STA230 R Shiny Project (F25)

Description

This project is based on the R Shiny Project from Ryan Miller: https://remiller1450.github.io/s230f23/rshiny_project.html

The focus of this project is on creating interactive data exploration application. The primary product is an R Shiny application that allows a user to thoughtfully explore a data set.

In addition to creating your app, you will need to write a 1-2 page summary of the app that includes the 1 page justification of difficulty (see below).

\(~\)

Deadlines

• Friday, Dec.12: You should have a dataset with appropriate data processing completed so that it is useable with RShiny.
• Friday Dec.19: Final Project submission

\(~\)

Submissions

• All submissions will be to the “Midterm 4: R Shiny” assignment on gradescope
• All submissions should consist of zip files to allow for submission to the assignment (data, project summary, data cleaning and app code)

App Expectations

Your finished Shiny app is expected to include:

• Data visualization
  • Your app should use either ggplot2, plotly, or leaflet in some capacity. You may choose to use more than one of these packages.
• User options
  • A user of your app should be able to manipulate or explore your data in at least three ways. Some possibilities include:
    • selecting a specific variable for a particular visual aesthetic (x, y, color, fill, etc.)
    • choosing a scale transformation (log2, log10, scale reversal, color palate changes, etc.)
    • selecting faceting variables and/or manipulating facet formatting
    • hoverable or clickable content that displays additional information (markers, map regions, etc.)
    • different app pages or panels that a user can select to display different content

\(~\)

\(~\)

Groups

Groups are free-choice, max of 3 members, MINIMUM of 2. The final project is intended to be collaborative. See me if you do not have a group member after class on Monday 12/8.

\(~\)

Intermediate Progress

By the end of the day on Friday 12/12, you are expected to have code that cleans/manipulates your data to the point where you can create a sketch version of some type of graphic that you intend for your Shiny app to display. The sketch graphic you share does not need to ultimately be used in your app.

\(~\)

Submission

• By Friday Dec. 19, you must have a working version of your app that runs on your local PC.
• You must have a 1-2 page summary explaining the context of the data, goal of the displays, how cleaning was performed, and what motivated the particular graphical displays used in the app.

\(~\)

Assessment Details

App Code

• Code should be neatly formatted with appropriate comments
• Code should run on another PC without issues (assuming all packages are installed and all data files are moved to the proper location)
• Any substantive data cleaning is done as a preliminary step using a separate script, with the cleaned output (exported using write.csv()) being used in the final app
• Code should be reasonably clean and efficient.

Aesthetics

• The app’s user interface should be professional in appearance. This includes aesthetically pleasing labels, colors, and other design elements
• Figures and output generated by the app should also be of professional quality. This includes colors, labels, themes, visual choices, etc.

Function

• Your app should include at least 3 features that the user can manipulate
• Certain combinations of feature inputs should not be able to break your app
• Proper choices should be made to accommodate issues related to reactivity, challenges in graphing large amounts of data, etc.

Difficulty

• To account for the substantial variability of work required to effectively utilize different data sources you will be asked to argue for the level of difficult of your project.
• Your score in this category will be determined by the instructor, but your written narrative will be taken into consideration when determining that score.
• Additional details are provided in the section below

\(~\)

Level of Difficulty

One goal of this project is to afford you the opportunity to work with a topic that you find interesting. Unfortunately, real-world projects rarely utilize all areas of the data science workflow/life cycle equally. For example, some projects will require you to devote 90% of your time to data cleaning and manipulation in order produce a few relatively simple visualizations or models. Other projects might involve data come in a relatively clean format, and the majority of your time is spent making highly detailed visualizations or sophisticated models.

To address these differences, you will be asked to submit a \(\leq1\)-page written argument describing your project’s level of difficulty as part of the project summary. More specifically, you should argue that your project had “A-level”, “B-level”, or “C-level” difficulty, providing clear reasons and justification for your rating.

Hallmarks of an A-level project:

• Extensive data manipulation - substantial use of several distinct functions/methods covered in the tidyr, dplyr, merging and joining, stringr, and lubridate labs.
• Professional quality visualizations - thoughtfully constructed graphics that use customized labels, colors, and design choices
• Great functionality - options that are truly useful to an app user for gaining novel insights into your data. These generally go beyond simple filtering or variable selection options (though some of those are good too!)
• Code efficiency - legible and efficient code.
• Self-study - going beyond the contents of our in-class labs and researching methods and tools to accomplish something new. For example, you might learn about advanced UI features, or new types of graphs (including radar charts, treemaps, stacked area charts, sankey diagrams, etc.)

Note: If one area of your project is lacking in difficulty you should spend more time on and excel on others.

\(~\)

Additional Comments

R Shiny is a great technology to share and display your data science skills. I encourage you to consider hosting your finished app on shinyapps.io and storing your app’s code on github. If relevant, this allows you to include links to your project in a resume or cover letter to an internship or job opportunity. You can also embed a hosted R Shiny app directly into a personal webpage (if you have one).

Potential Projects

1. Economics-Trade:
   • Using BLS data https://www.bls.gov/bls/proghome.htm
   • Investigate relationships between Import/Export Price Indices and Consumer Expenditures
   • DO NOT make a map
2. Sports:
   • https://remiller1450.github.io/project_ideas.html
   • Investigate the types of figures that allow viewers to understand sports statistics
3. Education:
   • Use the college Scorecard Data: https://collegescorecard.ed.gov/data
   • Investigate relationships and effective presentations related to understanding college information
4. Other:
   • None of these projects sound interesting to you, so you’d rather work on something else
   • If this is the case, please talk to me EARLY to double check the data

GENERATIVE AI

In general, I am not a proponent of using generative AI in learning environments. I believe it is intellectually lazy (link 1 and link 2) and burgeoning research shows we fool ourselves into thinking we are learning effectively (link), when we really are not. That said, it can be an effective tool when it comes to synthesizing a desired output that would require extensive knowledge or time commitment, when developing that skill is not the primary objective.

The goal for this assignment is to think deeply about effective displays and getting a working interface that allows us to explore and learn from the dataset we are using, not to master RShiny. You may use generative AI to facilitate the data cleaning and ShinyApp coding for this project, but you MAY NOT use it to determine what should be included in the app, to make your summary, or to guide any other aspect of your project. Use of generative AI must be acknowledged in your summary, including a detail of which components of the project it was used for. Extensive use of generative AI without acknowledgement will result in a non-satisfactory score for the lab and will violate the Academic Honesty policy.

In addition, all code, methods, analysis, and AI prompts and output must be included in the project summary. All group members are responsible for understanding all aspects of your submission. Failure to explain any part of these in the summary will also result in a non-satisfactory score and could be grounds for violation of the Academic Honesty policy depending on how AI factors into this.