In the last few years, businesses have adopted machine learning and AI technology on a large scale. With this mass adoption, programming languages which are actualizing ML and AI have blossomed.

R programming language is amongst the few ML programming languages which are satisfying the need of data scientists in the best possible way. In the last five years, the popularity of the R programming language has crossed many milestones.

Thus, considering the gravity, in this blog I will share every little thing about R programming language with the use of R as well as a tutorial of machine learning using R. So, without any further due, let’s dive straight into it.

## R programming language – Introduction

R is a programming language and free software environment commonly used in machine learning and AI applications like statistical computing, data analytics, and scientific research.

Almost all statisticians, data scientists, and marketers are using the R programming language to retrieve, process, analyze, visualize and present the data. Its expressive syntax and easy-to-use interface are the major two reasons behind its increasing popularity in recent years.

R programming language is mainly used for developing statistical software and do data analysis. Though R has the command-line interface, it supports several graphical user interfaces such as R Studio which shows the output in a graphical way. R Studio is generally called IDS (Integrated Development Environment).

Cross-platform interoperability is the very useful feature of the R. Since it supports cross-platform, one can write code of R in Window, Mac or even Linux. Not only this, he can easily transfer the code from one platform to another.

O’Reilly Media conducted a survey in 2014 and according to that survey, data scientists are paid a median of $98,000 worldwide. Knowing R language also makes you stand out if you are applying for any kind of development job.

## R Programming Language – History

R is the implementation of S programming language which is nothing but another statistical programming language. S was written by John Chambers in 1976 while he was working in the Bell Labs.

Later, using the concept of S, Ross Ihaka and Robert Gentleman created R at the University of Auckland, New Zealand. They named their written language after the first names of themselves. The initial version of the R was released in 1995 and the stable version of R was released in 2000. Currently, R Development Core Team is maintaining the language and bringing new changes.

## R Programming Language – usefulness

Following are the most common applications of the R which you may notice in real-life.

### 1) Data science

Mobile apps and computer systems have already been collecting data but with the advent of IoT devices or internet-connected sensors, the game of data has been changed. Now, businesses are collecting a heap of data that contains really vital information about users, their purchasing behavior and many other business aspects. To retrieve meaningful relationships between different attributes of that data, data scientists use R programming language along with RStudio and apply concepts of data science.

### 2) Statistical Computing

Statisticians love R, and there is a reason for it.

To derive the meaningful relationships between the different attributes of the data, one should apply a lot of mathematical formulas. And for data storing limited attributes, it would be easy to apply mathematical formulas manually. But in the case with big data where we are storing millions of attributes in the data, one simply cannot apply mathematical formulas on it manually; the only option is to deploy R which stores every statistical function and a rick package repository with 91000 plus packages. With R’s expressive syntax, those from the non-computer science background can also write statistical functions and quickly import, clean and analyze the data.

R shows the output in charts. Meaning, you can create a visualization of any dataset for better understanding and implementation.

### 3) Machine Learning

Machine learning is really sought after technology which predicts the future based on the historical data. It uses an army of machine learning algorithms to do this almost impossible task. These machine learning algorithms are so complex, that it needs various packages which contain some part of their business logic. And guess what?

R carries various packages for common machine learning algorithms like Linear and non-linear regression, decision trees, linear and non-linear classification and many more. Machine learning engineers are using R for almost all fields like Finance, Genetics Research, retail, Marketing, and Health care.

### R Programming Language – Top ‘users’

Following is the list of top companies which are using R to satisfy their forecasting and data analytics needs.

Companies | Why they are using R? |

| |

Ford | They are using R to analyze social media to take design decisions for their cars. |

Twitter relies on R to monitor user experience. | |

New York Times | They are using R for data journalism and infographics. |

Human Right Data Analysis Group | Using R, they measure the effect of war on mankind. |

Uber | Uber calculates the driving time and pick-up time using R. |

Microsoft | Microsoft has released
Microsoft R Open with enhanced R distribution and Microsoft R Server. |

## R programming language – alternatives

There are a lot of options available when it comes to a language that you use for statistical computing and graphics. Following are the most reliable alternatives to the R programming language.

### Python

Python is a really popular interpreted, high-level, general-purpose programming language. The major two reasons why Python is the most suitable alternative of the Python are its easy-to-use code and simple syntax.

