Visualise distances between natural languages.
This uses the ASJP database to determine distance between pairs of languages. I use the Normalized Levenshtein Distance as described here, followed by MDS to (approximately) map the calculated distance matrix onto a 2D surface. It gives some quite nice clustering.
To change the languages plotted, just change the 'duolingo' variable. Language codes are normally as you'd expect, but can be checked in the dataset.tab file.
