Chapter 11: Sorting

Sorting is a fundamental operation in data science - from organizing results to preparing data for analysis to optimizing algorithm performance. Python provides powerful built-in sorting capabilities that are both fast and flexible, enabling you to sort any data structure with custom logic.

The sorted() Function

Python's sorted() function returns a new sorted list:

sorted() creates a new list, leaving the original unchanged. It works with any iterable.

Alphabetical Sorting

Sorting strings uses alphabetical (lexicographic) order:

Uppercase letters come before lowercase in ASCII sorting. Use str.lower() as a key for case-insensitive sorting.

Reverse Sorting

The reverse parameter sorts in descending order:

reverse=True inverts the sort order for any sorting operation.

The list.sort() Method

Lists have a .sort() method that sorts in-place:

.sort() modifies the original list and returns None. Use it when you don't need to preserve the original order.

sorted() vs list.sort()

Key differences between the two approaches:

Use sorted() for immutable objects or when preserving the original. Use .sort() for efficiency when you don't need the original.

Sorting Dictionaries

Sorting a dictionary sorts its keys:

To sort by values, use the key parameter with a lambda function.

Sorting Tuples

Tuples sort by first element, then second, and so on:

Tuples provide natural sorting for paired data.

Sorting with Custom Keys

The key parameter lets you specify custom sorting logic:

The key function extracts the value to sort by. It's called once per element.

Sorting by String Length

Common use case: sort strings by length:

key=len passes the built-in len function as the sorting key.

Custom Sort Functions

Define complex sorting logic with custom functions:

Return tuples from key functions for multi-level sorting (primary, secondary, etc.).

Sorting Objects

Sort custom objects using lambda or key functions:

Lambda functions provide concise attribute access for sorting objects.

Sorting by Multiple Criteria

Sort by multiple attributes using tuples:

Use negative values in tuples to reverse specific sort criteria.

Case-Insensitive Sorting

Handle mixed-case strings:

key=str.lower normalizes strings for comparison without modifying them.

Sorting with operator Module

The operator module provides efficient key functions:

itemgetter is faster than lambda for simple index/attribute access.

Sorting with attrgetter

For object attributes, use attrgetter:

attrgetter is cleaner and faster than lambda for attribute access.

Stable Sorting

Python's sort is stable - equal elements maintain their original order:

Stable sorting preserves relative order, useful for multi-pass sorting.

Sorting Pandas Series

Pandas Series have built-in sorting methods:

sort_values() returns a new sorted Series.

Sorting Pandas DataFrames

DataFrames sort by columns:

Use by parameter to specify sort column(s).

Sorting by Multiple Columns

Sort DataFrames by multiple columns with different orders:

The ascending parameter can be a list matching the columns in by.

In-Place DataFrame Sorting

Sort DataFrames in-place to save memory:

inplace=True modifies the DataFrame directly without creating a copy.

Sorting with Missing Values

Handle NaN values in sorting:

Use na_position='first' or 'last' to control NaN placement.

Practical Example: Ranking

Create rankings based on sorted data:

Combine sorting with enumeration to create rankings.

Practical Example: Top N

Extract top performers using sorting:

Combine sort_values() with head() or tail() for top/bottom N selection.

Quiz: Test Your Knowledge

Summary

Sorting is essential for organizing and analyzing data. Python provides flexible, efficient sorting through sorted() and list.sort(), with customization via key functions. Pandas adds powerful DataFrame sorting with sort_values() for multi-column operations.

Key takeaways: