Not clear to understand :)

[keith.briggs@bt.com]

These are *not* equivalent in floating point, of course.  The first is better.   
Even better:

  for (i = 1; i <= n; i++)
    {
      m_x += (x[i * xstride] - m_x) / i;
      m_y += (y[i * ystride] - m_y) / i;
    }

Keith

-----Original Message-----
From: gsl-discuss-owner@sourceware.org
[mailto:gsl-discuss-owner@sourceware.org]On Behalf Of Serhiy Lisovenko
Sent: 22 November 2007 11:34
To: gsl-discuss@sourceware.org
Subject: Not clear to understand :)


in file fit/linear.c

  for (i = 0; i < n; i++)
    {
      m_x += (x[i * xstride] - m_x) / (i + 1.0);
      m_y += (y[i * ystride] - m_y) / (i + 1.0);
    }

and other same...

code equivalent to

  for (i = 0; i < n; i++)
    {
      m_x +=x[i * xstride];
      m_y +=y[i * ystride];
    }
m_x/=(double)n;
m_y/=(double)n;

But the second code is more clear and some faster (no extra divisions).