 Usage:  {ll1,ll2} = glmlld(code,eta,y{,opt})  
 
 Input:

  code                   text string, the short code for the model (e.g. 
                         "bipro" for probit or "noid" for linear model). 
                         Append "fs" to the code for the expected ll2 
                         instead of ll2 (-> Fisher scoring algorithm). 
                         
  eta                    n x d matrix, the index values. 
                         
  y                      n x d matrix, the response values. 
                         
  opt                    optional, a list with optional input. The macro 
                         "glmopt" can be used to set up this parameter. 
                         
  opt.pow                scalar, power for power link. If not given, set 
                         to 0 (logarithm). 
                         
  opt.nbk                scalar, extra parameter k for negative binomial 
                         distribution. If not given, set to 1 (geometric 
                         distribution). 
                         
 Output:

  ll1                    n x d matrix, 1st derivative of log-likelihood. 
                         
  ll2                    n x d matrix, 2nd derivative of log-likelihood. 
                         
--------------------------------------------------------------
(C) MD*TECH Method and Data Technologies, 17.8.2000
