mlpack IRC logs, 2017-10-04

Logs for the day 2017-10-04 (starts at 0:00 UTC) are shown below.

October 2017
--- Log opened Wed Oct 04 00:00:52 2017
12:29 -!- kaushik_ [uid193796@gateway/web/] has joined #mlpack
15:32 -!- govg [~govg@unaffiliated/govg] has quit [Ping timeout: 246 seconds]
17:07 < rcurtin> zoq: I refactored the SGD-like optimizers, maybe the patch would be useful for you too---
17:07 < rcurtin> I put it in on the 'batch' branch, commits f29a9ddb fddae469 b3f9a456
17:07 < rcurtin> the logistic regression function changes need a lot of help, I am now debugging those; but at the very least the optimizer infrastructure seems to be ok in those patches I pushed
18:37 < zoq> rcurtin: Where does the initial batchSize (32) come from?
19:27 < rcurtin> I haven't tuned it, just a guess
19:29 < rcurtin> there are some bugs I am fixing now, I will push a fix shortly
19:40 < kaushik_>
19:41 < kaushik_> rcurtin: cli_test doesn't compile now and i get errors as above ^
19:41 < rcurtin> kaushik_: comment out cli_test.cpp, cli_bindings_test.cpp, and python_bindings_test.cpp
19:41 < kaushik_> ok
19:41 < rcurtin> these are the errors that I will fix once you finish the rest of the conversion; I can't fix them until everything uses serialize() instead of Serialize() though
19:42 < rcurtin> or, I could fix them now, but it would be a huge amount of pointless work :) much easier to do after everything is changed
20:53 < kaushik_> rcurtin: can you see the latest changes in methods/hoeffding_trees from PR. I am probably doing something wrong there.
20:55 < kaushik_>
21:11 < rcurtin> kaushik_: I can't see the code so I can't help debug
21:12 < rcurtin> ah nevermind I see that the changes were already pushed
21:22 < kaushik_> rcurtin: maybe you should first see hoeffding_tree_impl.hpp
21:22 < kaushik_>
21:24 < kaushik_> rcurtin: here,
21:28 < rcurtin> why does irccloud's pastebin hijack the pgup and pgdown keys on my browser
21:28 < rcurtin> ugh
21:31 < rcurtin> kaushik_: typically this happens when there is no default constructor and boost::serialization needs to default-construct an object
21:31 < rcurtin> but I don't see what you have changed in the code that should cause the default constructor to be needed where it wasn't before
21:32 < rcurtin> can you rerun with the -ftemplate-backtrace-limit=0 option to gcc?
21:32 < rcurtin> you can set that by either setting CMAKE_CXX_FLAGS when you configure CMake or modifying CMAKE_CXX_FLAGS in CMakeLists.txt
21:32 < rcurtin> basically the backtrace is not long enough, some of it is omitted:
21:33 < rcurtin> /usr/include/boost/serialization/vector.hpp:159:9: [ skipping 39 instantiation contexts, use -ftemplate-backtrace-limit=0 to disable ]
22:03 < kaushik_> rcurtin:
22:15 < rcurtin> hang on, let me take a look...
22:15 < rcurtin> I have to feed my cat first, she is not going to stop making noise until I do :)
22:18 < zoq> Fritz? Can't remember the name of the other cat.
22:18 < rcurtin> ha, yeah, the other one is named Drusilla
22:18 < zoq> ah, right
22:19 < rcurtin> kaushik_: ok, I see, the problem comes from line 767 in hoeffding_tree_impl.hpp:
22:19 < rcurtin> ar & BOOST_SERIALIZATION_NVP(numericSplits);
22:19 < rcurtin> I think that because we are now serializing a vector, boost::serialization will call vector.resize(), which needs access to a default constructor
22:19 < rcurtin> I guess that in this case, we can modify the numeric and categorical split types to have a default value for the number of classes
22:26 < rcurtin> zoq: I pushed some more fixes to the 'rcurtin/batch' branch, now logistic regression and NCA work correctly
22:27 < kaushik_> rcurtin: ah ok I see.
22:27 < kaushik_> will do that by tomorrow. Its 4 AM and I need to sleep :D
22:30 < zoq> oh, get some sleep
22:36 < zoq> rcurtin: Looks like you are almost done.
23:11 < rcurtin> zoq: yeah, just a few more to do
--- Log closed Thu Oct 05 00:00:53 2017