Nuffing.com

Python is a popular high-level programming language used mainly for data science, automation, web development, and Artificial Intelligence. It is a general-purpose programming language supporting functional programming, object-oriented programming, and procedural programming. Over the years, Python is known to be the best programming language for data science, and it is commonly used by big tech companies for data science tasks.

In this tutorial, you will learn why Python is so popular for data science and why it will stay popular in the future.

What Can Python Be Used For?

As said earlier, Python is a general-purpose programming language, which means that it can be used for almost everything.

One common application of Python in web development is where Django or Flask is used as the backend for a website. For example, Instagram’s backend runs on Django, and it’s one of the largest deployments of Django.

You can also use Python for game development with Pygame, Kivy, Arcade, etcetera; though it’s rarely used. Mobile app development is not left out, Python offers many app development libraries such as Kivy and KivyMD which you can use for developing multiplatform apps; and many other libraries like Tkinter, PyQt, etc.

The main talk of this tutorial is the application of Python in Data Science. Python has been proven to be the best programming language for Data Science and you will know why in this tutorial.

What Is Data Science?

According to Oracle, data science combines multiple fields, including statistics, scientific methods, artificial intelligence (AI), and data analysis, to extract value from data. It encompasses preparing data for analysis, including cleansing, aggregating, and manipulating the data to perform advanced data analysis.

Data science is applicable in different industries, and it’s helping to solve problems and discover more about the universe. In the health industry, data science helps doctors to make use of past data in making decisions, for example, diagnosis, or the right treatment for a disease. The education sector is not left out, you can now predict students dropping out of school, all thanks to data science.

Python Has a Simple Syntax

What else can make programming a lot easier than having an intuitive syntax? In Python, you need just one line to run your first program: simply type print(“Hello World!”) and run – it’s that easy.

Python has a very simple syntax, and it makes programming a lot easier and faster. There is no need for curly braces when writing functions, no semicolon is your enemy, and you don’t even need to import libraries before you write basic code.

This is one advantage Python has over other programming languages. You have fewer tendencies to make errors, and you can easily notice bugs.

Data Science is one complex field you can’t do without needing any help. Python offers all the help you need through its wide community. Whenever you get stuck, just browse it and your answer is waiting for you. Stack Overflow is a very popular website where questions and answers are posted to programming problems.

If your problem is new, which is rare, you can ask questions and people would be willing to provide answers.

Python Offers All the Libraries

You badly need water, and you have just two cups on the table. One is a quarter filled with water while the other one is almost full. Would you carry the cup with much water or the other one, though they both have water? You’d want to carry the cup containing a lot of water because you really need water. This is relatable to Python, it offers all the libraries you’d ever need for data science, you would definitely not want to use another programming language with only a few libraries available.

You will have a great experience working with these libraries because they are really easy to use. If you need to install any library, search for the library name at PyPI.org and follow the instructions towards the end of this article to install the library.

Related: Data Science Libraries for Python Every Data Scientist Should Use

Numerical Python – NumPy

NumPy is one of the most commonly used data science libraries. It allows you to work with numeric and scientific tasks in Python. Data is represented using arrays or what you may refer to as lists, which can be in any dimension: 1-dimensional (1D) array, 2-dimensional (2D) array, 3-dimensional (3D) array, and so on.

Pandas

Pandas is also a popular data science library used in data preparation, data processing, data visualization. With Pandas, you can import data in different formats such as CSV (comma-separated values) or TSV (Tab-separated values). Pandas works like Matplotlib because it allows you to make different types of plots. Another cool feature Pandas offers is that it allows you to read SQL queries. So, if you have connected to your database, and you want to write and run SQL queries in Python, Pandas is a great choice.

Matplotlib and Seaborn

Matplotlib is another awesome library Python offers. It has been developed on top of MatLab – a programming language used mainly for scientific and visualization purposes. Matplotlib allows you to plot different kinds of graphs with just a few lines of code.

You can plot graphs to visualize any data, helping you to gain insights from your data, or giving you a better representation of the data. Other libraries like Pandas, Seaborn, and OpenCV also use Matplotlib for plotting sophisticated graphs.

Seaborn (not Seaborne) is just like Matplotlib, just that you have more options – to give different parts of your graphs different colors, or hues. You can plot nice graphs and customize the look to make the data representation better.

Open Computer Vision – OpenCV

Perhaps you want to build an Optical Character Recognition (OCR) system, document scanner, image filter, motion sensor, security system, or anything else related to computer vision, you should try OpenCV. This amazing and free library offered by Python allows you to build computer vision systems over just a few lines of code. You can work with images, videos, or even your webcam feed and deploy.

Scikit-learn – Sklearn

Scikit-learn is the most popular library used specifically for machine learning tasks in data science. Sklearn offers all the utilities you need to make use of your data and build machine learning models in just a few lines of code.

There are various machine learning tasks like linear regression (simple and multiple), logistic regression, k-nearest neighbors, naive bayes, support vector regression, random forest regression, polynomial regression, including classification and clustering tasks.

Though Python is simple because of its syntax; there are tools that have been specifically designed with data science in mind. Jupyter notebook is the first tool, it is a development environment built by Anaconda, to write Python code for data science tasks. You can write and instantly run codes in cells, group them, or even include documentation, as provided by its markdown capability.

