Sorry, I don't understand what you're asking. The way it works now is: each document is read and summarized into a bunch of word frequencies, once. It looks l

Sorry, I don't understand what you're asking. The way it works now is: each document is read and summarized into a bunch of word frequencies, once. It looks like each is being read many times, but since I'm memoizing, it only happens once. Is that helpful or are you suggesting something else?