A model is a set of causal dependencies and equations defining the mathematical relationship among variables. A view is a visual representation of some subset of those relationships. Models are complete but might be complicated. Views, even for the most complicated model, can be made simple and clean. Most models are a collection of overlapping views, each view showing a simple portion of the underlying model.
Every view can have different content and different visual characteristics. Most commonly views are used to represent different parts of a model, but they can also be used to represent the same part in different ways. For example, it is possible to have one view that is a stock and flow representation and a second that is a causal loop representation of the same structure.
NOTE For models with one view the model and the view are essentially the same thing.
Having many views allows you to:
| • Represent complex models with combinations of simple diagrams. |
| • | Visualize local structure accurately and completely without having to worry about disturbing other clearly presented structures. |
| • | Make distinct representations of the same concepts for different purposes. |
| • | Localize structure for making and testing changes. |
When you make changes to a view you might, or might not, affect other views. Changes that affect only the appearance of a view and do not change the logical structure of the model do not influence any other view. Thus, you can remove variables from a view, change the colors and fonts in a view, or reposition variables within a view without affecting any other views.
On the other hand, changes that do affect the logical structure of a model will change related views. For example, if you introduce a new causal connection between two variables, then that causal connection will be duplicated in all views containing those variables. Similarly, if you delete a variable from the model, it will be deleted from all views.
When you look at a view, it will be updated with changes you have made to other views. Vensim introduces any necessary changes in the simplest possible way. For example, if you have added a causal connection from Population to births, a straight arrow will connect these two variables in other views. If another view has the variable births but not Population, then Population will be added as a Shadow variable to that view, with a straight line connection made between the two.
The continual updating of views guarantees that every view is always completely accurate without your intervention.
