Batch Iterator for Dataset

Problem

Implement a batch iterator for a dataset. Given a 2D NumPy array X and an integer batch_size, yield successive batches of rows from X. If the dataset size is not evenly divisible by the batch size, the last batch should contain the remaining rows.

Solution

import numpy as np

def batch_iterator(X, batch_size=64):
    n_samples = X.shape[0]
    for i in range(0, n_samples, batch_size):
        yield X[i:i + batch_size]

Explanation

1. Get the total number of samples from the first dimension of the array.
2. Use range with step batch_size to iterate over starting indices.
3. Slice the array from index i to i + batch_size. Python slicing naturally handles the last batch if it is smaller than batch_size.

Complexity

• Time: O(n) across all yields, where n is the number of rows
• Space: O(batch_size) per yielded batch (views, not copies, in NumPy)