Leapcell

Posted on Jan 30

Functional Programming in Python: A Deep Dive

#python #django #flask #fastapi

Overview of Functional Programming

Functional programming refers to a style where every part of the code is immutable and consists of pure functions. A pure function is one that is independent of others and, given the same input, always produces the same output.

In addition, functional programming has the characteristic of allowing a function to be passed as an argument to another function and also to return a function.

Examples of Pure Functions

For example, to double the values of elements in a list, the following function can be used:

def multiply_2(list):
    for index in range(0, len(list)):
        list[index] *= 2
    return list

It should be noted that this code is not in the form of a pure function because the values of the elements in the list are modified. If the multiply_2() function is called multiple times, the results will be different each time.

To make multiply_2() a pure function, a new list needs to be created and returned, like this:

def multiply_2_pure(list):
    new_list = []
    for item in list:
        new_list.append(item * 2)
    return new_list

Advantages and Disadvantages of Functional Programming

The advantages of functional programming mainly lie in the fact that the pure function and immutable characteristics make the program more robust and easier to debug and test. The disadvantages are mainly that there are many restrictions and it is difficult to write.

Note that in pure functional programming languages (such as Scala), there are no variables in the functions written, so it can be guaranteed that as long as the input is determined, the output is determined. In programming languages that allow the use of variables, due to the uncertain state of variables inside the function, the same input may result in different outputs.

Python allows the use of variables, so it is not a pure functional programming language. Python only provides partial support for functional programming, mainly including the three functions map(), filter(), and reduce(), which are usually used together with lambda anonymous functions. Next, the usage of these three functions will be introduced one by one.

Python's Functional Programming Support Functions

Python map() Function

The basic syntax of the map() function is as follows:

map(function, iterable)

Here, the function parameter represents a function to be passed in, which can be a built - in function, a custom function, or a lambda anonymous function. The iterable represents one or more iterable objects, such as lists, strings, etc.

The function of the map() function is to call the specified function for each element in the iterable object and return a map object.
Note that the function returns a map object, which cannot be output directly. It can be displayed through a for loop or the list() function.

[Example 1] Multiply each element in the list by 2

listDemo = [1, 2, 3, 4, 5]
new_list = map(lambda x: x * 2, listDemo)
print(list(new_list))

The running result is:

[2, 4, 6, 8, 10]

[Example 2] Pass multiple iterable objects as parameters to the map() function

listDemo1 = [1, 2, 3, 4, 5]
listDemo2 = [3, 4, 5, 6, 7]
new_list = map(lambda x,y: x + y, listDemo1,listDemo2)
print(list(new_list))

The running result is:

[4, 6, 8, 10, 12]

Note that since the map() function is directly written in C language, it does not need to be called indirectly through the Python interpreter during runtime, and many optimizations have been made internally. Therefore, compared with other methods, this method has the highest running efficiency.

Python filter() Function

The basic syntax of the filter() function is as follows:

filter(function, iterable)

In this format, the funcition parameter represents a function to be passed in, and the iterable represents an iterable object.

The function of the filter() function is to use the function function to judge each element in iterable and return True or False. Finally, the elements that return True are combined into a new traversable collection.

[Example 3] Return all even numbers in a list

listDemo = [1, 2, 3, 4, 5]
new_list = filter(lambda x: x % 2 == 0, listDemo)
print(list(new_list))

The running result is:

[2, 4]

[Example 4] The filter() function accepts multiple iterable objects

listDemo = [1, 2, 3, 4, 5]
new_list = map(lambda x,y: x-y>0,[3,5,6],[1,5,8] )
print(list(new_list))

The running result is:

[True, False, False]

Python reduce() Function

The reduce() function is usually used to perform some cumulative operations on a collection. Its basic syntax is:

reduce(function, iterable)

Here, the function must be a function with two parameters, and the iterable represents an iterable object.

Note that since the reduce() function has been removed in Python 3.x and placed in the functools module, the functools module needs to be imported before using this function.

[Example 5] Calculate the product of elements in a list

import functools
listDemo = [1, 2, 3, 4, 5]
product = functools.reduce(lambda x, y: x * y, listDemo)
print(product)

The running result is:

Summary

Generally, when performing some operations on elements in a collection, if the operations are very simple, such as addition, accumulation, etc., map(), filter(), and reduce() should be considered first. In addition, when the data volume is very large (such as in machine learning applications), functional programming representation is usually preferred because of its higher efficiency.

Of course, when the data volume is not large, methods such as for loops can also be used. However, if some more complex operations need to be performed on the elements in the collection, considering the readability of the code, for loops are usually used.