![]() If a user filters on a certain criterion (e.g. An important decision for designers is how to treat missing values when filtering observations. Users see colleges appear and disappear on the map, as well the total number of colleges that match their criteria. In the context of searching among thousands of items, colleges and universities in our case, filtering is perhaps the most useful way of interacting with the data. size, selectivity, or type) are mapped to color and size of the markers. I use interactivity to filter which colleges are displayed, and to control which characteristics (e.g. Shiny provides a full slate of interactive elements: sliders, check boxes, drop-down lists. It combines nine different raw data files, selects and calculates variables and outputs a single data set used by the app. The code to update College Explorer is about 150 lines. Using API to access the data is ideal, but if API does not exist, raw data should be manipulated with code so that the manipulation is reproducible and can be easily updated. The key in data preparation is to automate the preparation as much as possible. my data comes from eight different IPEDS files and one CollegeScorecard file - two publicly available data sets). Data sources may need to be combined (e.g. most classifications of colleges and universities have 10-20 categories – a bit too many for effective visualization). college selectivity equals admitted students divided by total applications), and categorical variables might need to be consolidated (e.g. We rarely encounter data that is out-of-the-box ready for analysis or visualization. Compared to existing college search tools, this app combines quantitative and qualitative criteria with geographic location, enabling familiar map zooming and panning. When a user clicks on a marker, a pop-up window will appears with information and a link to the college or university website. The sliders and drop-down lists are driven by a variety of criteria: college or university selectivity, size, cost, SAT scores, salaries, etc. The colleges displayed, and the color and size of the markers are controlled with sliders and drop-down lists. colleges and universities while displaying their geographic location on a map. Rather than going line-by-line through code, I highlight key decisions, some programming issues and capabilities of Shiny, focusing on the potential for combining free data with open-source tools to create useful data products.Īs an illustration I use College Explorer – a Shiny web app I developed. I briefly compare that process to building a similar product in Tableau. I discuss developing an app using Shiny – a powerful R package. ![]() The purpose of this post is to discuss the key elements in developing an interactive web application that displays data with geographic component. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |