 Usage:  h = glmdiagh(bv,x{,w})  
 
 Input:

  bv                     p x p matrix, the estimated covariance 
                         matrix for the coefficients. 
                         
  x                      n x p matrix, the matrix of (observed) 
                         regressor variables. 
                         
  w                      optional, n x 1 matrix, the weights 
                         used to estimate the coefficients. 
                         
 Output:

  h                      n x 1 matrix, the diagonal elements of 
                         the `hat' matrix: 
                         W'^(1/2)X(X'WX)^(-1) X'W^(1/2). 
                         
--------------------------------------------------------------
(C) MD*TECH Method and Data Technologies, 17.8.2000
