NumPy Tutorial in Python

NumPy (Numerical Python) 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. NumPy is widely used in scientific computing, data analysis, and machine learning.

1. Installation

To install NumPy, you can use the following command:

pip install numpy

2. Importing NumPy

import numpy as np

This imports NumPy and gives it the alias np, which is the common convention.

3. Creating NumPy Arrays

NumPy arrays are the core data structure in NumPy. They are more efficient than Python lists and provide vectorized operations.

a. Creating Arrays from Lists

You can create a NumPy array from a Python list using np.array().

import numpy as np# Creating a 1D arrayarr = np.array([1, 2, 3, 4])print(arr)  # Output: [1 2 3 4]# Creating a 2D arrayarr_2d = np.array([[1, 2, 3], [4, 5, 6]])print(arr_2d)# Output:# [[1 2 3]#  [4 5 6]]

b. Creating Arrays with Zeros, Ones, and Random Values

• np.zeros(): Creates an array of zeros.

• np.ones(): Creates an array of ones.

• np.random.rand(): Creates an array of random numbers between 0 and 1.

# Array of zeroszeros = np.zeros((3, 3))  # 3x3 arrayprint(zeros)# Array of onesones = np.ones((2, 4))  # 2x4 arrayprint(ones)# Array with random valuesrandom_vals = np.random.rand(3, 3)  # 3x3 array with random valuesprint(random_vals)

4. Array Properties

You can access various properties of NumPy arrays:

arr = np.array([1, 2, 3, 4, 5])# Shape of the array (dimensions)print(arr.shape)  # Output: (5,)# Number of dimensionsprint(arr.ndim)   # Output: 1# Size of the array (total number of elements)print(arr.size)   # Output: 5# Data type of the arrayprint(arr.dtype)  # Output: int64 (or int32, depending on your system)

For 2D arrays:

arr_2d = np.array([[1, 2, 3], [4, 5, 6]])# Shape of the 2D arrayprint(arr_2d.shape)  # Output: (2, 3)

5. Array Indexing and Slicing

NumPy arrays can be indexed and sliced just like Python lists, but with more powerful options for multi-dimensional arrays.

a. Indexing

arr = np.array([10, 20, 30, 40, 50])# Accessing individual elementsprint(arr[0])  # Output: 10print(arr[3])  # Output: 40

b. Slicing

# Slicing from index 1 to 3print(arr[1:4])  # Output: [20 30 40]

c. Slicing for 2D arrays

arr_2d = np.array([[1, 2, 3], [4, 5, 6]])# Accessing a specific rowprint(arr_2d[1])  # Output: [4 5 6]# Slicing the 2D arrayprint(arr_2d[:, 1:3])  # Output: [[2 3] [5 6]]

6. Array Operations

NumPy arrays support a wide range of mathematical operations, and these operations are performed element-wise.

a. Arithmetic Operations

arr1 = np.array([1, 2, 3])arr2 = np.array([4, 5, 6])# Additionprint(arr1 + arr2)  # Output: [5 7 9]# Subtractionprint(arr1 - arr2)  # Output: [-3 -3 -3]# Multiplicationprint(arr1 * arr2)  # Output: [4 10 18]# Divisionprint(arr1 / arr2)  # Output: [0.25 0.4 0.5]# Exponentiationprint(arr1 ** 2)  # Output: [1 4 9]

b. Universal Functions (ufuncs)

NumPy provides universal functions (ufuncs) to perform element-wise operations on arrays.

arr = np.array([1, 4, 9])# Square rootprint(np.sqrt(arr))  # Output: [1. 2. 3.]# Exponentialprint(np.exp(arr))  # Output: [2.71828183e+00 5.45981500e+01 8.10308393e+03]# Logarithmprint(np.log(arr))  # Output: [0.         1.38629436 2.19722458]

7. Broadcasting

Broadcasting is a powerful feature in NumPy that allows you to perform operations on arrays of different shapes. When operating on arrays, NumPy compares the shapes of the arrays element-wise and "broadcasts" the smaller array across the larger one.

Example:

arr1 = np.array([1, 2, 3])arr2 = np.array([10])# Broadcasting: arr2 is broadcasted to match the shape of arr1print(arr1 + arr2)  # Output: [11 12 13]

8. Reshaping Arrays

You can change the shape of a NumPy array using the reshape() method.

arr = np.array([1, 2, 3, 4, 5, 6])# Reshaping to 2x3 matrixarr_reshaped = arr.reshape((2, 3))print(arr_reshaped)# Output:# [[1 2 3]#  [4 5 6]]

9. Statistical Operations

NumPy provides a variety of functions to calculate statistics.

arr = np.array([1, 2, 3, 4, 5])# Meanprint(np.mean(arr))  # Output: 3.0# Medianprint(np.median(arr))  # Output: 3.0# Standard Deviationprint(np.std(arr))  # Output: 1.4142135623730951# Sumprint(np.sum(arr))  # Output: 15

10. Linear Algebra in NumPy

NumPy also provides functions for linear algebra operations.

# Dot productarr1 = np.array([1, 2])arr2 = np.array([3, 4])print(np.dot(arr1, arr2))  # Output: 11 (1*3 + 2*4)# Matrix multiplicationarr_2d_1 = np.array([[1, 2], [3, 4]])arr_2d_2 = np.array([[5, 6], [7, 8]])print(np.matmul(arr_2d_1, arr_2d_2))# Output:# [[19 22]#  [43 50]]

11. Saving and Loading Arrays

You can save and load NumPy arrays to/from files using the np.save() and np.load() functions.

Saving an array:

arr = np.array([1, 2, 3, 4])np.save('my_array.npy', arr)

Loading an array:

loaded_arr = np.load('my_array.npy')print(loaded_arr)  # Output: [1 2 3 4]

12. Summary of Key Functions in NumPy

Conclusion

NumPy is a powerful and versatile library for numerical computing in Python. Whether you're working with simple arrays, performing complex mathematical operations, or handling multi-dimensional matrices, NumPy provides all the tools you need for efficient computation.