Comprehensive Guide to Advanced Python Programming

Introduction

In the previous article, we discussed the comprehensive overview of the “Python Programming Language and its built-in Data Structure.” We have also discussed the importance of learning Python to stay relevant in today’s competitive data science market, where many people have already been laid off due to the automation of tasks and the rise in Gen-AI and LLMs.

In this article, I’ll help you understand the core of the Advanced Python Topics, such as Classes and Generators, as well as some additional important topics from the perspective of a data scientist, along with a sample example.

By the end of this article, you will have a solid understanding of Python programming language, which is helpful for both interview preparation and day-to-day work as a data scientist and Python developer. By adopting these tips, you will write efficient code and boost your productivity while working with a team. 

Overview

1. Explore advanced Python concepts like Classes, Generators, and more tailored for data scientists.
2. Learn how to create custom objects and manipulate them effectively within Python.
3. Understand the power of Python Generators to save memory and streamline iteration processes.
4. Gain insights into various Python literals, including string, numeric, and Boolean types.
5. Enhance your coding efficiency with Python’s built-in functions and error management techniques.
6. Build a strong Python foundation, covering everything from basics to advanced topics, with practical examples for real-world application.

What is Advanced Python Programming?

Advanced Python Programming is studying and applying sophisticated Python concepts beyond basic programming. It includes topics like object-oriented programming (OOP), decorators, generators, context managers, and metaclasses. It also covers advanced data structures, algorithms, concurrency, parallelism, and techniques for optimizing code performance. Mastery of these concepts enables developers to write more efficient, scalable, and maintainable code suitable for complex applications in fields like data science, machine learning, web development, and software engineering.

A. Python Classes

Python allows the developer to create custom objects using the `class` keyword. The object’s blueprint can have attributes or encapsulated data, the methods or the class’s behavior.

Class Parentheses Optional but Not Function Parentheses

• It shows two ways to define a class: `class Container():` and `class Container:.`
• The parentheses after Container are optional for classes but are needed if you are inheriting from another class.
• In contrast, function parentheses are always required when defining a function.

class Container():
    def __init__(self, data):
        self.data = data

class Container: 
    def __init__(self, data):
        self.data = data

Wrapping a Primitive to Change Within a Function

The container is a simple class that wraps a primitive (in this case, an integer).

# The code defines a class called `Container` with a constructor method `__init__` 
# that takes a parameter `data` and assigns it to an instance variable `self.data`.
class Container:
    def __init__(self, data):
        self.data = data
    
def calculate(input):
    input.data **= 5
    
container = Container(5)
calculate(container)
print(container.data)

Output

3125

Compare Identity with “is” operator

c1 = Container(5)
c2 = Container(5)

print(id(c1), id(c2))
print(id(c1) == id(c2))  # returns False because they are different objects.
print(c1 is c2) # same objects but returns False because they are distinct instances.

Output

1274963509840 1274946106128

False

False

False

Now Compare by Value

The `eq method` added dynamically in the previous step is used for the equality check (c1 == c2).

c1 = Container(5)
c2 = Container(5)

print(c1 == c2)  # Compares by value
# This time, the result is True because the custom __eq__ method compares 
# the values inside the Container instances.

print(c1 is c2)  # Compares by identity (address)
# The is operator still returns False because it checks for identity.

Output

True

False

B. Python Generator

Generators are a special type of iterators, created using a function with the `yield` keyword, used to generate values on the fly.

Save memory with “Generators”

import sys

my_list = [i for i in range(1000)]
print(sum(my_list))
print("Size of list", sys.getsizeof(my_list), "bytes")


my_gen = (i for i in range(1000))
print(sum(my_gen))
print("Size of generator", sys.getsizeof(my_gen), "bytes")

Output

499500

Size of list 8856 bytes

499500

Size of generator 112 bytes

Fibonacci Generator and Yield

• `fib` is a generator function that generates Fibonacci numbers up to a specified count.
• `gen` is an instance of the generator, and next is used to retrieve the next values.
• The second loop demonstrates using a generator in a for loop to print the first 20 Fibonacci numbers.

def fib(count):
    a, b = 0, 1
    while count:
        yield a
        a, b = b, b + a
        count -= 1

gen = fib(100)
print(next(gen), next(gen), next(gen), next(gen), next(gen))

for i in fib(20):
    print(i, end=" ")

Output

0 1 1 2 3

0 1 1 2 3 5 8 13 21 34 55 89 144 233 377 610 987 1597 2584 4181

Infinite Number and Infinite Generator

• `math.inf` represents positive infinity.
• The `fib` generator is used with an infinite loop, and values are printed until the condition i >= 200 is met.

