PyInns - Home

List Comprehensions in Python – Best Practices for Data Science 2026

List comprehensions are one of Python’s most elegant and powerful features. They allow you to create new lists by transforming and filtering existing iterables in a single, readable line. In data science, they are widely used for data cleaning, feature engineering, and transforming datasets.

TL;DR — Basic Structure

[expression for item in iterable if condition]
More readable than traditional for loops with append()
Faster and more Pythonic for simple transformations

1. Basic List Comprehensions

scores = [85, 92, 78, 95, 88, 76, 91]

# Filter and transform in one line
high_scores = [score for score in scores if score >= 90]
squared_scores = [score ** 2 for score in scores]

print(high_scores)
print(squared_scores)

2. Real-World Data Science Examples

import pandas as pd

df = pd.read_csv("sales_data.csv")

# Example 1: Create new columns using list comprehension
df["log_amount"] = [0 if x <= 0 else round(x ** 0.5, 2) for x in df["amount"]]

# Example 2: Filter and extract specific data
high_value_customers = [
    (row.customer_id, row.amount, row.region) 
    for row in df.itertuples() 
    if row.amount > 1500
]

# Example 3: Clean and transform text data
clean_names = [name.strip().title() for name in df["customer_name"] if isinstance(name, str)]

3. When to Use List Comprehensions vs Traditional Loops

# Good use of list comprehension - simple and clean
prices = [float(x) for x in price_strings if x.strip() != ""]

# Better as traditional loop when logic is complex
processed = []
for x in price_strings:
    cleaned = x.strip()
    if cleaned:
        try:
            processed.append(float(cleaned))
        except ValueError:
            continue

4. Best Practices in 2026

Use list comprehensions for simple filtering and transformations
Avoid very complex list comprehensions — they can become hard to read
For complex logic, use a traditional for loop with .append()
Consider generator expressions (`(...)`) when working with very large data to save memory
Keep comprehensions focused on one clear transformation

Conclusion

List comprehensions are a hallmark of Pythonic code in data science. In 2026, they are the preferred way to create new lists through simple filtering and transformation. Use them liberally for clean, concise code, but switch to traditional for loops when the logic becomes complex or requires multiple steps. The goal is readability and maintainability, not cleverness.

Next steps: