cameron.freenode.net changed the topic of #mlpack to: http://www.mlpack.org/ -- We don't respond instantly... but we will respond. Give it a few minutes. Or hours. -- Channel logs: http://www.mlpack.org/irc/
govg has quit [Ping timeout: 255 seconds]
govg has joined #mlpack
stephentu has quit [Ping timeout: 245 seconds]
govg has quit [Ping timeout: 245 seconds]
govg has joined #mlpack
govg has quit [Quit: leaving]
govg has joined #mlpack
stephentu has joined #mlpack
< naywhayare>
stephentu: I agree with your assessment on #380... I want to leave it for a prospective GSoC applicant though
< naywhayare>
last year there were a lot of people who pointed out that there just aren't any easy bugs for a beginner to solve :)
< stephentu>
haha ok
< stephentu>
the CF stuff is messier than I thought
< naywhayare>
it is... this is one of the problems with GSoC
< naywhayare>
we get great students and they do great work, but sometimes as soon as the summer is over, they disappear without completely wrapping things up and get busy with other stuff
< naywhayare>
I put it on my todo list, but it always has to be below "doing research and graduating"
< stephentu>
ah i didnt realize parts of mlpack was written by gsoc people
< stephentu>
i just thought you were a machine
< naywhayare>
it says I've added and removed like 14 million lines of code, but most of that is just failed svn merges and copies (git is much better!)
< naywhayare>
most to all of the code in the library I've at least looked over, but that really doesn't scale as the library grows
< naywhayare>
I'm hoping that when I graduate at the end of the summer I can find a position that will let me continue working on mlpack (plus maybe some research)
< naywhayare>
although writing papers and devising new algorithms is fun and all, providing real-world high-quality implementations that work out of the box is really time-consuming and difficult, and it seems to me that most researchers often don't focus on that
< stephentu>
ya i hear you, writing solid implementations doesnt advance your academic career
< stephentu>
although honestly i think that attitude is really hurting research quality
< stephentu>
when all the prototypes are just so bad
< stephentu>
nobody bothers to compare to each other
< stephentu>
b/c nobody wants to touch that horrible code
< naywhayare>
the mlpack code was really, really horrifying when I "inherited" it
< naywhayare>
part of it was that the people writing it weren't thinking about how people might use their code
< naywhayare>
for instance, all-k-nearest-neighbors would... load the dataset, calculate nearest neighbors, print how long it took, and exit (note: no saving of results)