This code demonstrates positive infinity, a Fibonacci number generator, and how to break out of the generator loop based on a condition.

import math

# Printing Infinity: special floating-point representation 
print(math.inf)

# Assign infinity to a variable and perform an operation
inf = math.inf
print(inf, inf - 1)  # Always infinity, Even when subtracting 1, the result is still infinity


# Fibonacci Generator:
def fib(count):
    a, b = 0, 1
    while count:
        yield a
        a, b = b, b + a
        count -= 1

# Using the Fibonacci Generator:
# Use the Fibonacci generator with an infinite count
f = fib(math.inf)

# Iterate through the Fibonacci numbers until a condition is met
for i in f:
    if i >= 200:
        break
    print(i, end=" ")

Output

inf

inf inf

0 1 1 2 3 5 8 13 21 34 55 89 144

List from Generator

import math

# The code is creating a Fibonacci sequence generator using a generator function called `fib`.
def fib(count):
    a, b = 0, 1
    while count:
        yield a
        a, b = b, b + a
        count -= 1

# The `fib` function takes a parameter `count` which determines the number of Fibonacci numbers to generate.
f = fib(10)

# This code generates Fibonacci numbers and creates a list containing the square root of each Fibonacci number.
data = [round(math.sqrt(i), 3) for i in f]
print(data)

Output

[0.0, 1.0, 1.0, 1.414, 1.732, 2.236, 2.828, 3.606, 4.583, 5.831]

Simple Infinite Generator with “itertools”

The generator function could be simpler without having to take a max count property. This can be done easily with itertools.

import itertools

def fib():
    a, b = 0, 1
    while True:
        yield a
        a, b = b, b + a

# itertools.islice is used to get the first 20 values from an infinite generator.
print(list(itertools.islice(fib(), 20)))

Output

[0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89, 144, 233, 377, 610, 987, 1597, 2584, 4181]

Iterate Through Custom Type with iter

# Defines a custom LinkedList class with a Node class as an element.

class Node:
    def __init__(self, data, next_node=None):
        self.data = data
        self.next = next_node

class LinkedList:
    def __init__(self, start):
        self.start = start
    
    # The __iter__ method is implemented to allow iteration over the linked list.
    def __iter__(self):
        node = self.start
        while node:
            yield node
            node = node.next
            
ll = LinkedList(Node(5, Node(10, Node(15, Node(20)))))
for node in ll:
    print(node.data)

Output

5

10

15

20

C. Python Literals

Literals are the constants that provide variable values, which can be directly utilized later in expressions. They are just a syntax used in Python to express a fixed value of a specific data type.

For Example:

4x - 7 = 9

    # 4 : Coefficient
    # x : Variable
    # - and = : Operator
    # 7 and 9 : Literals (constants)

Types of Literals in Python

Python supports various types of literals, such as

In my previous article, I have discussed the collection literals, which you can refer to here. In this article, we will be discussing about : 

• String literals
• Numeric literals
• Boolean literals
• Special literals

Python String/Character Literals

A string literal is created by writing a text (i.e. the group of Characters) inside the single inverted commas(‘ ‘), double inverted commas(” “), or triple quotes (to store multi-line strings). 

For instance,

# in single quote
s = 'AnalyticalNikita.io'
print(s)

# in double quotes
d = "AnalyticalNikita.io"
print(d)

# multi-line String
m = '''Analytical
              Nikita.
                      io'''
print(m)

# Character Literal
char = "A"
print(char)

# Unicode Literal
unicodes = u"\u0041"
print(unicodes)

# Raw String
raw_str = r"raw \n string"
print(raw_str)

Output

AnalyticalNikita.io

AnalyticalNikita.io

Analytical

               Nikita.

                          io

A

A

raw \n string

Python Numeric Literals

There are three types of numeric literals in Python which are immutable by nature, namely:

1. Integer: These are both positive and negative numbers, including 0 — a decimal literal, a binary literal, an octal literal, and a hexadecimal literal. Note: While using the print function to display a value or to get the output, these literals are converted into decimals by default.
2. Float: These are the real numbers having both integer and fractional (denoted by decimal) parts.
3. Complex: These numerals are similar to the mathematics complex numbers, denote in the form of `a + bj`, where ‘a’ is the real part and ‘b‘ is the complex part.

