Learning Microsoft's DP-100 and saw this notebook:
The part of its code which I have question with is
# Define preprocessing for categorical features (encode the Age column) categorical_features =  categorical_transformer = Pipeline(steps=[ ('onehot', OneHotEncoder(handle_unknown='ignore'))])
here is the link for the entire notebook for you to see the full picture. I cannot copy the full pages of code here.
My question is: Why is the DiabetesPedigree column(which is the 7th column) treated as a categorical feature?
To me, it looks nothing different from the rest of other columns(except the Diabetic column).
If DiabetesPedigree could be treated as categorical why not Pregnancies?
What are the rules to tell whether a column of data is numeric or categorical?