Day 18 of 30Real Python & Data
🔢NumPy — Arrays & Vectors
🎯 Today's objective
Work with fast numerical arrays — the math layer of AI.
📖 Key concepts
- np.array() makes an n-dimensional array
- Vectorised math (no loops): a * 2
- Shape, mean, sum, dot product
- Arrays power tensors in deep learning
💻 Try this code
import numpy as np
a = np.array([1, 2, 3, 4])
print(a * 10) # [10 20 30 40]
print(a.mean(), a.sum())
print(np.dot(a, a)) # 1+4+9+16 = 30👥 Learn it your way
🧒 Young Explorers
NumPy lets you do math on a whole list at once — super fast and tidy.
Your task
Make an array of 5 test scores; print the mean and the scores boosted by 5.
💼 Professionals
Vectorise instead of looping; understand shapes/broadcasting — the #1 ML bug source.
Your task
Create a 3×3 array, print its shape, transpose, and column means.
🤖 AI connection
Neural networks ARE arrays of numbers ("tensors") multiplied together. NumPy is the gateway drug to PyTorch/TensorFlow.
✅ Quick check
Why use NumPy over plain lists for math?
It’s faster and lets you do math on whole arrays at once (vectorised).
What is a.mean()?
The average of the array’s values.