NumPy Essentials for AI

Why NumPy

Arrays are how AI represents data — vectors, matrices, batches of images. NumPy makes operating on them fast and concise.

import numpy as np
a = np.array([[1, 2], [3, 4]])
a.mean(axis=0)         # column means
a @ a                  # matrix multiply
(a > 2)                # boolean mask

Concepts that pay off

• Vectorization — operate on whole arrays, avoid Python loops.
• Broadcasting — combine arrays of different shapes.
• Axes — many functions take axis= to control direction.

These ideas carry straight over to PyTorch and TensorFlow tensors.