A popular alternative is Google Colaboratory, also known as Google Colab. They are similar and used for the same purpose but Google Colab has more advantages because of its cloud support. You have access to more space, not having to worry about your computer storage getting full. You can also share your notebooks, log in on any device and access it, or even save your notebook to GitHub.

How to Install Any Data Science Library in Python

Given you already have Python installed on your computer, this step-by-step section will guide you through how to install any data science library on your Windows computer. NumPy will be installed in this case, follow the steps below:

1. Press Start and type cmd. Right-click the result and choose Run as administrator.

1. You need PIP to install Python libraries from PyPi. If you already have, feel free to skip this step; if not, please read how to install PIP on your computer.
2. Type pip install numpy and press Enter to run. This process will install NumPy on your computer and you can now import and use NumPy on your computer. This process should look similar to the screenshot shown below, ignore the warning and blank spaces. (If you use Linux or macOS, simply open a terminal and enter the pip install command).

It’s Time to Use Python for Data Science

Among other programming languages like R, C++, and Java; Python stands to be the best for data science. This tutorial has guided you through why Python is so popular for data science. You now know what Python offers and why big companies like Google, Meta, NASA, Tesla, etcetera use Python.

Did this tutorial succeed in convincing you that Python will remain the best programming language for data science? If yes, go on and build nice data science projects; help make life easier.

About The Author

User account management is one of the many challenges of Linux system administrators. Some of the responsibilities of a system administrator are enabling/disabling user accounts, preserving the home directory, setting user permissions, assigning groups/shells to users, and managing passwords.

Effective control of user accounts is only possible after familiarity with the basics of Linux account management. Hence, this article is a stepping stone towards securing user accounts. It demonstrates how to create, delete and modify user accounts and manage predefined settings or files to build the most suitable and secure environment for Linux users.

How to Add User Accounts in Linux

As a word of precaution, any user who utilizes your Linux machine must have a separate user account. A user account allows you to separate your files in a safe space with the ability to tailor your home directories, path, environment variables, etc.

Before beginning with the creation of a new user, list the available user accounts with the help of the cut command as follows:

cut -d: -f1 /etc/passwd

The simplest way of creating a new user account in Linux is with the help of useradd. This utility offers various parameters to specify additional information while adding a new user. Some of the options are:

• -c: Adds description/comment to a user account.

  useradd -c "John Wise" john

• -d: Sets the home directory for the specified user. By default, the useradd command sets it to the username (/home/john), but you can replace it with the directory of your choice as follows:

  useradd -d /mnt/home/john

• -g: Allows you to set the primary group of a user. The user will be added to a group by default if you don’t add one during the creation process.
• -G: Adds the user to multiple groups.

  useradd -G juice,apple,linux,tech john

• -o: Creates a new user account using the UID of an existing user.
• -p: Used to add an encrypted password to the account. You can also add your password later using the passwd command.

  passwd john

For instance, here’s how you can use the useradd command and some of the above parameters to add a new user:

useradd -g tech -G apple,linux -s /bin/zsh -c "James Adem" adem

In the user creation process, the aforementioned command performs several actions:

Modify Default User Settings

The useradd command reads the default values from /etc/login.defs, /etc/useradd, and /etc/default/useradd. You can open the files in your favorite text editor in Linux, make and save the appropriate changes before using the command.

You can view some of the settings available inside login.defs using the following command:

cat /etc/login.defs | grep 'PASS\|UID\|GID'

The uncommented lines are keywords with values. For instance, the PASS_MAX_DAYS keyword sets a maximum of 9999 days for password expiration. Similarly, the PASS_MIN_LEN keyword requires the password length to be at least five characters. Lastly, the UID and GID keywords allow customization of the user and group ID ranges for any new user account.

You can also view/modify the default settings present inside the files by using the useradd command with the -D flag.

Note that you don’t use the -D flag to create a new account. Instead, it only allows you to change the default settings. Also, it supports changes for only a few parameters that the useradd command uses to create an account.

For instance, the following command changes the default shell to /bin/sh and the home directory to /home/new:

useradd -D -b /home/new -s /bin/sh

Modify User Groups on Linux

usermod is another simple yet straightforward Linux utility to modify user account details. It supports similar parameters or flags as the useradd command and that’s why its usage is quite simple.

For instance, you can change the default shell of the user adem from /bin/sh to /bin/bash as follows:

usermod -s /bin/bash adem

Now to include adem in the sales group, you’ll need to use the -aG flag as a simple -G flag will remove the user from the previously added supplementary groups: apple and linux.

usermod -aG sales adem
cat /etc/group | grep adem

Related: The Best Command Line Utilities for Viewing File Content in Linux

How to Delete User Accounts on Linux

Linux offers another command-line utility userdel to delete any user account. Here’s the basic syntax:

userdel username

However, it will only remove the account details from the /etc/passwd file. To remove the user’s home directory as well, use the -r flag, as follows:

userdel -r username

As a precaution, we recommend finding all the files owned by the user and reassigning them to any other existing user account. Use the find command to list all the files either owned by the user or assigned to a user ID you have removed or not associated with any user.

find / -user username -ls
find / -uid 504 -ls
find / -nouser -ls 

Linux User Account Management in a Nutshell

This article demonstrates Linux user account creation, deletion, and modification examples with tips and tricks for any beginner Linux user who wants to pursue system administration and learn user account management.

It also shows how to edit the configuration files to define UID and GID ranges and change the default settings for user account creation in Linux.

About The Author