NumPy Deep Dive

NumPy (Numerical Python) is the bedrock of scientific computing in Python. It provides the high-performance multidimensional array object and tools for working with these arrays.

1. Arrays vs. Lists 🏎️

Why do we use NumPy arrays instead of Python lists? Speed and Efficiency.

• Memory: NumPy arrays use less memory.
• Speed: NumPy operations are implemented in C, making them 50x faster.
• Convenience: Vectorized operations make code cleaner.

import numpy as np
import time

# Speed Test
size = 1000000
list1 = range(size)
list2 = range(size)
arr1 = np.arange(size)
arr2 = np.arange(size)

# Python List
start = time.time()
result_list = [(x+y) for x,y in zip(list1, list2)]
print(f"List time: {time.time() - start:.5f} seconds")

# NumPy Array
start = time.time()
result_arr = arr1 + arr2
print(f"NumPy time: {time.time() - start:.5f} seconds")

2. Creating Arrays 🏗️

# From a list
arr = np.array([1, 2, 3])

# Zeros and Ones
zeros = np.zeros((3, 3)) # 3x3 matrix of zeros
ones = np.ones((2, 4))   # 2x4 matrix of ones

# Ranges
range_arr = np.arange(0, 10, 2) # [0, 2, 4, 6, 8]
linspace_arr = np.linspace(0, 1, 5) # 5 numbers from 0 to 1

3. Dimensions, Shapes, and Reshaping 📐

Understanding dimensions is crucial for Deep Learning.

• ndim: Number of dimensions (axes).
• shape: Tuple of integers indicating the size of the array in each dimension.
• size: Total number of elements.

arr = np.array([[1, 2, 3], [4, 5, 6]])

print(arr.ndim)  # 2
print(arr.shape) # (2, 3)

# Reshaping
reshaped = arr.reshape(3, 2)
print(reshaped)
# [[1, 2],
#  [3, 4],
#  [5, 6]]

4. Broadcasting 📡

Broadcasting allows NumPy to perform operations on arrays of different shapes.

# Adding a scalar to a matrix
arr = np.array([1, 2, 3])
print(arr + 10) 
# Output: [11, 12, 13] (10 is "broadcast" to all elements)

# Adding a 1D array to a 2D array
matrix = np.ones((3, 3))
row = np.array([0, 1, 2])
print(matrix + row)
# Output:
# [[1., 2., 3.],
#  [1., 2., 3.],
#  [1., 2., 3.]]

5. Indexing and Slicing 🔪

arr = np.array([[1, 2, 3], [4, 5, 6], [7, 8, 9]])

# Standard Indexing
print(arr[0, 1]) # Row 0, Col 1 -> 2

# Slicing [start:end:step]
print(arr[0:2, 1:3]) 
# Rows 0 to 1, Cols 1 to 2
# [[2, 3],
#  [5, 6]]

# Boolean Indexing (Filtering)
print(arr[arr > 5]) 
# Output: [6, 7, 8, 9]

Interactive Challenge: Image Manipulation

Images are just NumPy arrays of pixels! Let's manipulate a "dummy" image.

Quiz

Key Takeaways

✅ NumPy is faster and more efficient than lists.
✅ Broadcasting simplifies math on different shapes.
✅ Reshaping is critical for preparing data for neural networks.

What's Next?

NumPy handles the numbers. Now let's handle structured data (tables) with Pandas.

Next Chapter: Mastering Pandas DataFrames.