#ste||ar on 2020-04-24 — irc logs at irclog.cct.lsu.edu

2020-02-24 20:46 hkaiser changed the topic of #ste||ar to: STE||AR: Systems Technology, Emergent Parallelism, and Algorithm Research | stellar.cct.lsu.edu | HPX: A cure for performance impaired parallel applications | github.com/STEllAR-GROUP/hpx | Buildbot: http://rostam.cct.lsu.edu/ | Log: http://irclog.cct.lsu.edu/ | GSoC: https://github.com/STEllAR-GROUP/hpx/wiki/Google-Summer-of-Code-%28GSoC%29-2020

00:03 nikunj has quit [Read error: Connection reset by peer]

00:03 nikunj has joined #ste||ar

00:08 nikunj has quit [Ping timeout: 256 seconds]

00:08 nikunj has joined #ste||ar

00:14 Nikunj__ has joined #ste||ar

00:14 Nikunj__ has quit [Remote host closed the connection]

00:15 Nikunj__ has joined #ste||ar

00:15 akheir1 has quit [Read error: Connection reset by peer]

00:15 akheir1 has joined #ste||ar

00:25 bita has quit [Ping timeout: 265 seconds]

00:25 shahrzad has quit [Remote host closed the connection]

00:25 shahrzad has joined #ste||ar

00:25 akheir1 has quit [Read error: Connection reset by peer]

00:26 akheir1 has joined #ste||ar

00:38 Nikunj__ has quit [Ping timeout: 260 seconds]

01:07 nikunj has quit [Ping timeout: 265 seconds]

01:09 nikunj has joined #ste||ar

01:17 nan77 has quit [Remote host closed the connection]

01:30 nikunj has quit [Ping timeout: 258 seconds]

01:32 bita has joined #ste||ar

01:33 nikunj has joined #ste||ar

01:53 shahrzad has quit [Ping timeout: 265 seconds]

02:01 nikunj has quit [Ping timeout: 250 seconds]

02:10 nikunj has joined #ste||ar

02:15 nikunj has quit [Ping timeout: 256 seconds]

02:16 hkaiser has quit [Quit: bye]

02:25 nikunj has joined #ste||ar

02:29 bita has quit [Quit: Leaving]

02:29 nikunj has quit [Ping timeout: 260 seconds]

02:37 nikunj has joined #ste||ar

02:42 shahrzad has joined #ste||ar

02:44 shahrzad has quit [Client Quit]

02:49 nikunj has quit [Ping timeout: 264 seconds]

02:53 nikunj has joined #ste||ar

03:06 nikunj has quit [Ping timeout: 256 seconds]

03:08 nikunj has joined #ste||ar

03:09 akheir1 has quit [Quit: Leaving]

03:15 nikunj has quit [Ping timeout: 264 seconds]

03:20 nikunj has joined #ste||ar

03:24 nikunj has quit [Ping timeout: 256 seconds]

03:27 nikunj has joined #ste||ar

03:38 nikunj has quit [Ping timeout: 240 seconds]

03:40 nikunj has joined #ste||ar

03:44 nikunj has quit [Ping timeout: 240 seconds]

03:56 nikunj has joined #ste||ar

04:09 nikunj has quit [Ping timeout: 265 seconds]

04:10 nikunj has joined #ste||ar

04:19 nikunj has quit [Ping timeout: 250 seconds]

04:20 nikunj has joined #ste||ar

04:24 nikunj has quit [Ping timeout: 256 seconds]

04:30 nikunj has joined #ste||ar

04:50 nikunj has quit [Ping timeout: 265 seconds]

04:51 nikunj has joined #ste||ar

05:00 nikunj has quit [Ping timeout: 256 seconds]

05:09 nikunj has joined #ste||ar

05:20 nikunj has quit [Ping timeout: 258 seconds]

05:20 nikunj has joined #ste||ar

05:24 karame_ has quit [Remote host closed the connection]

05:25 nikunj has quit [Ping timeout: 265 seconds]

05:28 nikunj has joined #ste||ar

05:35 nikunj has quit [Ping timeout: 264 seconds]

05:39 nikunj has joined #ste||ar

05:47 nikunj has quit [Ping timeout: 265 seconds]

05:50 nikunj has joined #ste||ar

05:54 nikunj has quit [Ping timeout: 260 seconds]

