ChanServ changed the topic of #mlpack to: "mlpack: a fast, flexible machine learning library :: We don't always respond instantly, but we will respond; please be patient :: Logs at http://www.mlpack.org/irc/
< UmarJ>
hi
< UmarJ>
is there a flag to specify where the python bindings are installed?
< _slack_mlpack_U0>
You can check the CMakelist.txt
< UmarJ>
hmm, couldn't find anything there. guess ill just delete them from /usr/.../site-packages and copy them from the build directory to where i want
< AnushKiniGitter[>
> `zoq on Freenode` Anush Kini (Gitter): imshow does not work with the C++ kernel, as a workaround you have to save the plot and use xeus image_from_file.
< AnushKiniGitter[>
Thanks. Just tried this out. It works
< AnushKiniGitter[>
(edited) ... works => ... works great.
< AnushKiniGitter[>
(edited) ... It works great. => ... It works.
< UmarJ>
rcurtin[m]: Thank you! I'll test it out (once I get the bindings to build haha >_>)
< rcurtin[m]>
if you're just looking to use the bindings, depending on your setup you may be able to just install from pip or conda, which might be easier :)
< UmarJ>
Oh, that's true, I will do that :)
mega-05 has joined #mlpack
mega-05 has quit [Remote host closed the connection]
ib07 has joined #mlpack
< rcurtin[m]>
removing `boost::visitor` from `NSModel` appears to have decreased compilation time for `mlpack_knn` from 60s to 48s... that was just one trial, but it looks promising
< zoq>
Hopefully we can see similar results for the ann part.
< rcurtin[m]>
I am hoping we see more than 20% speedup :)
< rcurtin[m]>
I think the ann code uses boost::visitor much more heavily than the NSModel code does
< rcurtin[m]>
I also didn't check RAM usage in my simple experiment :)
< zoq>
Yeah, that is another story, but that should improve as well.
< rcurtin[m]>
agreed! we'll find out
< rcurtin[m]>
I only have my phone to work on today, and the coding is so slow on it that I won't be able to finish the other Model classes like RSModel, RAModel, etc., but I should be able to do those pretty quickly when I have a keyboard again tomorrow :)
< rcurtin[m]>
my new computer parts are waiting for me when I get home too, so hopefully that should help accelerate the build/test cycle too :)
< zoq>
hehe, on the ann front there are still some layers left, but we will get there.