08:34 < HarshVardhanKuma> Hi guys. I've noticed something interesting in the mlpack_cf method. The documentation states use of a 3-d input matrix but the examples shown use a csv file. How is anyone supposed to write a csv file for a 3-d matrix? Any special formats? Wouln't it be easy if the input were indeed a 2-d matrix - the row id as the user id and the column id as the item id?

08:35 < HarshVardhanKuma> * Hi guys. I've noticed something interesting in the mlpack_cf method. The documentation states use of a 3-d input matrix but the examples shown use of a csv file. How is anyone supposed to write a csv file for a 3-d matrix? Any special formats? Wouln't it be easy if the input were indeed a 2-d matrix - the row id as the user id and the column id as the item id?

13:21 < rcurtin[m]> HarshVardhanKumar (Harsh Kumar): exactly, in the 3d format the third dimension is just the rating that the user gave to the item. if your situation doesn't have any ratings, you might consider just using '1' as the value in the third dimension 👍️

13:21 < rcurtin[m]> also, can any maintainer give a quick review to https://github.com/mlpack/mlpack/pull/2968 ? just a simple semicolon fix so that the doxygen docs build again 😃

16:49 < jjb[m]> Ryan Curtin sorry I couldn’t make the earlier meeting. I’m heavily booked during the usual 9 - 5 and 6-8 PM until June 11th.

16:50 < jjb[m]> zoq I think we’re slatted to meet in about 10 minutes. Are we using the usual MLPACK zoom link?

16:52 < rcurtin[m]> no worries jjb ! NippunSharma (Nippun Sharma) and I just caught up about the state of the PR and talked through a few design questions 👍️

16:53 < zoq> jjb[m]: Yes, and yes let's use the zoom room

16:55 < jonpsy[m]> jjb: say4n zoq Meetings about to start, the usual spot.

16:57 < jjb[m]> <https://zoom.us/j/3820896170> <-- usual spot

17:56 < heisenbuugGopiMT> Hey, is the previous meeting still going on?

17:59 < jonpsy[m]> heisenbuug (Gopi M Tatiraju): it was a meeting about MOO proposal with my mentors. Not the official mlpack meeting :-)

18:00 < heisenbuugGopiMT> Yea, I know...

18:00 < heisenbuugGopiMT> Actually I also planned a meeting with my mentor at 23:30 IST, so I was just seeing if the room is available...

18:01 < shrit[m]> The room is empty

18:01 < heisenbuugGopiMT> Okay...

18:01 < shrit[m]> You can joing

19:33 < jonpsy[m]> Oh, btw when are we planning to move towards c++20? I think c++20 concepts and modules would be great for a heavily templatized library such as mlpack.

19:50 < rcurtin[m]> jonpsy: we are super slow with adopting new C++ standards because we want to be reverse-compatible for older Linux distros (ubuntu LTS, RHEL) and also older Visual Studio versions (which typically have broken or buggy standard support)

19:50 < rcurtin[m]> so... I would guess C++20 support comes into the picture in... like... 2025 at the earliest

19:51 < rcurtin[m]> RHEL7's current compiler is gcc 4.x, for what it's worth... I think RHEL8 is gcc 7 or 8?

19:56 < rcurtin[m]> looks like MSVC is still implementing some C++20 (and some C++17) support in VS2019... which... strangely, is still what the most recent version of Visual Studio is called despite it being 2021

19:57 < rcurtin[m]> anyway, I'm not opposed to upgrading, let's just make sure we don't leave anyone who's on an older system unable to compile mlpack :)

20:15 < rcurtin[m]> shrit: I finally looked into my hard drive that's been clicking for years... I'm not sure how it ever ended up in that system, but it wasn't being mounted anywhere and wasn't used at all so I just unplugged and removed it... I guess the reason I got away with leaving it in there so long is that it was never doing anything 😃

20:40 < Aakash-kaushikAa> @rcurtin free indicator if your system has power, if RGBs weren't enough 😛

23:07 < jonpsy[m]> <rcurtin[m] "so... I would guess C++20 suppor"> 2025 looks fairly optimistic given we're still stuck in c+11 which is two generations behind c++20 (~ . ~ )