Another reason why Python is equivalent to R is that Python has most of the packages of R. Meaning, data scientists and researchers do not have to write more code in Python compared to R.

So, since both languages have a lot of similarities, data scientists and researchers opt for a language that they know better. However, according to their goals, they switch between Python and R. For instance, if they want to analyze the dataset and show findings in the research paper, they choose R. And if they want to write a data analysis program that runs in a distributed system and interacts with lots of components, they prefer Python.

### SAS

SAS (Statistical Analysis System) is a powerful statistical software and SMEs are using it heavily. Thanks to its GUI, comprehensive documentation, and reliable technical support, small and medium-sized enterprises are deriving a lot of benefits without investing a lot, both time and knowledge.

However, unlike R, SAS struggles with academic and research data. But SAS is very useful to do commercial analysis.

### SPSS

SPSS (Software package for statistical analysis) is another purposeful statistical tool. It is mainly exercised in social science. Using an SPSS is so easy that even a non-statistician can use it; it is just like Excel and perhaps just because of this, SPSS is the easiest to learn amongst enterprise statistical tools.

However, the only drawback of using SPSS is its pricing; it is expensive, really expensive. Currently, IBM is enjoying the ownership of the SPSS after acquiring it in 2009 for $1.2 billion.

## R Programming Language – Tutorial

Here is the fun part! Let’s see some basics of the R programming language with a problem statement and solution.

### Run R Programming in Windows

To download the R for Windows machine, follow the simple steps.

- Click here to go on the official site of R.
- Look at the sidebar, you will see CRAN link. Click on it.
- Select a mirror.
- Find “Download R for Windows” and click on it.
- Run the files and install R.

Here, two things are worth mentioning. First, if you are running R on a 32-bit machine, install only the 32-bit version of the R. And second, You need to install R-Studio separately.

### Installing R Studio

R studio is an IDE for running R. And to install it, you have to follow the simple process. Download the R Studio and install it. But make sure you install R before installing R studio.

After installing R Studio, when you open it for the first time, it will ask which version of R is running.

Also, R detects whether R is installed on your system or not. If it is not, a warning box pops up. And if R has already been installed, you will see the R Studio home screen.

Then, you need to create R script in which you can write the code and execute it. To create the R Script, click on File>New File>R Script. Once you do it, you can now see R Script Editor. The code written in R Script is debugged by the R Studio itself.

The following image shows 4 sections of the R studio GUI.

So, now when you know the basics of R and RStudio, let’s pore over the first R program.

### First R program** **

We will generate 10000 numbers in a non-uniform manner, and then organize them by taking frequency into account and create a chart.

Here is the code for it.

n <-floor(rnorm (10000, 500, 100)

T<-table(n)

barplot(t)

To run this program, all you need to do is, CTRL+A (which selects all lines) and press CTRL+Enter. The result would be like the following image. It is the graph showing the bell curve.

Before moving further, and counting the occurrence of each value, let’s understand what each line means.

The first line (n <-floor(rnorm (10000, 500, 100)) creates 10000 random numbers in such a pattern that 500 is the mean value of these numbers and 100 is the deviation of the numbers. And the floor function is responsible for removing decimal points.

Now, let’s study the second line of the code which counts the occurrence of each value.

T<-table(n) holds the Table function which takes 10000 numbers as input and gives the frequency of each value as output.

The last line of the code calculates frequencies of the bar graph. The bar graph function considers the table of frequency and generates the bar chart.

This is what R can do. If a company has a lot of unstructured data, it simply takes meaningful relationships out of unstructured data by just writing a code of 3 lines!

# In the nutshell

In so many ways, the R programming language is magic. It processes the big data and tells every little thing about people who buy something from a store. In fact, it tells what will they buy if they will come to visit the shop for the second time! Thanks to such kind of serviceability, R has become the first choice of data scientists. It shows the result into the graphical format and it enjoys cross-platform capability. However, compared to its alternatives, R is a bit hard language to implement. It follows the same concepts of other programming language but when it comes to the syntax of the commands, R is really unique. So, you always need to R separately and if you already know some other languages than R, make sure you don’t try to learn R by taking your known language as a reference.