News for hackers. Stuff that matters.
Re:ML Basics: K Nearest Neighbors in Ruby by aswanson on Thursday September 14, 2017 @01:10PM (#15250082)
I love ruby, but its falling almost irretrievably behind python in data analysis/visualization/machine learning libraries.
Re:ML Basics: K Nearest Neighbors in Ruby by dnc on Thursday September 14, 2017 @01:23PM (#15250217)
Same here, I use python at work for ML related stuff on daily basis for the reason pretty much everyone else chooses python over ruby ('that's what everyone else in my company uses and it just has more mature ML libraries and greater support'), but countless times have I caught myself thinking while writing python code: 'this could have been much cleaner and shorter code had I used ruby and its blocks'. It's a shame.
Re:ML Basics: K Nearest Neighbors in Ruby by jonnytran on Friday September 15, 2017 @12:51AM (#15254730)
I agree. I love Ruby. Not just the language, but the values of the community. Don't give up. Try SciRuby. I also recently discovered pycall, for those times when something really doesn't exist yet in Ruby. Of course, if it doesn't exist, that's an opportunity to make it yourself!

1: https://github.com/SciRuby

2: https://github.com/mrkn/pycall.rb

Re:ML Basics: K Nearest Neighbors in Ruby by xapata on Thursday September 14, 2017 @09:22PM (#15253854)
Would you like some code review to improve your Python?
Re:ML Basics: K Nearest Neighbors in Ruby by ngsayjoe on Thursday September 14, 2017 @01:37PM (#15250385)
I'm doing machine learning in Python ... this is because there is no choice. Given a choice, i'd use Ruby rather than Python.
Re:ML Basics: K Nearest Neighbors in Ruby by gaius on Thursday September 14, 2017 @02:02PM (#15250712)
because there is no choice

You could be using R or MATLAB certainly, there are still choices in this space

Re:ML Basics: K Nearest Neighbors in Ruby by sidlls on Thursday September 14, 2017 @03:20PM (#15251599)
C++, R, Matlab, etc. There is plenty of choice.
Re:ML Basics: K Nearest Neighbors in Ruby by ngsayjoe on Monday September 18, 2017 @11:42PM (#15281628)
All you mentioned is not better choice, C++ is not suited for ML / startups (I was a hardcore C++ programmer). As for R & Mathlab, i think they are not suited for ppl with CS background, they are more suited for people with maths background.
Re:ML Basics: K Nearest Neighbors in Ruby by pkd on Thursday September 14, 2017 @06:02PM (#15252872)
Actually I'd say it's catching up now because of http://sciruby.com/
Re:ML Basics: K Nearest Neighbors in Ruby by pugio on Thursday September 14, 2017 @06:42PM (#15253090)
SciRuby has been around for quite a while, and hasn't made significant inroads. Their development also seems to have stalled.