
Well, I cheated. The autocorrelation plot at the end of the last blog was particularly poor because I included the text header of the file and only processed the first 1000 words. It looked like this: Simply by removing the text header (which is clearly not part of our repeating patterns) the plot improves to […]