3. Adding new rule classes 3. Adding new rule classes However, before we add these conditional rules to the model we need to be wary. We will want to compare the model before and after changes. Even though Graphical-Belief has some very nice tools for comparing changes in a model after it is compiled, those tools don't work if we make structural changes to the model (adding variables or rules) and hence need to recompile it. To get around this difficulty, we can create a conditional distribution for the MOV which looks a lot like the old MOV-rule and does not depend on "Live-Steam" even though it is an input. After we compile the model, we can change the valuation and assess the effect with the standard history tools.
As there are two MOV's we can once again take advantage of Graphical-Belief's object-oriented model construction to organize the information. To do this, we will want to define a new class of MOV rule and make our changes there.
Figure 6. Model after creating "Steam-Mov-Rule".
We start by creating a new "Cond-Pmf" rule (previously we added "Cond-Pmf" to the template with the "Finding tool", Figure 1) and naming it "Steam-MOV-Rule". As this is a conditional rule, it has two input handles, one for input variables and one for output variables. We connect "Live-Steam" as an input and "MOV-25-B" as an output. The result looks like Figure 6.
Continue with this
example and edit the properties of the new rule.
Model Construction. Go back to the
begining of this example.
Return to
the main example page.
Back to overview of Graphical-Belief.
View a list
of Graphical-Belief in publications and downloadable technical
reports.
The Graphical-Belief user
interface is implemented in Garnet.
Get more
information about obtaining Graphical-Belief (and why
it is not generally available).
get
the home page for Russell Almond , author
of Graphical-Belief.
Click
here to get to the home page for Insightful (the company that StatSci
has eventually evolved into).