Feature engineering basics in Pandas - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ Predict Output

intermediate

Output of creating a new feature by combining columns

What is the output DataFrame after running this code that creates a new feature by combining two columns?

Pandas

import pandas as pd

df = pd.DataFrame({
    'height_cm': [170, 180, 160],
    'weight_kg': [70, 80, 60]
})
df['bmi'] = df['weight_kg'] / ((df['height_cm'] / 100) ** 2)
print(df)

A {'height_cm': [170, 180, 160], 'weight_kg': [70, 80, 60], 'bmi': [24.22, 24.69, 23.44]}

B {'height_cm': [170, 180, 160], 'weight_kg': [70, 80, 60], 'bmi': [24.22, 24.69, 23.44, 25.0]}

C {'height_cm': [170, 180, 160], 'weight_kg': [70, 80, 60]}

D SyntaxError

💡 Hint

Calculate BMI as weight divided by height in meters squared.

The new column 'bmi' is calculated by dividing weight in kg by the square of height in meters. The result is a new column added to the DataFrame.

❓ data_output

intermediate

Number of unique categories after encoding

After applying one-hot encoding to the 'color' column, how many new columns are created?

Pandas

import pandas as pd

df = pd.DataFrame({'color': ['red', 'blue', 'green', 'blue', 'red']})
df_encoded = pd.get_dummies(df, columns=['color'])
print(df_encoded.columns.tolist())

A 2

B 4

C 5

D 3

💡 Hint

Count unique values in the 'color' column.

There are three unique colors: red, blue, and green. One-hot encoding creates one column per unique category.

🔧 Debug

advanced

Identify the error in feature scaling code

What error does this code raise when trying to scale a feature using Min-Max scaling?

Pandas

import pandas as pd

df = pd.DataFrame({'score': [10, 20, 30, 40, 50]})
min_val = df['score'].min()
max_val = df['score'].max()
df['scaled'] = (df['score'] - min_val) / (max_val - min_val)
print(df)

A TypeError

B No error, but scaling is incorrect

C KeyError

D ZeroDivisionError

💡 Hint

Check the formula for Min-Max scaling carefully.

The code divides by max_val instead of (max_val - min_val), so the scaling is incorrect but no error occurs.

❓ visualization

advanced

Interpret the histogram of a newly created feature

Given this code that creates a new feature 'age_group' and plots its histogram, what will the histogram show?

Pandas

import pandas as pd
import matplotlib.pyplot as plt

df = pd.DataFrame({'age': [15, 22, 37, 45, 52, 67, 70]})
df['age_group'] = pd.cut(df['age'], bins=[0, 18, 35, 60, 100], labels=['Child', 'Young Adult', 'Adult', 'Senior'])
df['age_group'].value_counts().plot(kind='bar')
plt.show()

A Bar chart with counts: Child=1, Young Adult=2, Adult=2, Senior=2

B Bar chart with counts: Child=0, Young Adult=2, Adult=3, Senior=2

C Bar chart with counts: Child=1, Young Adult=1, Adult=2, Senior=3

D Bar chart with counts: Child=1, Young Adult=1, Adult=3, Senior=2

💡 Hint

Check which ages fall into each bin range.

The bins split ages into four groups. Counting how many ages fall into each group gives the bar heights.

🚀 Application

expert

Choosing the best feature transformation for skewed data

You have a feature with a highly skewed distribution. Which transformation is most appropriate to reduce skewness before modeling?

A Apply a logarithmic transformation (log(x + 1))

B Apply a Min-Max scaling

C Apply one-hot encoding

D Apply standardization (z-score scaling)

💡 Hint

Think about transformations that reduce skewness, not just scale data.

Log transformation reduces right skewness by compressing large values. Scaling methods do not reduce skewness.