Skip to content

Latest commit

 

History

History
112 lines (87 loc) · 9.95 KB

project2_desc.md

File metadata and controls

112 lines (87 loc) · 9.95 KB
Raw

GR5243 Fall 2018 Applied Data Science

Project 2 Shiny App Development (V 2.0)

--- App Improvement

In this project, working in teams, you will study a collection of R shiny apps previously developed by your peers to create an improved new app that is better, faster or more interesting. It should still be designed for exploring and visualizing a topic about New York City or United States using NYC Open Data or data.gov.

Challenge

For this project, we invite you to study a set of R/Shiny apps that were previously developed by students in ADS using either New York City's Open Data or datasets available on data.gov about United States.

Each team will be given a set of apps that used a common data source (will be posted in Piazza). These previous R/Shiny apps may have focused on different angles of the same data sets but often have similar functionalities.

Your challenge is to study them and create an improved version 2.0 app that integrates what you think are good about these apps and potential improvements.

A number of web applications using open data from data.gov can be found at https://www.data.gov/applications. You can recreate, extend and improve some of these applications' functionality or create something entirely from scratch.

For presentation, the team should present their shiny app as a product version 2.0 release demo and interesting findings from their experience. All shiny apps are required to be deployed (either on http://www.shinyapps.io/ or your own Shiny server).

All developments need to be carried out in group shared private repo on [https://www.github.com/TZstatsADS/] with clear project management log, taking advantage of GitHub issues.

Each week, we will give a tutorial in class and having live discussion and brainstorm sessions. The instruction team will join team discussions during class and online.

  • week 1: introduction and project description
  • week 2: shiny app development/shiny server setup/spatial visualization
  • week 3: [TBD] depends on topics emerge from online and in-class discussions.

Evaluation criteria

  • Value added of the improved app
    • Has it created more functionality?
    • Is it easier to use? Is it faster?
    • Does it display information more clearly?
    • Can it be used to answer new questions or exploring for new interesting trends?
    • [optional] are new data source integrated into this improved app?
  • Implementation of the app
    • are the data processed/integrated properly?
    • are necessary measures taken to account for data collection issues such as biases, missing values, etc?
  • Visualization
    • Is the information visualized clearly?
    • Is the visualization easy to understand?
  • Presentation and organization
    • Is the presentation convincing about the business value added of the improved apps?
    • Is the GitHub organized and prepared in a way that makes it easier for readers to appreciate the value of the app and understand how to use the app?

(More details will be posted as grading rubrics in courseoworks/canvas)

Suggested team workflow

  1. [wk1] Week 1 is the studying & planning week. Study previous apps assigned to you, read project description, data description, browse data and brainstorm for ideas.
  2. [wk1] Each team is strongly recommended to demonstrate project progress by posting a project plan with task assignments as issues on GitHub by September 27th.
  3. [wk1] As a team, study the assigned apps together (not necessarily face-to-face). During the discussion, discuss a number of key analyses/functionalities carefully and discuss possible improvements.
    • "What information is of value or interesting?"
    • "Which analysis/visualization is particularly informative?"
    • Good visualizations do not just look cool. They should be easy to understand and give the correct impression about data.
  4. [wk2] Based on outcomes from week 1 brainstorm sessions, identify additional data processing and modeling needed (start early on this one!)
  5. [wk2] Week 2 is the design and development week. Start learning and using the shiny library to do simple things. Some team members may know more about shiny than others. Share and learn together.
  6. [wk2] It is ok to have 2-3 leads to explore at the beginning of week 2 but it is better to converge on a single direction by the end of week 2 (October 3/4).
  7. [wk2] Construction of shiny platform can start early using fake/simulated data.
  8. [wk3] Week 3 is the shiny app version 2.0 implementation/production/testing week. By October 3/4 (the beginning of week 3), you should have all necessary data processing done and a project idea.
  9. [wk3] By October 3/4, You should layout a to-do list (as a 2nd set of GitHub issues) for the app production and divide up tasks. At the session on October 3/4, team members should work together and resolve any ambiguity about which team member should be doing what for the final stage of production of the app.

Working together

  • Setup a GitHub project folder from joining the GitHub classroom link with everyone listed as contributors. Everyone clones the project locally via your GitHub desktop and create a local branch.
  • The team can work in subgroups of 2-3, which might meet more frequently than the entire team. However, everyone should check in regularly on group discussion online and changes in the GitHub folder.
  • Learn to work together is an important learning goal of this course.

Resources

Here the examples are for inspiration. The order is abitrary. I do not endorse any of these analyses to be a model analysis for you to follow. The tools included are those I think might be useful. Please feel free to send me stories and/or tools that you would like to contribute to this list.

Stories

Blogs

News articles:

Examples

A small set of potential tools

*These are not fully tested and more tools can be found.

Paper to read