In-place operations for memory efficiency in NumPy - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ Predict Output

intermediate

Output of in-place addition with NumPy arrays

What is the output of the following code snippet?

NumPy

import numpy as np
arr = np.array([1, 2, 3])
arr += 5
print(arr)

A [1 2 3]

B [1 2 3 5]

C [5 7 8]

D [6 7 8]

💡 Hint

Remember that '+=' modifies the array in place by adding 5 to each element.

The '+=' operator adds 5 to each element of the array in place, changing the original array to [6, 7, 8].

❓ data_output

intermediate

Memory usage difference between in-place and out-of-place operations

Given the code below, which option correctly describes the memory usage behavior?

NumPy

import numpy as np
arr = np.ones((1000, 1000))
arr2 = arr * 2  # Out-of-place
arr *= 2       # In-place

A Neither arr2 nor arr *= 2 change memory usage

B arr2 uses new memory; arr *= 2 modifies arr without extra memory

C arr2 modifies arr in place; arr *= 2 creates a new array

D Both arr2 and arr *= 2 create new arrays in memory

💡 Hint

Multiplying with '*' creates a new array; '*=' modifies the existing array.

arr2 = arr * 2 creates a new array, using extra memory. arr *= 2 modifies arr in place, saving memory.

🔧 Debug

advanced

Identify the error in in-place operation

What error will this code raise and why?

NumPy

import numpy as np
arr = np.array([1, 2, 3], dtype=int)
arr += 2.5

A TypeError: Cannot cast float to int with casting rule 'same_kind'

B ValueError: operands could not be broadcast together

C No error, arr becomes [3.5, 4.5, 5.5]

D SyntaxError: invalid syntax

💡 Hint

Check if the operation tries to store floats into an integer array.

NumPy does not allow in-place addition of floats to an integer array without explicit casting, causing a TypeError.

🚀 Application

advanced

Using in-place operations to optimize a data transformation

You have a large NumPy array 'data' and want to normalize it by subtracting the mean and dividing by the standard deviation. Which code snippet uses in-place operations correctly to save memory?

A

mean = data.mean()
data -= mean
std = data.std()
data /= std

B

mean = data.mean()
data = data - mean
std = data.std()
data = data / std

C

mean = data.mean()
data -= mean
std = data.std()
data = data / std

D

mean = data.mean()
data = data - mean
std = data.std()
data /= std

💡 Hint

Look for use of '-=' and '/=' operators to modify 'data' in place.

Using '-=' and '/=' modifies the original array without creating new arrays, saving memory.

🧠 Conceptual

expert

Why prefer in-place operations in large data processing?

Which is the best reason to prefer in-place operations when working with large NumPy arrays?

A They automatically parallelize computations across CPU cores

B They always run faster because they use GPU acceleration

C They reduce memory usage by avoiding creation of temporary arrays

D They prevent any changes to the original data

💡 Hint

Think about memory and temporary data when processing large arrays.

In-place operations save memory by modifying data directly, avoiding extra copies.