Checking that Your Model is Complete
Before running an analysis, you may want to check that your model is complete. If it is not, DPL will warn you that it isn't and ask you to correct the problem.
- An influence diagram is incomplete if:
- Any node is missing a name;
- Any chance or decision event is missing state names; or
- Any chance event is missing probability data.
- A decision tree is incomplete if:
- There is an unattached subtree in the Decision Tree Pane that is not linked with the Perform Subtree command; or
- There are no get/pay expressions.
If you attempt to run an analysis from a model that is incomplete, DPL will display an error message. DPL will then highlight the node with missing information and open a dialog prompting you to provide the missing data.
If a value node is missing data, DPL will display a warning message and ask whether you would like to continue the run with zeros inserted for the missing data. If you deliberately added a value node without data, you should classify it as a dummy node. To do this edit the node and check the box labeled Dummy on the General tab.
Versions: DPL Professional, DPL Enterprise, DPL Portfolio