Syntax
where r and v are switches;
The first form of the command is used to export beliefs so that they follow a label. Typically the label
will be a string of alphabetic characters preceded by the symbol `@' (this symbol is used as
the prefix recognised by [B/D] as implying that the remainder of a string is a label name).
The label is output as one line of text. The second form of the command is identical to the
first except that no label is defined, and no label will be exported.
The list of element names and bases, N1, N2, ..., defines the collection of elements
for which a variance covariance matrix will
be exported. It is an error if the collection so defined is empty.
The elements in the
collection are output in an the order given by the ordering convention
(so that their ordering in the EXPORT: command line is
irrelevant). The variance-covariance matrix for the collection
corresponding to belief store number i is output, one number per line.
For example, suppose that the collection being output is
The latter two forms of the command, with r replacing v, duplicate the former two forms of
the command, except that the corresponding beliefs are output as correlations rather than
variances and covariances.
are the names of elements or bases; and L is the name of a
file label. Base notation (see §6.2.4) can be used with the names. i is
an integer representing a valid belief store number.
.
On successive lines the output consists of the following variances and
covariances: 
, 
, ...,
, 
, 
, ..., and so
forth.
