mlpack IRC logs, 2017-05-05

Logs for the day 2017-05-05 (starts at 0:00 UTC) are shown below.

>
May 2017
Sun
Mon
Tue
Wed
Thu
Fri
Sat
 
1
2
3
4
5
6
7
8
9
10
11
12
20
21
22
23
24
25
26
27
28
29
30
31
--- Log opened Fri May 05 00:00:53 2017
00:17 -!- benchmark [~benchmark@198-0-35-42-static.hfc.comcastbusiness.net] has joined #mlpack
00:17 -benchmark:#mlpack- ALLKRANN (-k 3 --first_leaf_exact) | covtype 458.73 (old) => 458.79 (new) => 0.06 (diff) | corel-histogram 7.81 (old) => 7.86 (new) => 0.05 (diff) |
00:17 -benchmark:#mlpack- ALLKRANN (-k 3 --sample_at_leaves --single_mode) | covtype 22.73 (old) => 22.89 (new) => 0.16 (diff) | corel-histogram 4.26 (old) => 4.26 (new) => 0.00 (diff) |
00:17 -benchmark:#mlpack- ALLKRANN (-k 3 --tree_type=cover) | covtype 563.86 (old) => 560.73 (new) => -3.13 (diff) | corel-histogram 31.40 (old) => 31.32 (new) => -0.08 (diff) |
00:17 -benchmark:#mlpack- ALLKRANN (-k 3 --alpha 0.9) | covtype 462.44 (old) => 461.59 (new) => -0.85 (diff) | corel-histogram 7.70 (old) => 7.72 (new) => 0.02 (diff) |
00:17 -benchmark:#mlpack- ALLKRANN (-k 3 --alpha 0.9 --tau 10) | covtype 459.31 (old) => 459.95 (new) => 0.64 (diff) | corel-histogram 7.32 (old) => 7.33 (new) => 0.01 (diff) |
00:17 -benchmark:#mlpack- ALLKRANN (-k 3 --first_leaf_exact --single_mode) | covtype 206.16 (old) => 205.91 (new) => -0.25 (diff) | corel-histogram 8.42 (old) => 8.39 (new) => -0.03 (diff) |
00:17 -benchmark:#mlpack- ALLKRANN (-k 3 --sample_at_leaves) | covtype 316.72 (old) => 316.26 (new) => -0.46 (diff) | corel-histogram 7.26 (old) => 7.27 (new) => 0.01 (diff) |
00:17 -benchmark:#mlpack- ALLKRANN (-k 3) | covtype 458.74 (old) => 458.69 (new) => -0.05 (diff) | corel-histogram 7.81 (old) => 7.85 (new) => 0.04 (diff) |
00:17 -benchmark:#mlpack- ALLKRANN (-k 3 --tree_type=cover --single_mode) | covtype 447.08 (old) => 443.61 (new) => -3.47 (diff) | corel-histogram 28.00 (old) => 28.45 (new) => 0.44 (diff) |
00:17 -benchmark:#mlpack- ALLKNN (-k 3 --single_mode --tree_type=cover) | covtype 325.21 (old) => 326.79 (new) => 1.57 (diff) | corel-histogram 188.66 (old) => 190.03 (new) => 1.37 (diff) |
00:17 -benchmark:#mlpack- ALLKNN (-k 3 --tree_type=cover) | covtype 180.24 (old) => 181.07 (new) => 0.82 (diff) | corel-histogram 113.02 (old) => 113.03 (new) => 0.01 (diff) |
00:17 -benchmark:#mlpack- ALLKNN (-k 3) | covtype 14.84 (old) => 14.81 (new) => -0.03 (diff) | corel-histogram 13.09 (old) => 13.04 (new) => -0.04 (diff) |
00:17 -benchmark:#mlpack- ALLKFN (-k 3 --tree_type=r) | covtype -1.00 (old) => -1.00 (new) => 0.00 (diff) | corel-histogram 148.85 (old) => 148.80 (new) => -0.05 (diff) |
00:17 -benchmark:#mlpack- ALLKFN (-k 3 --single_mode --tree_type=cover) | covtype 173.07 (old) => 173.33 (new) => 0.26 (diff) | corel-histogram 51.70 (old) => 51.93 (new) => 0.23 (diff) |
00:17 -benchmark:#mlpack- ALLKFN (-k 3 --single_mode --tree_type=r) | covtype 908.08 (old) => 904.59 (new) => -3.49 (diff) | corel-histogram 389.13 (old) => 392.82 (new) => 3.69 (diff) |
00:17 -benchmark:#mlpack- ALLKFN (-k 3 --tree_type=cover) | covtype 140.07 (old) => 139.70 (new) => -0.37 (diff) | corel-histogram 46.56 (old) => 46.67 (new) => 0.11 (diff) |
00:17 -benchmark:#mlpack- ALLKFN (-k 3) | covtype 12.22 (old) => 11.94 (new) => -0.28 (diff) | corel-histogram 9.59 (old) => 9.52 (new) => -0.07 (diff) |
00:17 -benchmark:#mlpack- Benchmarks 34 of 34 passed.
00:17 -!- benchmark [~benchmark@198-0-35-42-static.hfc.comcastbusiness.net] has quit [Client Quit]
00:24 < rcurtin> zoq: yeah, I try to avoid the unnecessary merges when I can
