Now consider the example code in Figure 11.2, where A,B,D,E,F,X,Y,Z are the names of elements or bases to take part in the adjustment. The influence option is used to switch on the influence diagram display. Next, the overwrite control is switched on (its default). When this command is switched on, the graphics screen is not cleared after any adjustment command, and therefore any graphics drawn are superimposed over previously drawn screens. The GCLEAR: command can be used to clear the graphics output. The sequence of commands shown in Figure 11.2 will achieve the following:
Figure 11.2: Drawing superimposed influence diagrams
It is also possible to issue an ADJUST: command without naming the base to be adjusted, as in
In such cases, the effect as far as the influence diagram is concerned is exactly the same as if you had issued the command
The overshade control is used to determine whether or not the
shadings for successive partial adjustments are redrawn or not, and only
applies when the overwrite control is switched off. The
ideas here are best illustrated by example. Suppose that you carry out
the adjustment and in doing so obtain an influence
diagram where the B node is shaded first for the adjustment by
and then by the partial adjustment given by fitting also on
. This gives a node with two contiguous shaded sectors, the
whole representing the adjustment
where
.
Suppose that you now adjust partially by
and that you have
switched off the overwrite control. This will result in another
sector of the node being shaded, to correspond to the partial
adjustment by E. However, the treatment of the shading of the former
shaded sectors is as follows. If the overshade control is
switched on, the sector corresponding to the adjustment by D is
reshaded for that overall adjustment, thus replacing the initial two
shaded sectors by one shaded sector. Otherwise the initial two shaded
sectors are left as they were.