 Usage:  {edr, eigen} = sir1 (matYX, nbslices)  
 
 Input:

  matYX                  n x (p+1) matrix containing the value of the dependent 
                         variable in its first column, the values of the p 
                         explanatory variables in the last p columns, 
                         for a sample of size n. 
                         
  nbslices               optional scalar parameter, 
                         number of slices for the non missing y's cases. 
                         (By default, it is equal to 5) 
                         
 Output:

  edr                    p x p matrix which contains the estimated EDR directions 
                         (in columns), sorted according to the increasing values 
                         of their corresponding eigenvalues. 
                         
                         
  eigen                  p x 1 matrix which contains the corresponding eigenvalues 
                         sorted according to the increasing values. 
                         
--------------------------------------------------------------
(C) MD*TECH Method and Data Technologies, 17.8.2000
