next up previous contents index
Next: Constructing variance matrices from Up: Declaring variances and covariances Previous: Overview

Declaring single covariances


tex2html_wrap_inline33790 tex2html_wrap_inline33790 Syntax

  1. BD>var : v(S,N tex2html_wrap_inline33794 ,N tex2html_wrap_inline33798 )= (E) tex2html_wrap_inline33712

  2. BD>var : v(S,N tex2html_wrap_inline33794 )=(E) tex2html_wrap_inline33712

where tex2html_wrap_inline34842 are the names of elements which must already exist, tex2html_wrap_inline34518 is a valid equation, and S is a belief store number.

tex2html_wrap_inline33806 tex2html_wrap_inline33806

The syntax for the VAR:  command given here is used for the declaration of single variances and covariances. The first form of the syntax is used to specify the covariance between the elements tex2html_wrap_inline34584 and tex2html_wrap_inline34590 in belief store number S to be the value of the equation E. The round brackets surrounding the equation are essential unless E is a real number. The two names tex2html_wrap_inline34584 and tex2html_wrap_inline34590 are allowed to be the same, implying of course the declaration of a variance. The second fom of the syntax is a special short form of the first syntax when tex2html_wrap_inline35046 and we wish to specify a variance. Examples are:

BD>var : v(2,temperature,bloodpressure)=2.5 tex2html_wrap_inline33712

BD>var : v(1,X)=(ln (%c)+ln (%d)) tex2html_wrap_inline33712



David Wooff
Wed Oct 21 15:14:31 BST 1998