Sometimes a matrix or a table works better than diagrams. A
matrix is especially good if connections between model elements are important. A
matrix also scales better than a diagram since more information can be shown in the
same space. With matrix representation we thus don??™t often need to partition model
elements into submodels or parallel models. Table representation is especially good in
showing properties of modeling elements: a parameter table is a classic example. The
challenge in a table is showing the dependencies among the elements.
The choice of the representational formcan also depend on the model manipulation
actions. For example, a matrix gives a basis to automatically identify high cohesion and
low coupling between model elements with diagonalization. Tables and matrices also
offer other model manipulation options, like sorting based on model information. A
classic example is ?¬?nding priorities based on the properties of model elements.
Diagrams are particularly good in ?¬?nding patterns and organizing model elements into
subdesigns. Ideally, we don??™t need to ?¬?x on one representational style. The samemodel
can be shown in different representational forms and model manipulation operations
could take place in any of the possible representations.
10.6.2 Symbol De?¬?nition Guidelines
Although each notation will look different depending on the domain, some principles
can be generalized.
Pages:
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475