# integer literal

# Binary Literals
a = 0b10100

# Decimal Literal
b = 50

# Octal Literal
c = 0o320

# Hexadecimal Literal
d = 0x12b

print(a, b, c, d)


# Float Literal
e = 24.8

print(e)

# Complex Literal
f = 2+3j
print(f)

Output

20 50 208 299

24.8

(2+3j)

Python Boolean Literals

Like other programming languages, there are only two Boolean literals in Python also, namely: True (or 1) and False (or 0). Python considers Boolean the same as a number in a mathematical expression.

Such as:

a = (1 == True)
b = (1 == False)
c = True + 4
d = False + 10

print("a is", a)
print("b is", b)
print("c:", c)
print("d:", d)

Output

a is True

b is False

c: 5

d: 10

Python Special Literal

Typically, ‘None’ is used to define a null variable.

hi = None
print(hi)

Output

None

Note: If  we compare ‘None’ with anything else other than a ‘None’, it will always return False.

hi = None
bye = "ok"
print(hi == bye)

Output

False

It is known as a special literal because, in Python, it is also used for variable declaration. If you do not know the number of variables, you can use `None`, as it will not throw any errors.

k = None
a = 7
print("Program is running..")

Output

Program is running..

D. Zip Function

We had already seen this function previously with respect to Python Built-in Data Structure when we had an equal lengths of iterables (such as lists, dictionaries, etc.) as arguments and `zip()` aggregates the iterators from each of the iterables. 

Using zip with equal number of iterators

# Example using zip with two lists
numbers = [1, 2, 3]
letters = ['a', 'b', 'c']

# Zip combines corresponding elements from both lists
zipped_result = zip(numbers, letters)

# Iterate over the zipped result
for number, letter in zipped_result:
    print(f"Number: {number}, Letter: {letter}")

Output

Number: 1, Letter: a

Number: 2, Letter: b

Number: 3, Letter: c

But what if we have an unequal number of iterators? In this case, we will use `zip_longest()` from the `itertools` module to aggregate the elements. If two lists have different lengths, it will aggregate `N/A`.

Using zip_longest from itertools

from itertools import zip_longest

# Example using zip_longest with two lists of different lengths
numbers = [1, 2, 3]
letters = ['a', 'b']

# zip_longest fills missing values with a specified fillvalue (default is None)
zipped_longest_result = zip_longest(numbers, letters, fillvalue='N/A')

# Iterate over the zipped_longest result
for number, letter in zipped_longest_result:
    print(f"Number: {number}, Letter: {letter}")

Output

Number: 1, Letter: a

Number: 2, Letter: b

Number: 3, Letter: N/A

Default Arguments

When you have default values, you can pass arguments by name; positional arguments must remain on the left.

from itertools import zip_longest


def zip_lists(list1=[], list2=[], longest=True):
    if longest:
        return [list(item) for item in zip_longest(list1, list2)]
    else:
        return [list(item) for item in zip(list1, list2)]
    
names = ['Alice', 'Bob', 'Eva', 'David', 'Sam', 'Ace']
points = [100, 250, 30, 600]

print(zip_lists(names, points))

Output

[['Alice', 100], ['Bob', 250], ['Eva', 30], ['David', 600], ['Sam', None], ['Ace', None]]

Keyword Arguments

You can also pass named arguments in any order and can skip them even.

from itertools import zip_longest


def zip_lists(list1=[], list2=[], longest=True):
    if longest:
        return [list(item) for item in zip_longest(list1, list2)]
    else:
        return [list(item) for item in zip(list1, list2)]


print(zip_lists(longest=True, list2=['Eva']))

Output

[[None, 'Eva']]

E. General Functions

“do-while” loop in python

while True:

    print("""Choose an option:
          1. Do this
          2. Do that
          3. Do this and that
          4. Quit""")
    
    # if input() == "4":
    if True: 
        break

Output

Choose an option:

1. Do this

2. Do that

3. Do this and that

4. Quit

enumerate() function insted of range(len())

fruits = ['apple', 'banana', 'kiwi', 'orange']

# Using enumerate to iterate over the list with both index and value
for index, fruit in enumerate(fruits):
    print(f"Index {index}: {fruit}")

print("\n")
# You can also specify a start index (default is 0)
for index, fruit in enumerate(fruits, start=1):
    print(f"Index {index}: {fruit}")

