How to Import a Library in Python

What you’ll build or solve

You’ll import a library the right way in Python, then use it without ImportError or confusing name clashes.

When this approach works best

This approach works best when you:

• Add a new library (like requests) and want to import it in a script.
• Organize your code into multiple files and import between them.
• Work in a project where the same code needs to run on different machines.

Avoid this approach when:

• You paste one small snippet into a Python REPL once. You can still import there, but you may not need project structure.

Prerequisites

• Python 3 installed
• You know what variables are
• You can run a .py file from the terminal or your editor

Note: People often say “library,” “module,” and “package” loosely to mean “something you can import.” This guide uses “library” to mean any importable code.

Step-by-step instructions

1) Import a built-in library

Built-in libraries ship with Python, so you can import them without installing anything.

import math

print(math.sqrt(49))

If it runs without installation, it’s part of the standard library.

2) Import a third-party library you installed

Third-party libraries come from pip. Install first, then import. For detailed pip usage and troubleshooting, see the guide on installing modules.

Install:

python -m pip install requests

Then use it:

import requests

resp = requests.get("https://example.com", timeout=10)
print(resp.status_code)

What to look for:

Use python -m pip so you install into the same Python that runs your script.

3) Import only what you need

This keeps your code shorter and clearer.

from math import sqrt, pi

print(sqrt(81))
print(pi)

If you want a shorter name, use an alias:

import datetime as dt

print(dt.date.today())

Aliases are common with longer names, like:

import numpy as np

4) Import your own code from another file

Create two files in the same folder.

utils.py:

def greet(name: str) -> str:
    return f"Hi, {name}!"

app.py:

from utils import greet

print(greet("Sam"))

Run:

python app.py

What to look for:

Both files must be in the same folder, and you must run the script from that folder.

5) Import from a package (folders with modules)

Use this structure:

my_project/
  app.py
  my_package/
    __init__.py
    utils.py

my_package/utils.py:

def slugify(text: str) -> str:
    return text.strip().lower().replace(" ", "-")

app.py:

from my_package.utils import slugify

print(slugify("Hello Imports"))

Run from the project root:

python app.py

What to look for:

Run python app.py from the root folder, not from inside my_package/.

Examples you can copy

1) Standard library import for file paths

from pathlib import Path

path = Path("data") / "notes.txt"
print(path)

2) Third-party import and version check

import requests

print(requests.__version__)

3) Cleaner imports with aliases

import numpy as np

arr = np.array([1, 2, 3])
print(arr.mean())

Common mistakes and how to fix them

Mistake 1: Naming your file the same as the library

You create a file named random.py, then write:

import random

print(random.randint(1, 10))

Why it breaks:

Python imports your local random.py instead of the real random library.

Correct approach:

• Rename your file to something else, like random_demo.py.
• Delete the __pycache__/ folder if it exists.
• Run your script again.

Mistake 2: Installing a package but still getting ModuleNotFoundError

You run:

pip install requests
python app.py

Why it breaks:

pip might install into a different Python than the one running your script.

Correct approach:

python -m pip install requests
python app.py

Mistake 3: Using a relative import in a script and getting an error

Inside my_package/utils.py:

from .helpers import clean

Then you run:

python my_package/utils.py

Why it breaks:

Relative imports need a package context. Running the file directly skips that context.

Correct approach:

Run from the project root:

python -m my_package

Or run a top-level script:

python app.py

Troubleshooting

If you see ModuleNotFoundError right after import x, run:

python -c "import sys; print(sys.executable)"

Then install with:

python -m pip install x

If you see ImportError: attempted relative import with no known parent package, stop running the module file directly. Run from the project root with:

python -m package_name

If an import works on your laptop but not another machine, create a requirements.txt file and install from it on both machines.

If your editor shows red underlines but the script runs, select the correct Python interpreter in the editor and restart it.

Quick recap

• Use import math for built-in libraries.
• Install third-party libraries with python -m pip install package, then import package.
• Use from x import y or import x as alias for cleaner code.
• Import your own files with from utils import func when files share a folder.
• Import packages with from my_package.module import thing and run from the project root.
• If imports fail, check the Python interpreter, current folder, and name conflicts.