I was playing around with R’s ggplot2 library when I came across a data file consisting of 500,000 IP geolocations. I downloaded the CSV file, which contains the latitudes and longitudes of each geolocation, and decided to plot the latitude and longitude values on a scatterplot to see what would happen. This was partially inspired by the Visualizing Friends work by Facebook, although it is a very quick bit of experimentation.
I had to make a decision regarding image size and alpha values. It seems like many places are very spread out compared to Europe, which is quite densely populated. I rendered the image to be 12700 x 12120 pixels, with a 1/5 alpha value on the points. This tries to show the dispersion of other places while highlighting the more compact areas of Europe, although it works best when you zoom in. I have different versions which show other continents a bit more clearly.
I didn’t create this to necessarily mean anything, I just wanted to see how ggplot2 would handle it. I’ll hopefully get a chance to play around with Geolocations again in the future, possibly looking at mapping Tweets in a similar way.
The full-size image. (Warning: 12MB, 12700 x 12120)
Reduced versions of the UK and the US, subset by Country (with different alpha values).