Lesson 8: Sorting and Ranking

Course: Data Engineering | Duration: 35 minutes | Level: Intermediate

Learning Objectives

By the end of this lesson, you will be able to:

• Sort a DataFrame by one or multiple columns using .sort_values()
• Sort a DataFrame by its row index using .sort_index()
• Assign ranks to rows using .rank()
• Retrieve the top or bottom N rows with .nlargest() and .nsmallest()

Prerequisites

• Lesson 7: Data Types and Type Inspection

Lesson Outline

Part 1: .sort_values() — Sort by Column Values

.sort_values() returns a new DataFrame sorted by the specified column(s).

import pandas as pd
 
df = pd.DataFrame({
    'name':       ['Alice', 'Bob', 'Carol', 'David', 'Eve'],
    'department': ['Engineering', 'Marketing', 'Engineering', 'Marketing', 'Engineering'],
    'salary':     [95000, 72000, 88000, 68000, 105000]
})
 
# Sort by salary ascending (default)
print(df.sort_values('salary'))
#     name   department  salary
# 3  David    Marketing   68000
# 1    Bob    Marketing   72000
# 2  Carol  Engineering   88000
# 0  Alice  Engineering   95000
# 4    Eve  Engineering  105000
 
# Sort by salary descending
print(df.sort_values('salary', ascending=False))
#     name   department  salary
# 4    Eve  Engineering  105000
# 0  Alice  Engineering   95000
# 2  Carol  Engineering   88000
# 1    Bob    Marketing   72000
# 3  David    Marketing   68000

Multi-column sort: pass a list of column names. The second column is used as a tiebreaker.

# Sort by department ascending, then by salary descending within each department
print(df.sort_values(['department', 'salary'], ascending=[True, False]))
#     name   department  salary
# 4    Eve  Engineering  105000
# 0  Alice  Engineering   95000
# 2  Carol  Engineering   88000
# 1    Bob    Marketing   72000
# 3  David    Marketing   68000

Part 2: .sort_index() — Sort by Row Index

.sort_index() sorts rows by their index labels. Useful after filtering (which preserves original row numbers) or after setting a custom index.

import pandas as pd
 
df = pd.DataFrame({
    'name':   ['Alice', 'Bob', 'Carol'],
    'salary': [95000, 72000, 88000]
}, index=['E003', 'E001', 'E002'])
 
# Sort by index (alphabetically for string index)
print(df.sort_index())
#       name  salary
# E001   Bob   72000
# E002  Carol   88000
# E003  Alice   95000
 
# Sort descending
print(df.sort_index(ascending=False))

Part 3: .rank() — Assign Ranks

.rank() returns a Series of float ranks. The method parameter controls how ties are handled.

import pandas as pd
 
df = pd.DataFrame({
    'name':   ['Alice', 'Bob', 'Carol', 'David', 'Eve'],
    'salary': [95000, 72000, 88000, 88000, 105000]  # Carol and David tie
})
 
# Default: 'average' — tied values get the average of the ranks they would have
df['rank_avg'] = df['salary'].rank(ascending=False)
print(df[['name', 'salary', 'rank_avg']])
#     name  salary  rank_avg
# 0  Alice   95000       2.0
# 1    Bob   72000       5.0
# 2  Carol   88000       3.5   # tie: (3+4)/2
# 3  David   88000       3.5   # tie: (3+4)/2
# 4    Eve  105000       1.0
 
# 'dense' — no gaps in ranking even after ties
df['rank_dense'] = df['salary'].rank(method='dense', ascending=False)
print(df[['name', 'salary', 'rank_dense']])
# 1.0, 4.0, 2.0, 2.0, 1.0  (no 3.0 gap — 1, 2, 4 not 1, 2, 3, 4, 5 with skip)

Part 4: .nlargest() and .nsmallest() — Top/Bottom N

These are efficient shortcuts for "sort descending, take first N rows" — faster than sorting the entire DataFrame when you only need the extremes.

import pandas as pd
 
df = pd.DataFrame({
    'name':   ['Alice', 'Bob', 'Carol', 'David', 'Eve'],
    'salary': [95000, 72000, 88000, 68000, 105000]
})
 
# Top 3 salaries
print(df.nlargest(3, 'salary'))
#     name  salary
# 4    Eve  105000
# 0  Alice   95000
# 2  Carol   88000
 
# Bottom 2 salaries
print(df.nsmallest(2, 'salary'))
#     name  salary
# 3  David   68000
# 1    Bob   72000

Use .nlargest()/.nsmallest() instead of sort_values(...).head(N) — they skip the full sort, making them faster on large DataFrames.

