4 ms·

I love ruby, but its falling almost irretrievably behind python in data analysis/visualization/machine learning libraries.

dnc 9y ago
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.
- xapata 9y ago
  Would you like some code review to improve your Python?
- jonnytran 9y ago
  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 https://github.com/SciRuby 2: https://github.com/mrkn/pycall.rb https://github.com/mrkn/pycall.rb
ngsayjoe 9y ago
I'm doing machine learning in Python ... this is because there is no choice. Given a choice, i'd use Ruby rather than Python.
- gaius 9y ago
  because there is no choice You could be using R or MATLAB certainly, there are still choices in this space
- sidlls 9y ago
  C++, R, Matlab, etc. There is plenty of choice.
  - ngsayjoe 9y ago
    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.
pkd 9y ago
Actually I'd say it's catching up now because of http://sciruby.com/ http://sciruby.com/
- pugio 9y ago
  SciRuby has been around for quite a while, and hasn't made significant inroads. Their development also seems to have stalled.

dnc 9y ago

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.

xapata 9y ago
Would you like some code review to improve your Python?
jonnytran 9y ago
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 https://github.com/SciRuby 2: https://github.com/mrkn/pycall.rb https://github.com/mrkn/pycall.rb

ngsayjoe 9y ago

I'm doing machine learning in Python ... this is because there is no choice. Given a choice, i'd use Ruby rather than Python.

gaius 9y ago
because there is no choice You could be using R or MATLAB certainly, there are still choices in this space
sidlls 9y ago
C++, R, Matlab, etc. There is plenty of choice.
- ngsayjoe 9y ago
  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.

pkd 9y ago

Actually I'd say it's catching up now because of http://sciruby.com/ http://sciruby.com/

pugio 9y ago
SciRuby has been around for quite a while, and hasn't made significant inroads. Their development also seems to have stalled.