nunique() for cardinality in Data Analysis Python - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ Predict Output

intermediate

Output of nunique() on a DataFrame column

What is the output of the following code snippet?

Data Analysis Python

import pandas as pd

data = {'fruit': ['apple', 'banana', 'apple', 'orange', 'banana', 'banana']}
df = pd.DataFrame(data)

result = df['fruit'].nunique()
print(result)

A 3

B 6

C 4

D 2

💡 Hint

Count how many unique fruit names appear in the list.

The column 'fruit' contains three unique values: 'apple', 'banana', and 'orange'. The nunique() function counts these unique values, so the output is 3.

❓ data_output

intermediate

Number of unique values per column

Given the DataFrame below, what is the output of df.nunique()?

Data Analysis Python

import pandas as pd

data = {'A': [1, 2, 2, 3], 'B': ['x', 'y', 'x', 'z'], 'C': [True, True, False, False]}
df = pd.DataFrame(data)

result = df.nunique()
print(result)

A

A    3
B    3
C    2
dtype: int64

B

A    4
B    3
C    2
dtype: int64

C

A    3
B    2
C    2
dtype: int64

D

A    3
B    3
C    3
dtype: int64

💡 Hint

Count unique values in each column separately.

Column A has values [1,2,2,3] with 3 unique values (1,2,3). Column B has ['x','y','x','z'] with 3 unique values (x,y,z). Column C has [True,True,False,False] with 2 unique values (True, False).

🔧 Debug

advanced

Identify the error in using nunique()

What error will this code produce?

Data Analysis Python

import pandas as pd

data = {'numbers': [1, 2, 3, 4]}
df = pd.DataFrame(data)

result = df.nunique(axis=2)
print(result)

A TypeError: axis must be 0 or 1

B ValueError: axis 2 is out of bounds for array of dimension 2

C AttributeError: 'DataFrame' object has no attribute 'nunique'

D No error, prints the number of unique values per row

💡 Hint

Check the valid axis values for DataFrame methods.

DataFrames have only two axes: 0 (rows) and 1 (columns). Using axis=2 is invalid and raises a ValueError about axis being out of bounds.

🚀 Application

advanced

Using nunique() to find columns with high cardinality

You have a DataFrame with many columns. Which code snippet correctly finds columns with more than 10 unique values?

Data Analysis Python

import pandas as pd

# Assume df is predefined

# Find columns with more than 10 unique values
high_card_cols = ?

A high_card_cols = df.columns[df.nunique() < 10].tolist()

B high_card_cols = df.columns[df.nunique(axis=1) > 10].tolist()

C high_card_cols = df.columns[df.nunique() > 10].tolist()

D high_card_cols = df.columns[df.nunique(axis=0) < 10].tolist()

💡 Hint

nunique() by default counts unique values per column.

df.nunique() returns a Series with unique counts per column. Filtering with > 10 selects columns with more than 10 unique values. axis=1 counts per row, which is not desired here.

🧠 Conceptual

expert

Understanding nunique() behavior with missing values

Consider a DataFrame column with values: [1, 2, 2, None, 3, None]. What is the output of df['col'].nunique(dropna=False)?

Data Analysis Python

import pandas as pd

data = {'col': [1, 2, 2, None, 3, None]}
df = pd.DataFrame(data)

result = df['col'].nunique(dropna=False)
print(result)

A 3

B 2

C 5

D 4

💡 Hint

dropna=False counts NaN as a unique value.

The unique values are 1, 2, 3, and None (NaN). With dropna=False, NaN is counted as a unique value, so total unique count is 4.