code

[elcarim]

"The code is strange, especially the variable corpus which seems to have no purpose at all."

"Assigning idf to wordIDFs[word] will always result in 1 for any given word."

"The tfidf value will always be equal to tf."

"We can achieve the same results by simply taking the most frequently occurring words!"

[jessthebp]

Oh, you're totally right! I mindlessly reworked some tf:idf for multi-document stuff and didn't... finish changing it for single document keyword extraction.

My original plan was to have it compare to a corpus of text in a list-- basically "your docs here" (since it's originally for SEO purposes, probably other ranking docs. But I didn't implement that and it's not reflected in the code

I'm going to update the TF IDF code to include an alternate corpus, and update the name of the tf_idf snippet to just TF. I legitimately really appreciate your callout here.