Lab#2

You will design and implement GradeEntry, LetterGradeEntry, and NumericGradeEntry below, so that you can create and run a le test lab02.py with code very similar to:if __name__ == ’__main__’: grades = [NumericGradeEntry(’csc148’, 87, 1.0), NumericGradeEntry(’bio150’, 76, 2.0), LetterGradeEntry(’his450’, ’B+’, 1.0)] for g in grades: # Use appropriate ??? methods or attributes of g in format print("Weight: {}, grade: {}, points: {}".format(g.?, g.??, g.???)) # Use methods or attributes of g to compute weight times points total = sum( # sum of the list of... [g.? * g.?? # ? and ?? are methods or attributes of g for g in grades]) # using each g in grades # sum up the credits total_weight = sum([g.? for g in grades]) print("GPA = {}".format(total / total_weight)) Important: Notice that the code above never checks whether a particular g is a LetterGradeEntry versus a NumericGradeEntry. You design the classes below so that it just does the right thing! Also notice the list comprehension: [g.? * g.?? for g in grades] Check out list comprehensions if needed. setup We assume that you either remember some of the setup techniques from last week, look for them in the lab#1 handout or consult your TA and other students. You?ll need to: 1  Log into your cdf account, start up Pycharm or another IDE for Python programs, and navigate to csc148/Labs/lab02  Open a web browser and navigate to the page with lab02 materials: http://www.cdf.toronto.edu/~csc148h/winter/Labs/lab02/ Here you'll nd materials for lab#2.  Download the le specs.txt from among the lab#2 materials, and save it under your own lab02 subdirectory. Open specs.txt in Pycharm (or some other IDE), and read through it. Check with your TA before moving on, in order to reassure yourself that you're on the right track. design GradeEntry Begin by designing the public interface of class GradeEntry, using the instructions below. You are not intended to be able to create instances of class GradeEntry, rather through inheritance you will be creating subclasses of GradeEntry. Here's what you need to do: 1. Create and open a new le with your editor called grade.py in the subdirectory lab02. 2. Perform an object-oriented analysis of the speci cations in specs.txt, following the same recipe we used in class for Point, Shape, Rational, etc.: (a) choose a class name (GradeEntry) and write a brief description in the class docstring. (b) write some examples of client code that uses your class (c) decide what services your class should provide as public methods, for each method declare an API1 (examples, header, type contract, description) (d) decide which attributes your class should provide without calling a method, list them in the class docstring Show your work to a TA before proceeding, in order to be reassured you are on the right track. implement GradeEntry The design is where the hard thinking should take place, so now it's time to implement GradeEntry. Remember that there is at least one method in GradeEntry that should only be implemented in its subclasses. In that case (or cases) the implementation for the method is an easy one-liner: raise NotImplementedError(’Subclass needed’) Here's what you need to do: 1. write the body of special methods __init__, __eq__, and __str__ 2. write the body of other methods Notice that there is no practical way to try out instances of GradeEntry until you have created one or more subclasses. There will be one or more methods that generate those annoying NotImplementedErrors. 1use the CSC108 function design recipe 2 design NumericGradeEntry The procedure for designing a subclass is similar to design of class GradeEntry, with a few important di erences: 1. Rather than class NumericGradeEntry:, your declaration will be class NumericGradeEntry(GradeEntry):. This tells Python (and human readers) that this class inherits attributes and methods from GradeEntry 2. Attributes and methods that you will use unchanged from GradeEntry (we say you inherit these) need not be mentioned in the class docstring. New attributes of NumericGradeEntry should be documented in the class docstring, as usual. 3. If you have new attributes for NumericGradeEntry that were not in GradeEntry, you will need to redesign the init method. This will include writing a new docstring for the method that says that NumericGradeEntry's initializer extends the initializer of GradeEntry, gives a (probably) new header and type contract, a new example of a call to the initializer, and says what the initializer does to the new attributes. 4. You will have (at least) one method that could not be implemented in class GradeEntry, but can now be implemented in NumericGradeEntry. Your docstring should say that the method in NumericGradeEntry (which should have the same name as in GradeEntry) overrides the corresponding method in GradeEntry. You should be careful that the type contract for the method in NumericGradeEntryEntry is consistent with the type contract for the corresponding function in GradeEntry. A good mental exercise is to convince yourself that any client code that uses an instance of GradeEntry without knowing that it is a LetterGradeEntry should experience results that are consistent with the public interface of GradeEntry.2 Show your work to a TA before moving on. design LetterGradeEntry This will be very similar to the procedure for designing NumericGradeEntry. Once again, devote special attention to the method(s) that were not implemented in GradeEntry, as well as the attributes that are new in LetterGradeEntry. implement both NumericGradeEntry and LetterGradeEntry Here is where we experience the labour-saving features of declaring subclasses. If we designed GradeEntry carefully, there should be a lot of code that was written in GradeEntrythat does not have to be re-written in NumericGradeEntry and LetterGradeEntry | the code is already there, through inheritance. Avoiding duplicate code also avoids many errors. For the attributes and methods of GradeEntry that will be used unchanged in these subclasses your implementation consists of...nothing. You did the work in GradeEntry. If you don't initialize any new attributes in a subclasses, you don't need to have additional documentation. We have already discussed how to document an init method that extends the one from GradeEntry under design. Here is how you would implement the extended init : 1. The rst statement in your implementation should be GradeEntry.__init__(self, ???) 2This subtle, yet very important, idea is a consequence of the Liskov Substitution Principle, and is worth taking the time to think through. ...where the question marks indicate possible values GradeEntry's initializer needs. 2. after the rst line, add code to initialize your new attributes that aren't inherited from GradeEntry. You should have some method(s) that are not inherited from GradeEntry. You should implement these as usual, even if they share a name with the corresponding method in GradeEntry. additional exercises As well as the various grade entries, here are some additional exercises in designing and implementing classes with inheritance. We set up each one so that one appropriate solution involves class Roster together with one of its subclasses. We certainly don't expect you to do this many exercises in the lab, but they are here for additional practice. We'll have a brief quiz in the last 15 minutes of the lab, which will involve an exercise similar to one of the additional exercises.
Powered by