Welcome to our digication page on our Shiny Web App that we developed in RStudio!
This shiny app was created to generate reactive graphs of a random sample from a population living on a virtual island. The source of the data was a University of New Mexico (UNM) virtual world with interactive “citizens.” We were able to engage with these virtual citizens to obtain information about their lifestyle that we could use in our data analysis. The UNM islands application (“Islands”) was very interactive--we were required to ask the virtual citizens for consent to be interviewed. From the different cities, we were able to gather information on height, weight, age, food preference, and physical activity levels. Further, we were able to derive body mass index (BMI) before we began our true data analysis on these variables’ relationships to one another. Some of the information that could best be derived from this data looks into the very strong relationships between height, weight, and BMI. These variables are all interconnected and generally show strong correlations due to their impact on overall body type.
The shiny app that we have developed allows users to explore into our dataset and the relationships amongst the variables that we recorded. Taking the data from a .csv file, R is able to build graphs that change based on the inputs from the user. Our user have several options that will impact the graph generated. A couple of our options include radio buttons that allow the user to alter the explanatory and response variables. When these values are changed, the scatterplot and line-of-best-fit change according to which variables are selected.
Another option for our app includes a slider input bar which adjusts the sample size. Changing the sample size will impact the regression line as the individual sample points have more or less influence. As the sample size decreases, the individual points have a greater impact on the regression. On the flip side, with a larger sample, the individual points have a lesser impact on the regression, which tends to reflect the true mean of the population
The last elements of our shiny app will build a box-and-whisker plot for the BMI variable to illustrate our descriptive statistics for that variable. The box-and-whisker plot is an image to encompass the trends in the dataset. The ‘whiskers’ indicate the maximum and minimum values of the dataset. The ends of the box are the 25th and 75th percentiles with a single vertical bar depicting the median. This too will be reactive to the changing sample size as the medians and quartiles change.