NumPy - Numerical Computing

Welcome to Module 8, @name! Darwin here with Data Science - the science of data! 📊🐍

Safari Analogy: Data Science is like the Safari analytics system - you collect data about animals, analyze population trends, visualize results, and make decisions based on them! 🦁📈

NumPy - The Foundation of Data Science

NumPy (Numerical Python) is the core library for numerical computing. All other DS libraries (Pandas, Scikit-learn) are built on top of it!

Installation

1pip install numpy

Creating Arrays

1import numpy as np
2
3# From a list
4arr = np.array([1, 2, 3, 4, 5])
5print(arr)  # [1 2 3 4 5]
6
7# 2D array - matrix
8matrix = np.array([
9    [1, 2, 3],
10    [4, 5, 6],
11    [7, 8, 9]
12])
13
14# Special arrays
15zeros = np.zeros((3, 4))       # 3x4 matrix of zeros
16ones = np.ones((2, 3))         # 2x3 matrix of ones
17identity = np.eye(3)           # 3x3 identity matrix
18range_arr = np.arange(0, 10, 2) # [0, 2, 4, 6, 8]
19linspace = np.linspace(0, 1, 5) # [0, 0.25, 0.5, 0.75, 1]
20random_arr = np.random.rand(3, 3)  # Random 3x3 matrix

Array Attributes

1arr = np.array([[1, 2, 3], [4, 5, 6]])
2
3print(arr.shape)   # (2, 3) - dimensions
4print(arr.ndim)    # 2 - number of dimensions
5print(arr.size)    # 6 - number of elements
6print(arr.dtype)   # int64 - data type

Array Operations

1# Safari data - populations in different reserves
2populations = np.array([120, 450, 85, 200, 330])
3
4# Basic statistics
5print(np.mean(populations))   # 237.0 - mean
6print(np.median(populations)) # 200.0 - median
7print(np.std(populations))    # 135.4 - standard deviation
8print(np.min(populations))    # 85
9print(np.max(populations))    # 450
10print(np.sum(populations))    # 1185
11
12# Vectorized operations (broadcasting)
13growth = populations * 1.1  # 10% growth
14normalized = (populations - np.mean(populations)) / np.std(populations)

Indexing and Slicing

1arr = np.array([10, 20, 30, 40, 50])
2
3# Indexing
4print(arr[0])      # 10 - first element
5print(arr[-1])     # 50 - last element
6
7# Slicing
8print(arr[1:4])    # [20 30 40]
9print(arr[:3])     # [10 20 30]
10print(arr[::2])    # [10 30 50] - every other
11
12# Boolean indexing - filtering
13endangered = arr[arr < 30]  # [10 20]
14
15# 2D indexing
16matrix = np.array([[1,2,3], [4,5,6], [7,8,9]])
17print(matrix[0, 1])   # 2 - row 0, column 1
18print(matrix[:, 0])   # [1 4 7] - entire column 0
19print(matrix[1, :])   # [4 5 6] - entire row 1

Reshape and Shape Manipulation

1arr = np.arange(12)  # [0, 1, 2, ..., 11]
2
3# Reshape
4matrix = arr.reshape(3, 4)  # 3x4 matrix
5flat = matrix.flatten()     # Flatten to 1D
6transposed = matrix.T       # Transpose

Matrix Operations

1A = np.array([[1, 2], [3, 4]])
2B = np.array([[5, 6], [7, 8]])
3
4# Element-wise operations
5print(A + B)  # Addition
6print(A * B)  # Element-wise multiplication
7
8# Matrix multiplication
9print(A @ B)  # or np.dot(A, B)
10
11# Linear algebra
12print(np.linalg.det(A))       # Determinant
13print(np.linalg.inv(A))       # Inverse matrix
14eigenvalues, eigenvectors = np.linalg.eig(A)