3 reasons why all teams should learn R

Digital Transformation in the Modern Era

In our increasingly hyper-connected and digitized world, tremendous amounts of data are generated daily from the many online interactions that take place. Many organizations recognize the value of this and have embarked on digital transformations to leverage the burgeoning volumes of data to drive their businesses.

As enterprises strive to become data-driven, there are several key components that they must get right, one of which is equipping the employees with the right data tools to do their best work.

R is a powerful tool that forms an integral part of the modern-day data science toolkit. In this blog post, we explore the capabilities of R and the compelling reasons why you should train your teams on it.

What is R?

R is an open-source programming language optimized for statistical analysis and data visualization. Developed in the early 1990s by statisticians Ross Ihaka and Robert Gentlemen, it has grown into a reputable data mining and analysis ecosystem.

Its goal is to create a more user-friendly way of performing statistics, data analysis, and data visualization. Despite its long history, R is still highly prevalent in data science today, and is often considered the programming language with the lower barrier to entry for beginners.

R is home to a rich community-driven ecosystem comprising more than 17,000 curated packages within the CRAN repository. R packages are akin to apps that allow practitioners to perform a variety of tasks on R. With audited contributions from data professionals, the repository consists of many models and tools that empower users to prepare data, build powerful statistical models, and create beautiful visualizations.

Here are some industry use cases of R:

• ANZ (Australia and New Zealand Banking Group) used R in credit risk analysis to assess the probability of loan default

• John Deere used R to forecast customer demand for their equipment so that they can adjust optimally to the factors that impact order fulfillment

• Zillow, a leading real estate marketplace in the USA, used R to estimate housing prices.

• The City of Chicago used R to predict which restaurants are likely to commit violations in sanitation inspections to prioritize these outlets for review.

• AirBnB developed internal R packages to facilitate the efficient movement of data across various storage locations (e.g. Presto, AWS S3)

If you are wondering how R got its name, it is based on the first names of the two creators (Ross Ihaka and Robert Gentleman) and a play on the name of the older S programming language, on which R is built.

How is R useful?

(i) Ease of use and accessibility

A key strength of R is the presence of numerous well-established packages for data manipulation and statistical analysis. Its open-source nature also means that anybody can gain access to R’s rich capabilities. Within the extensive R ecosystem, Tidyverse is the most famous collection of R packages for data science.

Tidyverse is a collection of easy-to-use packages designed for data import, manipulation, visualization, and reporting tasks. These packages share the same design, grammar, and data structures, which in turn streamlines the learning of R since getting familiar with one package allows you to transition over to the next one readily.

R is also commonly considered as one of the easier programming languages for data manipulation, so the barrier to learning and applying R is relatively low.

While spreadsheet software like Excel can perform data analysis, it struggles to handle long repetitive data manipulation tasks involving big datasets. If you have tried working on large datasets of more than 100,000 rows in Excel, you will understand how tediously slow the program can become.

On the contrary, R can produce detailed analyses efficiently, even for large datasets. This is helpful for projects where you need to process a myriad of large complex datasets repeatedly and cannot afford to consume too much time or computing resources.

(ii) Data manipulation and visualization

The Tidyverse collection contains packages designed for data-related tasks, including popular ones like:

• dplyr – Comprises a set of easy-to-understand commands for data manipulation

• tidyr – Provides functions to create data in a tidy format for analysis and storage

• ggplot2 – Consists of methods and functions for building impactful data visualizations

When used in tandem, these packages allow users to perform data manipulation and analysis efficiently and present insights effectively in highly precise and informative visualizations.

(iii) Reporting and dashboarding

After performing data manipulation, analysis, and modeling, the final (and arguably the most important) step is to ensure that the insights are communicated meaningfully.

Besides static visualizations built with ggplot2, interactive dashboards can be created for stakeholders to engage in self-service business intelligence. These dashboards allow business users to directly access the dashboards by themselves to answer their own data questions

Shiny is an excellent R package that enables people to build and publish dashboards for sharing with others easily. Its ease of use allows even those without much technical experience to create powerful and professional dashboards.

Here is an example of a Shiny dashboard used for the monitoring of New Zealand trade information:

Democratizing data science with R

According to Forrester, companies make fewer than 50% of their decisions based on data instead of gut feeling, experience, or opinion. To unlock the value of data, employees need to upskill and equip themselves with tools to learn from their data efficiently and effectively.

The good news is that powerful data tools need not cost much at all. R is a free, open-source programming language that makes it easy to perform critical data science tasks like data manipulation, modeling, and visualization.

R is a practical yet intuitive skill for technical and non-technical employees to learn and apply, given its relatively low barrier to entry. With these valuable data skills in place, enterprises will be on their way to achieving the positive business outcomes of operating as a data-driven company.