGarg108Gi> i am using $ cmake -D DEBUG=ON -D PROFILE=ON ../ and make -j1

11:51 < RishabhGarg108Gi> Linux 20.04. 8Gb RAM, i5 8th Gen with 8 cores of 1.6GHz

11:51 < RishabhGarg108Gi> Sorry Ubuntu 20.04

11:51 < zoq> RishabhGarg108Gi: So you could disable debug symbols -DDEBUG=OFF, that is the default setting, same for profile.

11:52 < zoq> RishabhGarg108Gi: If you don't need the cli executables turn them off as well -DBUILD_CLI_EXECUTABLES=OFF

11:53 < zoq> RishabhGarg108Gi: I guess you build the git master brnach? In which case no need to disable bindings support.

11:53 < zoq> RishabhGarg108Gi: If you build against the latest stable version you can turn of the python bindings -DBUILD_PYTHON_BINDINGS=OFF

11:54 < zoq> RishabhGarg108Gi: You can do the same for the other bindings, but as I said if you build the master branch no need to disable them.

11:55 < RishabhGarg108Gi> Actually my targets are to first use the library to get an overview of what it is and how things are done using the library. Then I would try to do something to contribute to it. And I work in python and C++. So, could you please suggest me cmake options according to these considerations.

11:55 < zoq> RishabhGarg108Gi: 6GB should be enough to build mlpack.

11:56 < RishabhGarg108Gi> I was also building the man pages. So do i really need to build them as well ?

12:03 < RishabhGarg108Gi> Actually the point at which I am running out of memory is when it starts to build the tests. So, is there any way by which I could build all the other things and the tests separately ? Would this strategy work ?

12:07 < zoq> RishabhGarg108Gi: The is no option in place for that, and 6GB free memory or SWAP should be enough.

12:08 < zoq> RishabhGarg108Gi: Do you build the git master branch?

12:09 < RishabhGarg108Gi> Yes @zoq . I am building from the master branch

12:09 < RishabhGarg108Gi> Actually I just ran cmake -D BUILD_TESTS=OFF -D BUILD_CLI_EXECUTABLES=OFF ../ and it completed without any issue.

12:09 < zoq> RishabhGarg108Gi: Let's see if "cmake -DBUILD_CLI_EXECUTABLES=OFF -DBUILD_TESTS=OFF .." works first.

12:10 < zoq> RishabhGarg108Gi: Okay, let's enable the tests next and see if that works as well.

12:11 < RishabhGarg108Gi> How do I enable tests ?? Should I like rerun the cmake command without -D BUILD_TESTS=OFF ?

12:11 < zoq> RishabhGarg108Gi: Yes, or -DBUILD_TESTS=ON

12:12 < RishabhGarg108Gi> okay. Lemme give it a try. Hoping that it would work :)

12:56 < RishabhGarg108Gi> It got built sucessfully🥳🥳. Thanks a lot @zoq .

12:57 < RishabhGarg108Gi> So is the build process complete now ??

13:13 < rcurtin> shrit[m]: I think I lost you on the call

14:56 < abernauer[m]> yashwants19 There isn't a bug just didn't notice the verbose argument was required.

14:57 < rcurtin> verbose shouldn't be a required option; are you sure it's required?

14:59 < abernauer[m]> When I ran the function yesterday without verbose = TRUE I didn't get any output just an empty list in R.

15:00 < rcurtin> well, that sounds expected... the argument isn't required, but it won't print anything unless you specify `verbose`

15:00 < rcurtin> that was preprocess_describe, I think?

15:00 < rcurtin> in this case, it sounds like the "right" thing happened

15:01 < abernauer[m]> yeah that was the function. Bindings work fine from my limited use so far.

15:01 < rcurtin> ok, awesome

15:01 < rcurtin> if the documentation doesn't point out that nothing will be printed unless `verbose` is specified, maybe we should update the docs for that one?

15:03 < abernauer[m]> Yeah that's probably a good idea. In the quick start example the verbose argument is set.

15:58 < jeffin143[m]> yashwants19: you are the real gangster 😂😂

16:36 < RishabhGarg108Gi> Also I have a suggestion that as the complete mlpack build is very memory intensive and many people with low end devices might be facing memory issues, so we could mention that in Installation docs that in such case people can build tests separately from the whole build. Coz when I built them separately then at max it was taking only 4Gb RAM and while building the complete project it was taking more than 7.5Gb RAM.

21:21 < jeffin143[m]> Is azure not able to handle our ?? Builds

21:21 < jeffin143[m]> Now publish artificates are failing across many pipelines

21:23 < zoq> jeffin143[m]: Looks like Azure has some issues.

21:24 < jeffin143[m]> Ahaa I see

21:24 < jeffin143[m]> There are so many projects using azure , how do they handle so much

21:24 < jeffin143[m]> I means it would be sooooooooo much of builds running

21:26 < shrit[m]> Actually Azure has some servers issue now in europe

21:27 < shrit[m]> So I do not know what happend to cereal build, I think it is building now