The R Project SVN R

Rev

Rev 89613 | Show entire file | Ignore whitespace | Details | Blame | Last modification | View Log | RSS feed

Rev 89613 Rev 90253
Line 52... Line 52...
52 
  with state vector \samp{a}, transitions \samp{a <- T a + R e},
 52 
  with state vector \samp{a}, transitions \samp{a <- T a + R e},
53 
  \eqn{e \sim {\cal N}(0, \kappa Q)}{e ~ N(0, kappa Q)} and observation
 53 
  \eqn{e \sim {\cal N}(0, \kappa Q)}{e ~ N(0, kappa Q)} and observation
54 
  equation \samp{y = Z'a + eta},
 54 
  equation \samp{y = Z'a + eta},
55 
  \eqn{(eta\equiv\eta), \eta \sim {\cal N}(0, \kappa h)}{eta ~ N(0, kappa h)}.
 55 
  \eqn{(eta\equiv\eta), \eta \sim {\cal N}(0, \kappa h)}{eta ~ N(0, kappa h)}.
56 
  The likelihood is a profile likelihood after estimation of
 56 
  The likelihood is a profile likelihood after estimation of
57 
  \eqn{\kappa}{kappa}.
 57 
  \eqn{\kappa}.
58 
 
 58 
 
59 
  The model is specified as a list with at least components
 59 
  The model is specified as a list with at least components
60 
  \describe{
 60 
  \describe{
61 
    \item{\code{T}}{the transition matrix}
 61 
    \item{\code{T}}{the transition matrix}
62 
    \item{\code{Z}}{the observation coefficients}
 62 
    \item{\code{Z}}{the observation coefficients}
Line 92... Line 92...
92 
}
 92 
}
93 
 
 93 
 
94 
\value{
 94 
\value{
95 
  For \code{KalmanLike}, a list with components \code{Lik} (the
 95 
  For \code{KalmanLike}, a list with components \code{Lik} (the
96 
  log-likelihood less some constants) and \code{s2}, the estimate of
 96 
  log-likelihood less some constants) and \code{s2}, the estimate of
97 
  \eqn{\kappa}{kappa}.
 97 
  \eqn{\kappa}.
98 
 
 98 
 
99 
  For \code{KalmanRun}, a list with components \code{values}, a vector
 99 
  For \code{KalmanRun}, a list with components \code{values}, a vector
100 
  of length 2 giving the output of \code{KalmanLike}, \code{resid} (the
 100 
  of length 2 giving the output of \code{KalmanLike}, \code{resid} (the
101 
  residuals) and \code{states}, the contemporaneous state estimates,
 101 
  residuals) and \code{states}, the contemporaneous state estimates,
102 
  a matrix with one row for each observation time.
 102 
  a matrix with one row for each observation time.