06:00 nikunj has joined #ste||ar

06:10 nikunj has quit [Ping timeout: 256 seconds]

06:15 nikunj has joined #ste||ar

06:29 nikunj has quit [Ping timeout: 258 seconds]

06:33 nikunj has joined #ste||ar

06:48 nikunj has quit [Ping timeout: 264 seconds]

06:58 nikunj has joined #ste||ar

07:07 nikunj has quit [Ping timeout: 265 seconds]

07:07 RostamLog has quit [Ping timeout: 260 seconds]

07:07 RostamLog has joined #ste||ar

07:09 nikunj has joined #ste||ar

07:13 nikunj has quit [Ping timeout: 265 seconds]

07:16 nikunj has joined #ste||ar

07:21 nikunj has quit [Ping timeout: 260 seconds]

07:24 Hashmi has joined #ste||ar

07:30 nikunj has joined #ste||ar

07:35 nikunj has quit [Ping timeout: 265 seconds]

07:36 nikunj has joined #ste||ar

08:08 mcopik has joined #ste||ar

08:09 mcopik has quit [Client Quit]

08:24 nikunj has quit [Ping timeout: 264 seconds]

08:26 nikunj has joined #ste||ar

08:30 nikunj has quit [Ping timeout: 256 seconds]

08:38 nikunj has joined #ste||ar

09:09 nikunj has quit [Ping timeout: 250 seconds]

09:10 nikunj has joined #ste||ar

09:15 nikunj has quit [Ping timeout: 264 seconds]

09:25 nikunj has joined #ste||ar

09:30 nikunj has quit [Ping timeout: 264 seconds]

09:33 Hashmi has quit [Quit: Connection closed for inactivity]

09:36 nikunj has joined #ste||ar

09:43 nikunj has quit [Ping timeout: 258 seconds]

09:54 nikunj has joined #ste||ar

09:58 nikunj has quit [Ping timeout: 256 seconds]

10:10 nikunj has joined #ste||ar

10:16 nikunj has quit [Ping timeout: 250 seconds]

10:18 nikunj has joined #ste||ar

10:35 nikunj has quit [Ping timeout: 264 seconds]

10:35 nikunj has joined #ste||ar

10:39 nikunj has quit [Ping timeout: 264 seconds]

10:40 nikunj has joined #ste||ar

10:42 weilewei has quit [Remote host closed the connection]

10:44 nikunj has quit [Ping timeout: 250 seconds]

10:47 nikunj has joined #ste||ar

10:54 nikunj has quit [Ping timeout: 240 seconds]

10:57 nikunj has joined #ste||ar

11:17 Nikunj__ has joined #ste||ar

11:17 nikunj97 has joined #ste||ar

11:22 Nikunj__ has quit [Ping timeout: 256 seconds]

11:34 hkaiser has joined #ste||ar

11:45 nikunj has quit [Ping timeout: 256 seconds]

11:53 nikunj has joined #ste||ar

12:00 nikunj has quit [Ping timeout: 256 seconds]

12:01 nikunj has joined #ste||ar

12:21 nikunj has quit [Ping timeout: 250 seconds]

12:25 nikunj has joined #ste||ar

12:29 nikunj97 has quit [Ping timeout: 258 seconds]

12:44 weilewei has joined #ste||ar

13:06 Hashmi has joined #ste||ar

13:37 nikunj has quit [Ping timeout: 265 seconds]

13:37 nikunj has joined #ste||ar

13:51 nan111 has joined #ste||ar

14:19 rtohid has joined #ste||ar

15:01 bita has joined #ste||ar

15:06 nikunj97 has joined #ste||ar

15:41 mdiers_1 has joined #ste||ar

15:43 mdiers_ has quit [Read error: Connection reset by peer]

15:43 mdiers_1 is now known as mdiers_

15:46 nan111 has quit [Remote host closed the connection]

15:52 nan11 has joined #ste||ar

16:36 karame_ has joined #ste||ar

16:38 weilewei has quit [Remote host closed the connection]

16:50 rtohid20 has joined #ste||ar

16:50 rtohid20 has left #ste||ar [#ste||ar]

16:51 rtohid68 has joined #ste||ar

16:51 weilewei has joined #ste||ar

16:53 rtohid has quit [Ping timeout: 240 seconds]

