NumPy

NumPy (Numerical Python)

NumPy is a powerful library for numerical computing in Python. It provides support for large, multi-dimensional arrays and matrices, along with a collection of mathematical functions to operate on these arrays. It is widely used for data analysis, machine learning, and scientific computing.

________________________________________

Key Features of NumPy:

1. Multidimensional Arrays (ndarray):

o NumPy introduces the ndarray, a fast, flexible, and efficient multi-dimensional array object. Arrays are more compact and faster than regular Python lists.

2. Mathematical Functions:

o NumPy provides a wide range of mathematical operations, such as linear algebra, statistics, Fourier transforms, and random number generation.

3. Broadcasting:

o Broadcasting allows NumPy to perform element-wise operations on arrays of different shapes. It is a powerful feature that makes array operations efficient without needing explicit looping.

4. Array Operations:

o NumPy supports operations like element-wise addition, subtraction, multiplication, division, and more, which can be performed on arrays without using explicit loops.

5. Linear Algebra:

o NumPy provides functions for matrix multiplication, eigenvalue decomposition, determinants, and other linear algebra operations.

6. Random Module:

o NumPy has a random module for generating random numbers and performing statistical sampling.

7. Fast and Efficient:

o NumPy is highly optimized for performance and is implemented in C, making it much faster than pure Python code for large data sets.

8. Compatibility with Other Libraries:

o NumPy arrays serve as the backbone for other data science and machine learning libraries like Pandas, Matplotlib, TensorFlow, and SciPy.

________________________________________

Basic NumPy Operations

1. Importing NumPy:

import numpy as np

2. Creating Arrays:

• From a List:

• arr = np.array([1, 2, 3, 4])

• print(arr)

• Creating a 2D Array:

• arr2d = np.array([[1, 2, 3], [4, 5, 6]])

• print(arr2d)

• Using Built-in Functions:

• zeros = np.zeros((3, 3))  # 3x3 array of zeros

• ones = np.ones((2, 4))    # 2x4 array of ones

• eye = np.eye(3)           # Identity matrix of size 3x3

3. Array Operations:

• Addition:

• arr1 = np.array([1, 2, 3])

• arr2 = np.array([4, 5, 6])

• result = arr1 + arr2

• print(result)  # Output: [5 7 9]

• Element-wise Operations:

• result = arr1 * arr2

• print(result)  # Output: [4 10 18]

• Dot Product (Matrix Multiplication):

• arr3 = np.array([[1, 2], [3, 4]])

• arr4 = np.array([[5, 6], [7, 8]])

• result = np.dot(arr3, arr4)

• print(result)

4. Array Indexing and Slicing:

• Indexing:

• arr = np.array([1, 2, 3, 4, 5])

• print(arr[2])  # Output: 3

• Slicing:

• print(arr[1:4])  # Output: [2 3 4]

• 2D Array Indexing:

• arr2d = np.array([[1, 2, 3], [4, 5, 6]])

• print(arr2d[1, 2])  # Output: 6

5. Reshaping Arrays:

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

reshaped = arr.reshape(2, 3)  # Reshaping to 2x3 array

print(reshaped)

6. Broadcasting Example:

Broadcasting allows operations on arrays of different shapes. For example:

arr1 = np.array([1, 2, 3])

arr2 = np.array([10])

result = arr1 + arr2  # arr2 is broadcasted to match the shape of arr1

print(result)  # Output: [11 12 13]

________________________________________

Mathematical Operations in NumPy:

1. Sum:

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

3. sum_arr = np.sum(arr)

4. print(sum_arr)  # Output: 15

5. Mean:

6. mean = np.mean(arr)

7. print(mean)  # Output: 3.0

8. Standard Deviation:

9. std_dev = np.std(arr)

10. print(std_dev)  # Output: 1.4142135623730951

11. Minimum and Maximum:

12. min_val = np.min(arr)

13. max_val = np.max(arr)

14. print(min_val, max_val)  # Output: 1 5

15. Matrix Operations (determinant, inverse):

16. mat = np.array([[1, 2], [3, 4]])

17. det = np.linalg.det(mat)

18. inv = np.linalg.inv(mat)

19. print(det, inv)

________________________________________

NumPy Random Module:

NumPy provides a random module to generate random numbers and perform statistical sampling.

1. Random Integer:

2. random_int = np.random.randint(1, 10)  # Random integer between 1 and 9

3. print(random_int)

4. Random Float:

5. random_float = np.random.rand(3, 2)  # 3x2 matrix of random floats between 0 and 1

6. print(random_float)

7. Random Normal Distribution:

8. normal_dist = np.random.randn(3, 3)  # 3x3 matrix from a normal distribution

9. print(normal_dist)

10. Random Sample from a Given Array:

11. sample = np.random.choice([10, 20, 30, 40, 50], size=3)

12. print(sample)

________________________________________

Advanced NumPy Operations:

1. Linear Algebra (Eigenvalues and Eigenvectors):

2. mat = np.array([[1, 2], [3, 4]])

3. eigenvalues, eigenvectors = np.linalg.eig(mat)

4. print("Eigenvalues:", eigenvalues)

5. print("Eigenvectors:", eigenvectors)

6. Singular Value Decomposition (SVD):

7. U, S, Vt = np.linalg.svd(mat)

8. print(U, S, Vt)

9. Solving Linear Systems:

10. A = np.array([[3, 1], [1, 2]])

11. B = np.array([9, 8])

12. X = np.linalg.solve(A, B)  # Solves the equation Ax = B

13. print(X)

________________________________________

Performance Considerations:

1. Vectorization:

o Instead of using loops, NumPy allows you to perform operations on entire arrays at once. This leads to significant performance improvements compared to native Python loops.

2. Memory Efficiency:

o NumPy arrays are more memory-efficient than Python lists due to their contiguous memory storage.

3. Interfacing with C/C++:

o NumPy allows you to integrate with low-level languages (C, C++) to achieve performance gains when necessary, particularly for large datasets or complex mathematical computations.