Practice

<PracticeBlock prompt="Sort the employee DataFrame first by department ascending, then by salary descending within each department. Print the sorted result." initialCode={import pandas as pd\n\ndf = pd.DataFrame({\n 'name': ['Alice', 'Bob', 'Carol', 'David', 'Eve'],\n 'department': ['Engineering', 'Marketing', 'Engineering', 'Marketing', 'Engineering'],\n 'salary': [95000, 72000, 88000, 68000, 105000]\n})\n\n# Sort by department ascending, then salary descending\nsorted_df = \nprint(sorted_df)\n} hint="Use df.sort_values(['department', 'salary'], ascending=[True, False]) — pass a list for both column names and ascending flags." solution={import pandas as pd\n\ndf = pd.DataFrame({\n 'name': ['Alice', 'Bob', 'Carol', 'David', 'Eve'],\n 'department': ['Engineering', 'Marketing', 'Engineering', 'Marketing', 'Engineering'],\n 'salary': [95000, 72000, 88000, 68000, 105000]\n})\n\n# Sort by department ascending, then salary descending\nsorted_df = df.sort_values(['department', 'salary'], ascending=[True, False])\nprint(sorted_df)} />

<PracticeBlock prompt="Use .nlargest(3) to find the 3 highest-paid employees. Then use .nsmallest(2) to find the 2 lowest-paid employees." initialCode={import pandas as pd\n\ndf = pd.DataFrame({\n 'name': ['Alice', 'Bob', 'Carol', 'David', 'Eve'],\n 'department': ['Engineering', 'Marketing', 'Engineering', 'Marketing', 'Engineering'],\n 'salary': [95000, 72000, 88000, 68000, 105000]\n})\n\n# Top 3 highest-paid\nprint("Top 3 earners:")\n\n\n# Bottom 2 lowest-paid\nprint("Bottom 2 earners:")\n\n} hint="Use df.nlargest(3, 'salary') and df.nsmallest(2, 'salary'). Pass the column name as the second argument." solution={import pandas as pd\n\ndf = pd.DataFrame({\n 'name': ['Alice', 'Bob', 'Carol', 'David', 'Eve'],\n 'department': ['Engineering', 'Marketing', 'Engineering', 'Marketing', 'Engineering'],\n 'salary': [95000, 72000, 88000, 68000, 105000]\n})\n\n# Top 3 highest-paid\nprint("Top 3 earners:")\nprint(df.nlargest(3, 'salary'))\n\n# Bottom 2 lowest-paid\nprint("\nBottom 2 earners:")\nprint(df.nsmallest(2, 'salary'))} />

Key Takeaways

• .sort_values('col') — sort ascending by default; ascending=False for descending
• Multi-column sort: pass lists for both column names and ascending flags
• .sort_index() — sort rows by their index labels; useful after filtering or reindexing
• .rank(method='dense') — assign numeric ranks; control tie handling with method=
• .nlargest(n, 'col') / .nsmallest(n, 'col') — efficient top/bottom N; faster than sort + head

Common Mistakes to Avoid

• Forgetting to assign: df.sort_values('salary') returns a new DF — original unchanged unless you reassign
• ascending with multi-sort: for multi-column sort, ascending must be a list matching the columns list: ascending=[True, False]
• Using sort_values for index: use .sort_index() for sorting by the row index, not .sort_values(df.index)

← Previous Lesson | Back to Course Overview | Next Lesson: Basic Statistics →