Compute initial values for the mean and covariance for the functions
emGauss and
daGauss.
This function is generic (see Methods); method functions can be
written to handle specific classes of data. Classes which
already have methods for this function include:
preGauss. A
default method operates on data frames and matrices.
USAGE:
Gauss.start(object, tol = sqrt(.Machine$double.eps))
REQUIRED ARGUMENTS:
object
either a matrix that may contain missing values or a
"preGauss"
object.
tol
lower limit on the diagonal entries of the covariance matrix.
VALUE:
a list containing the following components:
mu
column means for the data ignoring missing values.
sigma
diagonal matrix of column variances for the data ignoring missing values.
Variances less than or equal to
tol are replaced by 1.
NOTE:
The order of the variables will be different for
Gauss.start(x) and
Gauss.start(preGauss(x)) if the variables have been reordered by
preGauss.
REFERENCES:
Schafer, J. L. (1997), Analysis of Incomplete Multivariate Data,
Chapman & Hall, London.