17:09 rtohid68 has left #ste||ar [#ste||ar]

17:50 nikunj has quit [Read error: Connection reset by peer]

17:51 weilewei has quit [Remote host closed the connection]

17:56 rtohid has joined #ste||ar

18:45 karame_ has quit [Remote host closed the connection]

18:51 nan11 has quit [Remote host closed the connection]

18:53 kale_ has joined #ste||ar

18:54 kale_ has quit [Client Quit]

19:05 karame_ has joined #ste||ar

20:14 weilewei has joined #ste||ar

20:16 Hashmi has quit [Quit: Connection closed for inactivity]

20:24 <weilewei> hkaiser why am I getting this compilation error, when I try to pass barrier by reference to hpx::async function? error log: https://gist.github.com/weilewei/5c10873d1410eca3341ed8042307c19d code:

20:24 <weilewei> https://github.com/STEllAR-GROUP/DCA/blob/a0ea6a4daf4e029e84f3a085b13d3f37b8782f79/include/dca/phys/dca_step/cluster_solver/stdthread_qmci/stdthread_qmci_cluster_solver.hpp#L175

20:25 <hkaiser> weilewei: use std::ref(barrier)

20:26 <weilewei> hkaiser, it compiled! what's the magic (logic) here?

20:28 <hkaiser> you want to pass your barrier by reference, using std::ref() tells async to do that

20:28 <hkaiser> otherwise it will try to copy/move the argument

20:29 <weilewei> hkaiser I see, thanks

21:34 rtohid has left #ste||ar [#ste||ar]

22:13 <weilewei> The documentation link in this page is 404 not found: https://github.com/STEllAR-GROUP/hpx/tree/master/libs/mpi

22:15 <bita> hkaiser, do you have a minute?

23:06 <hkaiser> bita: here

23:06 <bita> In kmeans, to initialize the centroids, shuffle for matrices is used. Then the first k rows are selected (https://github.com/STEllAR-GROUP/phylanx/blob/master/examples/algorithms/kmeans/kmeans.cpp#L29). Although it helps with having unique rows, why can't we use std::uniform_int_distribution on indices and generate only k index? Isn't it less expensive?

23:06 <hkaiser> bita: idk, ask Parsa ;-)

23:07 <parsa> what's up

23:07 <bita> I would say Parsa did it to comply with numpy version

23:07 <bita> but I can change that too

23:07 <hkaiser> sure

23:08 <bita> parsa, is it right? do you remember it?

23:08 <weilewei> trying to build hpx with mpi future, found this error: https://gist.github.com/weilewei/b1b3073f37dff715db11465c0d57130b

23:08 <weilewei> hkaiser ^^ any idea about it?

23:09 <hkaiser> weilewei: things are now in namespace hpx::mpi::experimental

23:09 <hkaiser> uhh no

23:10 <hkaiser> hold on

23:10 <weilewei> that's hpx itself build

23:10 <hkaiser> this looks like an outdated installation of hpx or somesuch

23:10 <hkaiser> ohh no, wait

23:10 <parsa> bita: you could just shuffle indices but these codes were example codes and we were to compare their performance against their python version. i'm sure you could do a lot more optimization that that

23:11 <hkaiser> weilewei: we fixed that, but it's not in master yet: https://github.com/STEllAR-GROUP/hpx/pull/4540

23:11 <weilewei> hkaiser nice, I will try this branch then

23:11 <hkaiser> weilewei: the ticket is here: https://github.com/STEllAR-GROUP/hpx/issues/4539

23:11 <hkaiser> describes exactly your problem and why it happens

23:12 <bita> thanks parsa. As I needed swap for shuffle, I would go with this one and I know the numpy version can change using np.random.randint

23:12 <weilewei> hkaiser that's great!

23:12 <parsa> bita: that surely is fine. just remember none of these are intended to be performant implementations anyways

23:13 <bita> got it :)

23:14 <parsa> their main purpose is to drive phylanx development and we added a lot of things to get them to work. do you anything you want with it

23:14 <bita> nod

23:17 <weilewei> hkaiser adding barrier does not help the multi-threaded ring G algorithm... also moved MPI_barrier at the end of gaint for-loop

23:18 <weilewei> hkaiser maybe some logic are wrong in my algorithm, or some hidden race condition... going to try hpx mpi future version to see if it resolve problems or not