The R Project SVN R

Rev

Rev 88690 | Show entire file | Ignore whitespace | Details | Blame | Last modification | View Log | RSS feed

Rev 88690 Rev 88694
Line 140... Line 140...
140 
 
 140 
 
141 
  Missing values are allowed, and are handled exactly in method \code{"ML"}.
 141 
  Missing values are allowed, and are handled exactly in method \code{"ML"}.
142 
 
 142 
 
143 
  If \code{transform.pars} is true, the optimization is done using an
 143 
  If \code{transform.pars} is true, the optimization is done using an
144 
  alternative parametrization which is a variation on that suggested by
 144 
  alternative parametrization which is a variation on that suggested by
145 
  Jones (1980) and ensures that the model is stationary.  For an AR(p)
 145 
  \bibcitet{R:Jones:1980} and ensures that the model is stationary.
- 
 
 146 
  For an AR(p)
146 
  model the parametrization is via the inverse tanh of the partial
 147 
  model the parametrization is via the inverse tanh of the partial
147 
  autocorrelations: the same procedure is applied (separately) to the
 148 
  autocorrelations: the same procedure is applied (separately) to the
148 
  AR and seasonal AR terms.  The MA terms are not constrained to be
 149 
  AR and seasonal AR terms.  The MA terms are not constrained to be
149 
  invertible during optimization, but they will be converted to
 150 
  invertible during optimization, but they will be converted to
150 
  invertible form after optimization if \code{transform.pars} is true.
 151 
  invertible form after optimization if \code{transform.pars} is true.