Just as there are several methods for organizing entire software systems, there are different strategies for formally expressing computational processes within those structures.
188.8.131.52 Imperative computation
Most of you have probably been exposed to an imperative model of computer application, in which we taught of programs as giving a sequential set of instructions to the computer to work something. And, fact that is how the internals of the processors of computers are typically created. So, in C or JAVA or C++, you write typically methods that has of lists of instructions to the computer:
1. Put this value in this variable
2. Square the variable
3. Divide it by pi
4. If the result is larger than 1, return the result
In this model of computation, the primitive computational elements are basic arithmetic operations and assignment instructions. We may combine the elements using sequences of statements, and functional structures such as if, for, and while. We can give away from the details of a computation by defining a procedure that does it. Now, the person only requires to know the specification of the method, but not the implementation information, in order to use it.
184.108.40.206 Functional computation
Another style of programming is the functional model. In this model, we gain power through function invokes. Rather than describing the computer to do things, we ask it questions: What is 4 + 5? What is the square root of 6? What is the largest element of the list?
These questions can all be expressed as asking for the value of a function applied to some arguments. But where do the method come from? The answer is, from other methods. We initiate with some set of basic methods (like "plus"), and use them to construct more complex functions.
This method would not be powerful without the mechanisms of conditional evaluation and re cursion. Conditional functions tells one question under some situations and another question under other situations. Recursion is a method that lets the definition of a function refer to the function being described. Recursion is as useful as iteration.
In this model of computation, the primitive computational elements are typically basic arithmetic and list functions. We combine components using function composition (using the output of one function as the input to other function), if, and recursion. We use method definition as a method of abstraction, and the concept of higher-order functions (passing functions as arguments to other functions) as a way of capturing common high-level patterns.
220.127.116.11 Data structures
In either type of asking the computer to perform work for us, we have another kind of modularity and abstraction, which is centered around the organization of data.
At the most primitive level, computers operate on collections of (usually 32 or 64) bits. We may halt such a collection of bits as representing different things: a positive number, a signed value number, a ?oating-point value, a Boolean number (false or true), characters, or an address of some other data in the memory of the system. Python provides us the ability to work directly with all of those primitives.
There is only so much you can do with a single value, though. We could like to create computer programs that perform on representations of documents or maps or circuits or social networks. To do so, we require to aggregate primitive data elements into more complex data structures. These can added arrays, lists, dictionaries, and other data structure of our own devising.
Here, again, we gain the power of abstraction. We may prepare programs that do operations on a data structure representing a social network, as like, without having to work about the details of how the social network is represented in terms of bits in the machine.
18.104.22.168 Object-oriented programming: computation + data structures
Object-oriented programming is a model that gives the ideas of modularity and abstraction to execution and data at the same time.
An object is a data structure, together with a set of procedures that operate on the data. Basic methods may be written in an imperative or a functional model, but in the end there is imperative assignment to state variables in the object.
One major new type of abstraction in OO programming is "generic" programming. It may be that all objects have a method called print related with them. So, we may ask any object to print itself, without knowing how it is created. Or, in a graphics system, we may have several of different objects that know their x, y coordinates on the screen. So each of them can be asked, in the similar way, to say what their position is, even though they may be shown very differently inside the objects.
In addition, most object-oriented systems support inheritance, which is a way to make new types of objects by saying that they are mostly like another type of object, but with some errors. This is another way to take advantage of abstraction.
Python as well as other modern programming languages, such as Ruby, Java and C++, support all of these programming models. The programme needs to choose which programming model best suits the program. This is a point that we will give to throughout the course.