Skip to main content

Modules, Imports, and Common Standard Libraries

One of Python's strengths is its rich standard library. Many needs don't require writing from scratch -- thinking first about "does the standard library already have this?" can usually save a lot of time.

Basic import syntax​

import math
import json

Access functions from a module using the module name:

print(math.sqrt(16))

You can also import only the objects you need:

from math import sqrt, pi

If a module name is long, you can alias it:

import numpy as np
import pandas as pd

Modules and packages​

  • A single .py file can be considered a module.
  • A directory containing multiple modules can typically be considered a package.

When files start multiplying, I prefer splitting by responsibility rather than piling all functions into one script.

if __name__ == "__main__"​

The purpose of this check is to make a file both importable and directly runnable:

def main():
    print("run as script")


if __name__ == "__main__":
    main()

Common scenarios:

  • When imported: only reuse functions, don't execute test code
  • When run directly: execute demos, debugging, or command-line entry points

Frequently Used Standard Libraries​

1. math​

Suitable for basic math operations:

import math

print(math.sqrt(9))
print(math.floor(3.8))
print(math.pi)

2. collections​

My most commonly used items:

  • Counter: Counting
  • deque: Double-ended queue
  • defaultdict: Dictionary with default values
from collections import Counter, deque, defaultdict

print(Counter("banana"))
queue = deque([1, 2, 3])
mapping = defaultdict(list)

3. pathlib​

For handling paths in modern Python, I recommend pathlib over string concatenation everywhere:

from pathlib import Path

root = Path("data")
file_path = root / "users.csv"
print(file_path.exists())

4. json​

Suitable for handling JSON files and API responses:

import json

payload = {"name": "alice", "score": 95}
text = json.dumps(payload, ensure_ascii=False)
restored = json.loads(text)

5. itertools​

This module is great for handling combinations, permutations, and iterators:

from itertools import product

for pair in product([1, 2], ["a", "b"]):
    print(pair)

When to write your own vs. check the standard library first​

My current habit:

  • First determine if it's a common problem: paths, counting, queues, JSON, time, combinations, random numbers
  • If so, check the standard library first
  • Only when the standard library isn't enough, bring in third-party libraries