UFO Sightings App

Ben Bray
Developing Data Products - Johns Hopkins Data Science

  1. This simple App provides several alternative visualizations for UFO Sighting Data

  2. User can adjust the sort of the barchart and highlight a certain state of interest

  3. Coloring highlights trends among certain regions

The National UFO Reporting Center gives a listing of number of sightings per state. However, thinking about this phenomenon on a state by state basis causes problems. Since I'm from California, the state with the most UFO sightings, I'd like to point out...



More populous states have more people to see (or think they see) UFOs.



Larger states have more sky. A meteor is much more likely to hit Texas than to hit Rhode Island.



Although these ideas are fairly obvious, the purpose of the app is to make them visually obvious.

 
#number of sightings in California
CalSightings<-ufo$UFO_Sightings[ufo$State=="CALIFORNIA"]
#highest number of sightings of any state
MaxSightings<-max(ufo$UFO_Sightings)
#Sorted by number of sightings per 100,000 people
#finds ranking of California when ordered by sightings per capita
CalRankingPerCapita<-which(ufo[order(-ufo[,6]), c(1)]=="CALIFORNIA")

California had 10871 UFO sightings, the maximum number of sightings for any state was the same: 10871. California leads the pack here. However, when you adjust for population, California's ranking falls to 22. Who came in first?

ufo[order(-ufo[,6]), c(1)][1]

## [1] WASHINGTON
## 51 Levels: ALABAMA ALASKA ARIZONA ARKANSAS CALIFORNIA ... WYOMING

The contrast for California is made clear in the screen captures from the app shown to the left. I've added a legend signifying region by color. I've done this so that users of the app can indulge in their favorite regional stereotypes, or speculate about where the aliens plan to attack first, whichever is their preference. There definitely does seem to be something going on in the western part of the country.



A total of 81821 sightings have been recorded in the US. The image above was supposed to be a background for this slide. It shows up in knitr, or from the slidify index.html doc, but is not visible from R-pubs. I do not know the reason for this.