Grading

For the most part, grading considers the following points.

1. Documentation of Requirements  
   • A short section on what you think the problem is.
   • List of requirements and ``pointers'' from the list to test cases verifying requirements met.
2. Preliminary Design (Architecture). 
   • Do type and routine names make sense?
   • Does choice of types make sense?
   • Pre- and post-conditions on routines. Pointers to tests verifying conditions met.
   • Organization of types as to best use of space.
   • Completeness of the design. What level does it get to?
3. Code 
   • Does the code have useful comments or just filler?
   • Are overloads and operator definitions obvious?
   • Use of exceptions (throw/catch/try)
   • Maximum reuse of previous assignments?
4. Testing 
   • Test data complete and reasonable.
   • Primitives completely tested?
   • Are derived routines correctly tested?
   • Is the test output usable and in the document?
5. Learning Recap available from every member? 

The weights heavily favor points 1, 2, and 4 at a total of with respective weights 20%, 30%, and 30% respectively. The coding part gets the remaining 20%. Item 5 will be required of each individual independently. This becomes a 20 point homework assignment. It is graded on how convincing you are about what you learned.