CS350: Data Structures Class Notes
Class 2: Wednesday, September 3, 2008
Announcements:
We are changing classrooms. We will now meet in Olin 212.
There are links to code discussed today on the class web page.
Abstract Data Types or ADT
When programming, we use modules/modularity to partition our code for three reasons:
• It simplifies the task of coding—we can focus on individual task rather than the
whole problem
• It allows us to eliminate redundancies—we can reuse code
• It allows us to isolate errors—this simplifies debugging and unit testing
The key ideas are information hiding and encapsulation.
There are two parts to what we are doing. We want to separate the “what” from the
“how”. Both must be addressed, but it is easier if we do this separately. The “what”
defines the external interface. The “how” is the implementation.
An ADT is composed of two parts, the data and the operations on the data. Contrasted to
a data structure in a programming language, the data structure only include the data, not
the operations.
E.g., an ADT for Booleans would include: TRUE, FALSE, AND, OR, and NOT.
The correspondence between an ADT and a class definition in Python should be obvious.
Rational Numbers
Data (a/b)
numerator, denominator
Operations
Rational—constructor
numIs—accesser function that returns the numerator
denIs—accesser function that returns the denominator
+ or __add__—adds two rational numbers: a/b+c/d (ad+bc)/bd
– or __sub__—subtract two rational numbers: a/b-c/d (ad–bc)/bd
* or __mul__—multiplication: a/b*c/d ac/bd
/ or __div__—division: a/b / c/d ad/bc
<, ==, >, etc. or __cmp__—comparison –1 for less-than, 0 for equal, +1 for greater-than
Other issues including printing, etc.
