Stylometry analysis using word length

Based on a technique by TC Mendenhall.

Disclaimer: This algorithm isn't an accurate method of determining authorship, since the output varies greatly between verse and prose. However, it's still pretty interesting to compare different authors in this way, and the method does have some merits.

Required libraries: numpy, pandas, matplotlib, seaborn

Data sources

http://cs.stanford.edu/people/karpathy/char-rnn/shakespeare_input.txt

http://shakespeare.mit.edu/

http://www.gutenberg.org/ebooks/author/410

http://www.elizabethanauthors.org/king-leir-1605-1-16.htm

http://extra.shu.ac.uk/emls/iemls/renplays/miseries.htm

http://www.gutenberg.org/files/1962/1962-h/1962-h.htm

http://www.gutenberg.org/ebooks/author/296

https://en.wikipedia.org/wiki/Edward_de_Vere,_17th_Earl_of_Oxford#Literary_reputation