Performing Linear Regression Using Ruby

#1

Post by ucsd_surfNerd » Tue, Sep 11, 2012, 11:41 PM UTC

Performing Linear Regression Using Ruby
sharethrough.com

Re: Performing Linear Regression Using Ruby

#2

Post by sciboy » Wed, Sep 12, 2012, 1:44 AM UTC

B = (X^TX)^(-1)X^TY anyone?

Re: Performing Linear Regression Using Ruby

#3

Post by EzGraphs » Wed, Sep 12, 2012, 1:45 AM UTC

Ruby is great for data prep, basic calculations, web app development, and scraping aggregating data and R for visualization. I find them to be a joyful combination. Great for small data sets, quick estimations, and various small projects.

Larger data sets and performance intensive operations are better handled in Python (or Java, C++, etc). Lots of statistical analysis is way below the threshold of Hadoop and company.

Re: Performing Linear Regression Using Ruby

#4

Post by niggler » Wed, Sep 12, 2012, 1:49 AM UTC

What would be a cool followup post is an incremental solution that used ruby blocks.

Here: http://news.ycombinator.com/item?id=4508837

Re: Performing Linear Regression Using Ruby

#5

Post by carbocation » Wed, Sep 12, 2012, 1:52 AM UTC

sciboy wrote: post #2

B = (X^TX)^(-1)X^TY anyone?

Definitely. The jump from univariate regression to multivariate regression is a non-obvious step because the univariate approach does not generalize in a straightforward fashion.

Re: Performing Linear Regression Using Ruby

#6

Post by gergles » Wed, Sep 12, 2012, 1:56 AM UTC

I think it is especially important to note that linear regression assumes that the relationship between the variables is, well, linear, and that in the real world, it very rarely actually is.

At best, a big asterisk should come from any of these results if you didn't have someone with actual experience validate your design/proposed analyses first.

Re: Performing Linear Regression Using Ruby

#7

Post by zmjones » Wed, Sep 12, 2012, 2:02 AM UTC

And this is why we don't do statistical programming in Ruby.

Re: Performing Linear Regression Using Ruby

#8

Post by oscilloscope » Wed, Sep 12, 2012, 2:15 AM UTC

sciboy wrote: post #2

B = (X^TX)^(-1)X^TY anyone?

http://en.wikipedia.org/wiki/Regression_analysis#General_lin...

Re: Performing Linear Regression Using Ruby

#9

Post by thauck » Wed, Sep 12, 2012, 2:23 AM UTC

EzGraphs wrote: post #3

Ruby is great for data prep, basic calculations, web app development, and scraping aggregating data and R for visualization. I find them to be a joyful combination. Great for small data sets, quick estimations, and various small projects. Larger data sets and performance intensive operations are better handled in Python (or Java, C++, etc). Lots of statistical analysis is way below the threshold of Hadoop and company…

I mostly agree with this, because I do like Ruby even though I don't use it.

The missing link here, and the reason Python gets more love from the data community, is that Python scales down to the smaller data sets as well as it handles big ones. (Not sure if you ment it couldn't, but the distinction you make implies that.)

Re: Performing Linear Regression Using Ruby

#10

Post by cheald » Wed, Sep 12, 2012, 2:29 AM UTC

We do regressions with rb-gsl.

    require 'gsl'
    x = GSL::Vector.alloc(array_of_x_values)
    y = GSL::Vector.alloc(array_of_y_values)
    c0, c1, cov00, cov01, cov11, chisq, status = GSL::Fit::linear(x, y)

It's not nearly as much work, and it's much faster than doing it in pure Ruby. :)

(It also does weighted regressions and exponential fitting, among a host of other things. That wheel's gone done been invented already.)