Questions about the c-tf-idf

Hi Maarten, I have some questions about the c-tf-idf implementation

Is there a reason why you’re using “A” (average number of words per class)? I saw your past medium post about c-tf-idf and I don’t see a clear connection between the two formulas. According to the code, it seems more like a regularization value, is that right? Also, just to be sure:

• does tf_x,c refer to the count of word x divided by the total words across all documents in c?
• does f_x refer to the count of word x divided by the total number of documents?

Thanks, and I appreciate all the work you put in this library 😃

[MaartenGr]

Just to be sure: I assume that the given image in the code is still a leftover from the old implementation, am I right?

The image that you can find here is currently most up-to-date and should reflect the procedure accurately.

I really appreciate your work and the excellent documentation.

Thank you for your kind words! If you have any other questions, please let me know.

[lkluo]

Thanks @MaartenGr, it is very clear.