Stop 03 · Python for Data ScienceTopics · 9/9

Writing Clean, Reusable Code

In short

A notebook of 200 copy-pasted cells works once. Clean functions with clear names work next month, for your teammate, and in production. It's the habit that gets data scientists promoted.

Turning ad-hoc notebook cells into functions and modules that can be reused across projects — a habit that separates hobbyist scripts from production-adjacent data science work.

The mascot arranges small spice bowls inside a round steel masala box.

Think of it like this

Mummy's masala dabba

Every spice sits in its own small bowl in the round steel box, always in the same spot. Anyone in the family can cook from it without asking. Clean code is the same: each job in its own named function, kept where people expect it.

Watch it happen

Same analysis, two styles

Copy-paste notebook

Same cleaning pasted in five cells
Names like df2, df_final_v3
One fix means five edits

Small functions

load_orders() used everywhere
Clear names and a docstring
Fix once, rerun everything

Explore freely in a notebook, then move repeated steps into named functions in a .py file you import.

The big ideas

What you need to know

One function, one job

load, clean, summarise. Small functions are easier to test, reuse and read.

Names explain intent

monthly_revenue beats df3. A good name saves a comment.

Move code into modules

Put shared functions in a utils.py file and import them into any notebook.

Format and test

Tools like ruff or black tidy formatting. A few pytest tests catch broken cleaning logic early.

See it in code

Walk through the code

python

1import pandas as pd2 3def load_orders(path: str, min_amount: float = 0) -> pd.DataFrame:4    """Read orders, fix types and drop tiny test orders."""5    df = pd.read_csv(path, parse_dates=["date"])6    df["city"] = df["city"].str.strip().str.lower()7    return df[df["amount"] >= min_amount]8 9def city_summary(df: pd.DataFrame) -> pd.DataFrame:10    return df.groupby("city")["amount"].agg(["count", "sum"])11 12if __name__ == "__main__":13    print(city_summary(load_orders("orders.csv", min_amount=10)))

1/4Type hints and a docstring say what goes in, what comes out and why.

Two small, named functions with type hints and a docstring. Import them into any notebook with from utils import load_orders.

Save yourself the pain

Common mistakes

Oops

Hardcoding file paths and magic numbers deep inside functions.

Do this instead

Pass them in as parameters with sensible defaults, like min_amount=0.

Oops

Writing one giant function that loads, cleans, models and plots.

Do this instead

Split it. Each piece can then be reused and tested separately.

Your mission

Refactor an old notebook

~40 min

Quick check

What does if __name__ == "__main__": do in a Python file?

Go deeper

1Refactor a messy notebook into a set of clearly named, reusable functions.
2Learn to add docstrings explaining what a function does and what it expects.
3Practice moving reusable logic out of a notebook into a separate Python script.