Repost about L-inf calculation

Here is my original post as a follow up of question raised by others, seems the mentor have no activity recently, so I just repost it as a new question and context link attached below.

Please read the text from here

NOTE L-infinity norm will only detect skew for the categorical features. Instead of specifying an infinity_norm threshold, specifying a jensen_shannon_divergence threshold in the skew_comparator would detect skew for both numeric and categorical features.

I see and explored online to try to understand this concept. BTW, I think this is a very important and not sure why it even not mention in neither lecture / code assignment nor extended reading, I only found a link to wiki about Chebyshev distance which I think only applicable to categorical feature.

The course provides a good introduction to MLOps. It’s hard for any course to be exhaustive. Since this is an advanced specialization, please explore newer concepts and have fun.