__getitem__ and __len__ in PyTorch - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ Predict Output

intermediate

Output of __getitem__ in a PyTorch Dataset

Consider this PyTorch Dataset class. What will be the output of dataset[2]?

PyTorch

import torch
from torch.utils.data import Dataset

class SimpleDataset(Dataset):
    def __init__(self):
        self.data = [10, 20, 30, 40, 50]
    def __len__(self):
        return len(self.data)
    def __getitem__(self, idx):
        return self.data[idx] * 2

dataset = SimpleDataset()
output = dataset[2]

A 30

B 60

C 40

D IndexError

💡 Hint

Remember __getitem__ returns data at index multiplied by 2.

The __getitem__ method returns self.data[idx] * 2. At index 2, self.data[2] is 30, so output is 60.

❓ Model Choice

intermediate

Choosing __len__ for a Custom Dataset

You have a dataset class with a list of 1000 images. Which implementation of __len__ correctly returns the dataset size?

A

def __len__(self):
    return 1000

B

def __len__(self):
    return self.images[0]

C

def __len__(self):
    return len(self.images)

D

def __len__(self):
    return self.images

💡 Hint

The length should reflect the number of items in the images list.

The __len__ method should return the number of items in the dataset, which is the length of the images list.

🔧 Debug

advanced

Debugging __getitem__ Index Error

This Dataset class raises an IndexError when accessing dataset[5]. Why?

PyTorch

class MyDataset:
    def __init__(self):
        self.data = [1, 2, 3, 4, 5]
    def __len__(self):
        return len(self.data)
    def __getitem__(self, idx):
        return self.data[idx + 1]

dataset = MyDataset()
output = dataset[5]

A Because data list is empty

B Because __len__ returns wrong length

C Because __getitem__ returns a list instead of a number

D Because __getitem__ accesses index 6 which is out of range

💡 Hint

Check how the index is used inside __getitem__.

dataset[5] calls __getitem__(5), which returns self.data[6]. Since data has length 5, index 6 is out of range causing IndexError.

❓ Hyperparameter

advanced

Effect of __len__ on DataLoader Batching

If a Dataset's __len__ returns 50 but actually contains 100 samples, what happens when using a DataLoader with batch size 10?

A DataLoader will load only 50 samples in 5 batches

B DataLoader will load all 100 samples in 10 batches

C DataLoader will raise a ValueError due to mismatch

D DataLoader will load 100 samples but batch size will be ignored

💡 Hint

DataLoader uses __len__ to know dataset size.

DataLoader relies on __len__ to know how many samples to load. If __len__ returns 50, it will load only 50 samples in 5 batches of 10.

🧠 Conceptual

expert

Why Implement Both __getitem__ and __len__ in PyTorch Dataset?

Why is it important to implement both __getitem__ and __len__ methods in a PyTorch Dataset class?

A Because DataLoader uses __len__ to know dataset size and __getitem__ to fetch samples by index

B Because __getitem__ initializes the dataset and __len__ trains the model

C Because __len__ loads data and __getitem__ shuffles it

D Because __len__ is used only for printing and __getitem__ is optional

💡 Hint

Think about how DataLoader interacts with Dataset.

DataLoader calls __len__ to know how many samples exist and __getitem__ to get each sample by index during training.