the maximum number of iterations in the EM algorithm. Alternatively,
maxit can be a list containing some or all parameters in the list
returned by
emGauss.control.
tolerance
convergence criterion for the EM algorithm. This may either be a
number or a vector of two numbers. By default, convergence is assumed
when the maximum relative change in the parameter estimates is
less than
tolerance[1]. You may also specify tolerance[2], in which
case convergence also occurs when the absolute change in the
log-likelihood from one iteration to the next is less than
tolerance[2]. By default, the log-likelihood is not used in
checking for convergence.
rcmin
parameter used to determine when the variance-covariance matrix is
singular. Typically,
sqrt(.Machine$double.eps) is acceptable.
last
the sequence of iteration numbers that are saved
is determined as the final
last iterations.
trace
if TRUE, additional information is printed during the EM algorithm.
VALUE:
a list containing the control values described above.