We assume throughout this section that you have already used the
GRID: command (or the GRID0: or NODE: and
ARC: commands) to identify the nodes and arcs of interest.
Suppose, for example, that you wish to adjust the collection 
by
the collections 
and 
, and so are about to issue the
command
If we assume that you wish all these collections to appear on the influence diagram, then beforehand you should associate nodes on the diagram with these collections, and draw arcs between them as appropriate. The TESTGRID: command can be used quickly to show all the nodes and arcs relevant to the adjustment, without showing any other information.
A single option, the influence option, is used to require the display of the interactive construction of an influence diagram for the current adjustment. If the influence option is switched on, no other output is available interactively. However, results from an adjustment remain available (until overwritten by another adjustment) via later use of the SHOW: command. Figure 9.2 charts the different options available when drawing influence diagrams. The kind of shading used to shade the nodes to show summary and diagnostic information from an adjustment can relate either to the original quantities or to the canonical quantities. By default, shadings relate to the original quantities; otherwise the wheel control can be set to produce the canonical wheel.
We consider two kinds of influence diagram constructions: those resulting from a single adjustment, and those constructed from a sequence of adjustments. We discuss each of these as in the following examples. Influence diagrams can only be constructed under the fitting of sources of information; not the withdrawing of them.
