verne.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/
< rcurtin> lozhnikov: responding to #664 is my first priority tomorrow, I should have that done by 1500 UTC
< lozhnikov> rcurtin: ok
Mathnerd314 has quit [Ping timeout: 246 seconds]
marcosirc has joined #mlpack
Mathnerd314 has joined #mlpack
nilay has joined #mlpack
sumedhghaisas has joined #mlpack
< sumedhghaisas> @marcosirc: hey marcos...
< sumedhghaisas> I am looking into your pull request for issue 674... I think boost::variant will be a simpler solution... about the right solution I still have doubts... :)
< marcosirc> Hi @sumedhghaisas
< marcosirc> So, are you thinking in another option?
< sumedhghaisas> noo... I am just considering the performance in using boost variant...
< sumedhghaisas> but on top of that is performance a big issue here??
< sumedhghaisas> I mean the code we are trying to refactor... will it be called too many number of times?? I don't think so...
< sumedhghaisas> In that scenario we should go ahead with boost::variant... thats most elegant...
< sumedhghaisas> I even like your solution but in that we need to change the abstraction completely...
< marcosirc> Yes! I agree. It won't be called many times, so I think we shouldn't worry about performance here.
< marcosirc> Thanks for your feedback.
< marcosirc> I have to see why it failed in appveyor...
< sumedhghaisas> yeah in this case usability is more important than performance...
< sumedhghaisas> ohh yeah... I noticed that....
< sumedhghaisas> but I anyways hate windows environment...
< marcosirc> yeah me too! :)
< marcosirc> I will read the logs to find the reason.
< sumedhghaisas> unfortunately my personal computer is not working... so its in repair... so I have to do everything on my company laptop... which is windows only...
< sumedhghaisas> :(
< marcosirc> Ups! It must be annoying!
< marcosirc> have you installed cygwin?
< sumedhghaisas> I have installed MinGW... I am not sure what is the difference between MinGW and cygwin... which is better???
nilay has quit [Ping timeout: 250 seconds]
< marcosirc> I think there are far more tools for cygwin than for MinGW...
< marcosirc> I used cygwin for some months at the beginning of the year. I was stolen my personal laptop, so I borrowed a windows pc. I managed to do a lot of things with cygwin.
kwikadi has quit [Remote host closed the connection]
kwikadi has joined #mlpack
marcosirc has quit [Quit: WeeChat 1.4]
nilay has joined #mlpack
marcosirc has joined #mlpack
nilay has quit [Ping timeout: 250 seconds]
nilay has joined #mlpack
< nilay> zoq: Hi, In the descritize function to sample the locations j1, j2 do i have to sample them with or without replacement?
< nilay> and also to insure that in binary_vec1 and binary_vec2 all (binary_vec1[j1], binary_vec2[j2]) pairs are distinct, i can choose binary_vec1 from all odd locations and binary_vec2 from all even locations. Do you think this is a good idea
< nilay> ensure*
< nilay> or do you think ensuring such a thing(all pairs among the 2 vectors are unique) is not a big deal as probability of any pair being same location in both vectors is just (1/256)
< nilay> assuming our each binary vector is of size 256.
tsathoggua has joined #mlpack
tsathoggua has quit [Client Quit]
< zoq> nilay: I'm not sure, do you mean the permutation function? If that's the case you should sample without replacement.
< zoq> nilay: Choosing bin_vec from either all odd locations or even locations, sounds like a good idea.
< nilay> i think choosing from all locations also isn't bad, if we see the probability
< zoq> yeah, I think you are right, it's your call. I'm fine with both variants.
nilay_ has joined #mlpack
< nilay_> Ok great
nilay_ has quit [Ping timeout: 250 seconds]
marcosirc has quit [Quit: WeeChat 1.4]
sumedhghaisas has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client]