Changing the world doesn’t mean everyone has to make a big discovery; it can be as simple as becoming a better communicator and finding more creative ways to pass the knowledge we have onto the people who need it! 

Nowadays, we all live in a big data era, and data can come from every corner of our lives. How to be an effective story teller with big messy data is incredibly important for our lives. Traditionally, analyzing data and reporting the results are only the responsibilities of statisticians or analysts. However, with the development of technology, the process of storytelling with big data can be efficient and creative by using the Shiny package (a Web Application Framework package made for R). 

“Shiny” is an open source package for the R programming software that provides an elegant and powerful framework for building web applications. In addition, Shiny helps you turn your statistical or mathematical analyses into interactive web applications for better and more efficient storytelling with your data. Shiny provides opportunities for building web applications without requiring HTML, CSS, or JavaScript knowledge, so the learning curve is not steep at all. 

Here is a screenshot of one of my demo shiny apps.  This plot shows the population density in Canada. Each point represents a city in Canada, and the size of the point represents the population density in that city. 

 

Figure 1:  Population Density in Canada

If the points are too dense to distinguish, users can zoom in to investigate more about the area they are interested in. Additionally, the users can mouse over a specific point to get more information about the city it represents. Here is the screenshot of the plot zoomed in: 

Figure 2:  Zoomed-In View of the Population Density of Canada

Since Shiny apps are written with R software (one of many statistical programming languages), you can use any of its available packages when you build your shiny apps, which means you can directly apply any statistical package or any statistical functions usable with R into your app instead of writing them by yourself. Also, you have the flexibility of designing the output of the results. Here is a regression analysis of an open source data set (the iris one). 

 

Figure 3:  Regression Analysis of the Iris Dataset

After the users get the analytical results or stories, they will usually want to save the results in an elegant report with the current date in PDF or any other kind of format. Therefore, it is important that an interactive web app for analytical purposes has an automatic reporting system. 

In terms of a reporting system, you can set one up for your Shiny app through Markdown. Markdown (syntax) can turn your analytical results into high quality documents, reports, presentations and dashboards. You have options to control which plots should be put into your report and you can write down your comments for each plot. After finishing, you can download your whole report by simply clicking the “download” button. 

Here is the screenshot of the PDF report from my demo app:

 

Figure 4:  Demo Analytical Report

Figure 5 is part of the PDF report from the app.  

Figure 5:  Analytical Results (Demo Analytical Report of Canada Map)

In educational fields, Shiny apps have been applied widely for helping students gain a better and deeper understanding of mathematical and / or analytical fields. For example, by using the interactive functions of Shiny apps, students will gain experience of how parameters affect statistical distributions or graphs. It is better to learn math or statistics by “playing” with them instead of memorizing tedious formulae from textbooks. 

In veterinary medical fields, Shiny apps with automatic reporting systems can help veterinarians or vet med-related businesses have a better and deeper understanding of their data and make better decisions by telling a story with their data. 

In addition, the automatic reporting system can help users record their data stories or reports. In this way, users can save a lot of time when reporting or analyzing their data weekly or monthly. 

The Beef Cattle Institute of K-State Vet Med is currently working on building an analytical dashboard with Shiny and Markdown for helping beef cattle industries to gain insights and make better decisions with beef cattle data. The Beef Cattle Institute will also provide analytics mobile apps to fit the needs of veterinarians and beef producers. 

Note:  The respective apps were created using R, Markdown syntax, and Javascript, as well as the Shiny app.  While the code is shareable, the files are way too long and complicated (six different files with thousands of lines for each demo app; three different files with hundreds of lines for each voice one).  Anyone interested in the code may contact the author directly.  

Acknowledgments:  Thanks for the help from Wendy L. Michaels, Alyssa Toillion, Yihui Xie (from RStudio, Inc.), and the K-State Beef Cattle Institute. 

 

Jiena Gu is the project coordinator of the Beef Cattle Institute (K-State Vet Med). Originally from the Guangdong Province in China, Gu came to K-State with a degree in bioengineering. She received her master’s degree in statistics from Kansas State University after she finished her Graduate Certificate in Public Health. She is interested in storytelling with big data by using various tools such as data visualization and applied statistics. Her expertise is in big data analytics, applied statistics and create interactive applications with automatic reporting system. She may be reached at jiena@vet.k-state.edu.