In WQ5, Q4 if you use the numerical features that are provided earlier in the quiz instead of treating all non-object columns as numeric, you end up with a qualitatively different result. I found this confusing because explicit instructions for determining which features are categorical are provided while instructions to use all numeric data as features (e.g. including the incrementing id column) are not. Taking that with the list that is provided earlier can lead students to a different result than is intended.
We can emphasize a bit more the following sentence:
Create a preprocessor by dealing separately with the numerical and categorical columns. For the sake of simplicity, we will define the categorical columns as the columns with an object data type.
to:
Create a preprocessor by dealing separately with the numerical and categorical columns. For the sake of simplicity, we will define the categorical columns as the columns with an object data type while other columns would be consider as numerical columns.
Changes have been done here: FIX be more specific · INRIA/scikit-learn-mooc@0c2be15 · GitHub
FUN will be soon updated
Changes have been done here: FIX be more specific · INRIA/scikit-learn-mooc@0c2be15 · GitHub
FUN will be soon updated