00:26 -!- aashay [uid212604@gateway/web/irccloud.com/x-ixickrwtnjmzuxnf] has quit [Quit: Connection closed for inactivity]
05:23 -!- mikeling [uid89706@gateway/web/irccloud.com/x-coidfgkjozfefydk] has joined #mlpack
06:50 -!- aashay [uid212604@gateway/web/irccloud.com/x-rdydmryvozbferys] has joined #mlpack
07:14 -!- mentekid [~yannis@cpc92878-cmbg18-2-0-cust813.5-4.cable.virginm.net] has joined #mlpack
07:19 -!- mentekid [~yannis@cpc92878-cmbg18-2-0-cust813.5-4.cable.virginm.net] has quit [Client Quit]
07:21 -!- mentekid [~yannis@cpc92878-cmbg18-2-0-cust813.5-4.cable.virginm.net] has joined #mlpack
07:26 -!- mentekid [~yannis@cpc92878-cmbg18-2-0-cust813.5-4.cable.virginm.net] has quit [Client Quit]
08:52 -!- witness_ [uid10044@gateway/web/irccloud.com/x-ixljjrzjmfqdeuux] has joined #mlpack
09:19 -!- aashay [uid212604@gateway/web/irccloud.com/x-rdydmryvozbferys] has quit [Quit: Connection closed for inactivity]
10:28 -!- govg [~govg@unaffiliated/govg] has joined #mlpack
14:48 -!- Trion [~Thunderbi@117.215.245.182] has joined #mlpack
16:54 -!- Trion [~Thunderbi@117.215.245.182] has quit [Quit: Have to go, see ya!]
17:25 -!- vinayakvivek [uid121616@gateway/web/irccloud.com/x-mtrutypmqimplgph] has joined #mlpack
17:39 < rcurtin> zoq: I am having a lot of deja vu about #993; wasn't there some other change that was made recently to print all the tests?
17:39 < rcurtin> I seem to remember finding the --list_content option (and then forgetting it until you pointed it out again), but I can't remember anything else about the discussion
17:46 < zoq> rcurtin: I think we thought they deprecated the option, but it looks like the option was introduced in 1.60.0.
17:48 -!- mikeling [uid89706@gateway/web/irccloud.com/x-coidfgkjozfefydk] has quit [Quit: Connection closed for inactivity]
17:48 < rcurtin> 1.60, not 1.39?
17:49 < zoq> I think 1.60 at least there is no list-content for 1.59: http://www.boost.org/doc/libs/1_60_0/libs/test/doc/html/boost_test/utf_reference/rt_param_reference/list_content.html
17:53 < zoq> I tested it on masterblaster and there is no list_content option
17:53 < rcurtin> yes, masterblaster is 1.58
17:54 < rcurtin> ok, so I guess the comment you made in #993 has a typo? "for newer boost versions (>= 103600) there is already an option"
17:54 < rcurtin> 103600 would be 1.36, which is much older
17:54 < rcurtin> it seems like in 1.60 perhaps there was a rewrite of the input option parsing
17:55 < rcurtin> in 1.58 (and probably older), changing "tests" to "--print_tests" works just fine, I guess because they are just ignoring options they don't know
17:55 < rcurtin> so I suspect it may be possible to just "#if BOOST_VERSION < 106000" and then if that's true, add the option --list_content, to backport it
17:56 < rcurtin> but I am not sure exactly where that parsing change is, if it's 1.60.0 or 1.59.0 or what
17:59 < zoq> you are right, my bad it's 106000
18:01 < zoq> I think at the end the version doesn't matter, since if list-content is available it's not overwritten by our version.
18:02 < rcurtin> ah, right, good point!
18:02 < rcurtin> but I think the format that --list_content will give is slightly different than the format your patch gives
18:04 < zoq> yeah, if we are going to go this way I'll change the output accordingly
18:04 < rcurtin> sure, as long as that output format works for whatever purpose you have for it
21:29 -!- vinayakvivek [uid121616@gateway/web/irccloud.com/x-mtrutypmqimplgph] has quit [Quit: Connection closed for inactivity]
--- Log closed Sat May 06 00:00:54 2017