Computing the scatter of data around linear fits

The Python method below computes the scatter of data around a given linear model.

Let me know if you find any bugs.

   
 def scatterfit(x,y,a=None,b=None):  
  """  
 Compute the mean deviation of the data about the linear model given if A,B  
 (y=ax+b) provided as arguments. Otherwise, compute the mean deviation about   
 the best-fit line.  
   
 x,y assumed to be Numpy arrays. a,b scalars.  
 Returns the float sd with the mean deviation.  
   
 Author: Rodrigo Nemmen  
  """  
   
  if a==None:   
  # Performs linear regression  
  a, b, r, p, err = scipy.stats.linregress(x,y)  
    
  # Std. deviation of an individual measurement (Bevington, eq. 6.15)  
  N=numpy.size(x)  
  sd=1./(N-2.)* numpy.sum((y-a*x-b)**2); sd=numpy.sqrt(sd)  
    
  return sd