Output

Index 0: apple

Index 1: banana

Index 2: kiwi

Index 3: orange

Index 1: apple

Index 2: banana

Index 3: kiwi

Index 4: orange

Wait with `time.sleep()`

import time

def done():
    print("done")
   
def do_something(callback):
    time.sleep(2) # it will print output after some time for ex 2 means 2.0s
    print("Doing things....") # callback functions as an argument and prints "Doing things...." before calling the provided callback.
    callback() # Call the provided callback function 

    
# Call do_something with the done function as the callback
do_something(done)

Output

Doing things....

done

Sort complex iterables with `sorted()`

dictionary_data = [{"name": "Max", "age": 6},
                   {"name": "Max", "age": 61},
                   {"name": "Max", "age": 36},
                   ]

sorted_data = sorted(dictionary_data, key=lambda x : x["age"])
print("Sorted data: ", sorted_data)

Output

Sorted data: [{'name': 'Max', 'age': 6}, {'name': 'Max', 'age': 36}, {'name': 'Max', 'age': 61}] 

Get the Python version

If you’re curious to find the Python version on which you are working, you can use this code:

from platform import python_version

print(python_version())

Output

3.9.13

Get the Docstring of the objects

We can also use, `__doc__` to return the document of the functions, which provides all the details of the object, explaining its parameters and its default behavior.

print(print.__doc__)

Define default values in Dictionaries with .get() and .setdefault()

You can use `.setdefault()` function to insert key with a specified default value if the key is not already present in your dictionary.  Else `.get()` will return None, if the item has no specified key.

my_dict = {"name": "Max", "age": 6}                   
count = my_dict.get("count")
print("Count is there or not:", count)

# Setting default value if count is none
count = my_dict.setdefault("count", 9)
print("Count is there or not:", count)
print("Updated my_dict:", my_dict)

Output

Count is there or not: None

Count is there or not: 9

Updated my_dict: {'name': 'Max', 'age': 6, 'count': 9}

Using “Counter” from collections

• counter(): returns a dictionary of count elements in an iterable.

from collections import Counter

my_list = [1,2,1,2,2,2,4,3,4,4,5,4]
counter = Counter(my_list)
print("Count of the numbers are: ", counter)

most_commmon = counter.most_common(2) # passed in Number will denotes how many common numbers we want (counting starts from 1-n)
print("Most Common Number is: ", most_commmon[0]) # printin zeroth index element from 2 most common ones

Output

Count of the numbers are: Counter({2: 4, 4: 4, 1: 2, 3: 1, 5: 1})

Most Common Number is: (2, 4)

Merging two dictionaries using **

d1 = {"name": "Max", "age": 6}   
d2 = {"name": "Max", "city": "NY"}   

merged_dict = {**d1, **d2}
print("Here is merged dictionary: ", merged_dict)

Output

Here is merged dictionary: {'name': 'Max', 'age': 6, 'city': 'NY'}

F. Syntax Error Vs. Runtime Error 

There are mainly two types of errors that can occur in a program, namely:

1. Syntax errors: These types of errors occur during the time of compilation due to incorrect syntax. 
2. Runtime errors: These types of errors occur during the time of execution of the program also known as Exceptions in Python.

For hands-on experience and better understanding opt for the – Learn Python for Data Science Course

Conclusion

Congratulations! By now, I believe you have built a strong foundation in Python Programming. We have covered everything from Python Basics, including Operators and literals (numbers, strings, lists, dictionaries, sets, tuples), to Advanced Python topics such as Classes and Generators.

To level up your production-level coding skills, I have also discussed the two types of errors that can occur while writing a program. This way, you’ll be aware of them, and you can also refer to this article, where I have discussed how to Debug those Errors.

Additionally, I have compiled all the codes in a Jupyter Notebook, which you can — find here. These codes will serve as a quick future syntax reference.

Frequently Asked Questions

Nikita Prasad 13 Aug, 2024

Hi-ya!!! 👋 I'm Nikita Prasad Data Analyst | Machine Learning and Data Science Practitioner ↪️ Checkout my Projects- GitHub: https://github.com/nikitaprasad21 Know thy Author: 👩🏻‍💻 As an analyst I am eager to gain a deeper understanding of the data lifecycle with a range of tools and techniques, distilling down data for actionable takeaways using Data Analytics, ETL, Machine Learning, NLP, Sentence Transformers, Time-series Forecasting and Attention to Details to make recommendations across different business groups. Happy Learning! 🚀🌟