 Usage:  {xs, mh} = looreg (x, y, h)  
 
 Input:

  x                      n x p matrix, observations of explanatory variables 
                         
  y                      n x 1 matrix, observations of the dependent variable 
                         
  h                      n x p matrix. or 1 x p vector, 
                         or n x 1 vector, or scalar depending on whether 
                         one wants to use global or local bandwidth and if 
                         one uses the same bandwidth in all directions or 
                         one for each direction. 
                         
 Output:

  xy                     n x (p+1) matrix, the sorted data, the first p columns 
                         contain the sorted observations of the explanatory 
                         while the last column contains the sorted observations 
                         of the dependent variable 
                         
  mh                     n x 1 matrix, the sorted leave-one-out regression estimates 
                         
--------------------------------------------------------------
(C) MD*TECH Method and Data Technologies, 17.8.2000
