Using label encoding to anonymise multiple columns of different data types

Question

I am trying to anonymise my df of 30+ columns. I only need to anonymise 2 columns, one column (ID) is an int data type and the other (type) is object data type.

What