Sorting MultiIndex in Pandas - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ Predict Output

intermediate

Output of sorting a MultiIndex DataFrame by level

What is the output of the following code snippet?

Pandas

import pandas as pd

index = pd.MultiIndex.from_tuples([('b', 2), ('a', 1), ('b', 1), ('a', 2)])
df = pd.DataFrame({'value': [4, 3, 2, 1]}, index=index)
sorted_df = df.sort_index(level=0)
print(sorted_df)

A

       value
(b, 1)      2
(b, 2)      4
(a, 1)      3
(a, 2)      1

B

       value
(a, 1)      3
(a, 2)      1
(b, 1)      2
(b, 2)      4

C

       value
(a, 2)      1
(a, 1)      3
(b, 2)      4
(b, 1)      2

D

       value
(b, 2)      4
(b, 1)      2
(a, 2)      1
(a, 1)      3

💡 Hint

Sorting by level=0 sorts by the first index level alphabetically.

The DataFrame is sorted by the first level of the MultiIndex ('a' before 'b'). Within each group, the order of the second level remains as in the original data.

❓ data_output

intermediate

Number of rows after sorting MultiIndex with ascending=False

Given the following DataFrame, how many rows does the sorted DataFrame have after sorting by both levels descending?

Pandas

import pandas as pd

index = pd.MultiIndex.from_tuples([('x', 3), ('y', 1), ('x', 1), ('y', 2)])
df = pd.DataFrame({'score': [10, 20, 30, 40]}, index=index)
sorted_df = df.sort_index(level=[0,1], ascending=False)
print(len(sorted_df))

A 1

B 2

C 3

D 4

💡 Hint

Sorting does not remove rows, it only reorders them.

Sorting by both levels descending keeps all rows but changes their order. The number of rows remains 4.

🔧 Debug

advanced

Identify the error in sorting MultiIndex with invalid level

What error does the following code raise?

Pandas

import pandas as pd

index = pd.MultiIndex.from_tuples([('a', 1), ('b', 2)])
df = pd.DataFrame({'val': [5, 6]}, index=index)
sorted_df = df.sort_index(level=2)
print(sorted_df)

A IndexError: Too many levels: Index has only 2 levels, not 3

B No error, prints sorted DataFrame

C TypeError: level must be int or str

D KeyError: 2

💡 Hint

Check the number of levels in the MultiIndex before sorting by level.

The MultiIndex has only 2 levels (0 and 1). Trying to sort by level=2 causes an IndexError about too many levels.

🚀 Application

advanced

Sorting MultiIndex DataFrame by multiple levels with mixed ascending order

You have a MultiIndex DataFrame with two levels: 'city' and 'year'. You want to sort it so that cities are in ascending order but years are in descending order. Which code snippet achieves this?

Pandas

import pandas as pd

index = pd.MultiIndex.from_tuples([('NY', 2020), ('LA', 2019), ('NY', 2019), ('LA', 2020)])
df = pd.DataFrame({'population': [8, 4, 7, 5]}, index=index)

A df.sort_index(level=['city', 'year'], ascending=[True, False])

B df.sort_index(level=['city', 'year'], ascending=[False, True])

C df.sort_index(level=[0, 1], ascending=[False, False])

D df.sort_index(level=[1, 0], ascending=[True, False])

💡 Hint

The ascending parameter accepts a list matching the levels order.

To sort 'city' ascending and 'year' descending, pass ascending=[True, False] with level order ['city', 'year'].

🧠 Conceptual

expert

Effect of sort_index on MultiIndex with duplicates

Consider a MultiIndex DataFrame with duplicate index entries. What is the effect of calling sort_index() on this DataFrame?

A It raises an error because duplicates are not allowed in MultiIndex.

B It removes duplicate index entries and sorts the DataFrame.

C It sorts the DataFrame by the MultiIndex levels but keeps duplicate index entries intact.

D It sorts the DataFrame and resets the index to default integers.

💡 Hint

Think about whether sorting changes data or just order.

sort_index() only reorders rows based on index levels. It does not remove duplicates or reset the index.