From: William James <w_a_x_man@...>
Date: 2006-08-22T00:50:20+09:00
Subject: Re: You can write Fortran in any language

Bil Kleb wrote:
> Here is an example used for computing correlations.
> It's almost a direct translation of several hundred
> lines of Fortran 77.
>
> If you can bare to look, please help, even it is simply
> to say "yo bonehead, why didn't you just require 'some_library'?"
>
> Thanks,
> --
> Bil
> http://fun3d.larc.nasa.gov
>

>
> cat << EOF > correlation.rb
> # Returns an array with correlation and y-normalized correlation^2
> # Assumes two nested arrays of variables and samples.  For example,
> # here's a case with 3 samples of 3 input variables and 2 output variables:
> #
> #   x = [ [a1,a2,a3], [b1,b2,b3], [c1,c2,c3] ]
> #   y = [ [r1,r2,r3], [s1,s2,s3] ]
>
> def correlation(x,y)
>
>    fail 'size arrays unequal' if x.first.size != y.first.size # lame
>
>    n = x.first.size
>
>    sum_x = Array.new(x.size){0.0}
>    sum_x_sq = Array.new(x.size){0.0}
>    for i in 0..x.size-1
>      sum_x[i]    = x[i].inject(0) { |sum, e| sum + e }
>      sum_x_sq[i] = x[i].inject(0) { |sum, e| sum + e**2 }
>    end

sum_x = x.map{|a| a.inject(0){|sum,e| sum + e} }
sum_x_sq = x.map{|a| a.inject(0){|sum,e| sum + e**2 } }


>
>    sum_y = Array.new(y.size){0.0}
>    sum_y_sq = Array.new(y.size){0.0}
>    for i in 0..y.size-1
>      sum_y[i]    = y[i].inject(0) { |sum, e| sum + e }
>      sum_y_sq[i] = y[i].inject(0) { |sum, e| sum + e**2 }
>    end
>
>    sum_xy = Array.new(x.size){ Array.new(y.size){0.0} }
>    for k in 0..n-1
>      for i in 0..x.size-1
>        for j in 0..y.size-1
>          sum_xy[i][j] += x[i][k]*y[j][k]
>        end
>      end
>    end

sum_xy = x.map{|xa|   y.map{|ya|
    xa.zip(ya).inject(0){|sum,na| sum + na.first*na.last}}}