Numpy - work with arrays and matrices

Introduction to NumPy: The Foundation for Scientific Computing in Python

NumPy is a fundamental Python library for scientific computing. It provides powerful support for multi-dimensional arrays and matrices, along with a wide range of functions to operate on them. NumPy is an indispensable tool for data analysts, machine learning specialists, and scientific researchers.

NumPy allows you to efficiently process large volumes of data and perform mathematical operations, replacing loops with compact and vectorized code. The library includes built-in functions for linear algebra and statistics.

Installation and Import of the NumPy Library

To start working with NumPy, you need to install it and import it into your project.

Installing NumPy

The library is installed using pip: [ pip install numpy ]

Importing NumPy

To use NumPy in your code, import it using the import statement:

import numpy as np

Handling Missing Values (NaN)

NumPy uses np.nan to represent missing values.

arr = np.array([1, 2, np.nan, 4, 5])
print(np.isnan(arr)) # [False False  True False False]

Special functions such as nanmean(), nanstd(), nansum() are used to handle NaN values, ignoring the missing values.

NumPy Performance vs. Pure Python

NumPy is written in C and runs significantly faster than pure Python, especially when performing operations on large arrays. This is due to efficient vectorization and optimization.

import time
import numpy as np

size = 100000

# Python list
lista = list(range(size))
listb = list(range(size))

starttime = time.time()
resultpython = [a + b for a, b in zip(lista, listb)]
pythontime = time.time() - starttime

# NumPy array
anp = np.arange(size)
bnp = np.arange(size)

starttime = time.time()
resultnumpy = anp + bnp
numpytime = time.time() - starttime

print(f"Python time: {pythontime:.4f} seconds")
print(f"NumPy time: {numpytime:.4f} seconds")

The difference in performance can reach tens and hundreds of times on large data volumes.

Real-World Use Cases of NumPy

NumPy is used in various fields:

• Science: processing experimental data
• Finance: analyzing time series
• Machine learning: preparing and normalizing data, modeling
• Computer graphics: working with images
• Education: visualizing mathematical functions and algorithms

Frequently Asked Questions (FAQ)

How does a NumPy array differ from a Python list?

A NumPy array is a data structure that stores numbers in the form of tables (tensors), has a fixed data type, and supports vectorized operations. A Python list is more flexible but less efficient for numerical computations.

How to create a zero matrix of size 3x3?

matrix = np.zeros((3, 3))

How to transpose a matrix?

transposedmatrix = matrix.T

How to multiply matrices?

result = matrix1 @ matrix2

How to find the inverse of a matrix?

inversedmatrix = np.linalg.inv(matrix)

Main Categories of NumPy Functions

NumPy provides a wide range of functions for various tasks:

• Array creation: NumPy offers many functions for creating and initializing arrays with various initial values (zeros, ones, random numbers, etc.).
• Mathematical operations: The library supports many mathematical operations on arrays, including arithmetic operations, statistical functions, and linear algebra functions.
• Linear algebra functions: NumPy provides functions for performing linear algebra operations such as matrix multiplication, eigenvalue computation, etc.
• Trigonometric and logarithmic functions: The library offers functions for working with trigonometric and logarithmic operations, making it easy to perform complex calculations.

Conclusion

NumPy is a powerful and flexible tool for working with data. Its capabilities make it essential for learning and using in scientific and analytical applications.