Data type optimization in Data Analysis Python - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ Predict Output

intermediate

Output of memory usage after optimization

Given a pandas DataFrame with integer columns, what is the output of the memory usage after converting the columns to int8?

Data Analysis Python

import pandas as pd
import numpy as np

df = pd.DataFrame({
    'A': np.random.randint(0, 100, 1000),
    'B': np.random.randint(0, 100, 1000)
})

before = df.memory_usage(deep=True).sum()
df['A'] = df['A'].astype('int8')
df['B'] = df['B'].astype('int8')
after = df.memory_usage(deep=True).sum()
print(after < before)

A False

B True

C TypeError

D MemoryError

💡 Hint

Think about how changing data types to smaller ones affects memory.

Converting integer columns from default int64 to int8 reduces memory usage, so after < before is True.

❓ data_output

intermediate

Resulting data types after optimization

What are the data types of the columns after applying the following optimization?

Data Analysis Python

import pandas as pd

df = pd.DataFrame({
    'col1': [1, 2, 3],
    'col2': [1000, 2000, 3000],
    'col3': ['a', 'b', 'c']
})

df['col1'] = df['col1'].astype('int8')
df['col2'] = df['col2'].astype('int16')
result = df.dtypes.to_dict()

A {'col1': 'int8', 'col2': 'int16', 'col3': 'object'}

B {'col1': 'int64', 'col2': 'int64', 'col3': 'object'}

C {'col1': 'float64', 'col2': 'float64', 'col3': 'object'}

D {'col1': 'int8', 'col2': 'int8', 'col3': 'object'}

💡 Hint

Check the explicit astype conversions.

The code converts col1 to int8 and col2 to int16. col3 remains object.

🔧 Debug

advanced

Identify the error in data type conversion

What error will this code raise when trying to convert the column to int8?

Data Analysis Python

import pandas as pd

df = pd.DataFrame({'values': [0, 127, 128, 255]})
df['values'] = df['values'].astype('int8')

A OverflowError

B No error, conversion succeeds

C ValueError

D TypeError

💡 Hint

Check the range of int8 and the values in the column.

Values like 128 and 255 exceed the range of int8 (-128 to 127), causing a ValueError.

🚀 Application

advanced

Choosing optimal data types for mixed data

You have a DataFrame with columns: age (0-120), income (0-1,000,000), and gender ('M' or 'F'). Which data types optimize memory without losing information?

A {'age': 'int8', 'income': 'int32', 'gender': 'category'}

B {'age': 'int16', 'income': 'int16', 'gender': 'object'}

C {'age': 'int8', 'income': 'float64', 'gender': 'object'}

D {'age': 'int32', 'income': 'int64', 'gender': 'category'}

💡 Hint

Consider the value ranges and best types for categorical data.

age fits in int8 (0-127 covers 0-120), income fits in int32 (max ~2 billion), and gender is best as category for two values.

🧠 Conceptual

expert

Impact of data type optimization on performance

Which statement best describes the impact of data type optimization on data processing performance?

A Data type optimization has no effect on performance, only on memory usage.

B Smaller data types always speed up computations because less memory is used.

C Using larger data types improves performance by avoiding overflow errors.

D Optimizing data types reduces memory usage but can sometimes slow down computations due to type conversions.

💡 Hint

Think about trade-offs between memory and CPU operations.

While smaller data types reduce memory, some operations may require upcasting or conversions, which can slow down processing.