8. Numerical Technique: Linear Modeling

Given input data points (x₁,y₁), (x₂,y₂), ... (x_n,y_n), the linear approximation y = m*x + b with minimum error is given by equations 3.1 and 3.2, page 130:

Excerpts from chapter3_10.c:

   int count=0;
   double x, y, sumx=0, sumy=0, sumx2=0, sumxy=0, denominator, m, b;

      while ((fscanf(zone,"%lf %lf",&x,&y)) == 2)
      {
         ++count;
         sumx += x;
         sumy += y;
         sumx2 += x*x;
         sumxy += x*y;
       }

      denominator = sumx*sumx - count*sumx2;
      m = (sumx*sumy - count*sumxy)/denominator;
      b = (sumx*sumxy - sumx2*sumy)/denominator;