Automated Topic Analysis of the Voynich Manuscript

[MarcoP]

I am not sure I understand correctly, but the PCA K=2 plots show a single outlier at the top that makes the plots basically useless? Is it so? Or maybe it's a set of samples that generate identical outlier dots?

Anyway, if that actually is the situation, investigating and removing, or fixing, that sample (or samples) could improve the quality of the overall analysis.