In Using numerical and categorical variables together notebook, Handling categorical data section, in the explanation of ColumnTransformer, second step:
Indeed, each subset that has been created will is affected a specific transformer.
should be:
Indeed, each subset that has been created will be affected a specific transformer.
Next sentence:
Thus, a
ColumnTransformeris just another transformer: it gets some input data and will output some transformed data
In the following pipeline description:
the
predictmethod makes predictions on new data;
A more general remark, there is some inconsistency with paragraphs titles, some use infinite form, like Dispatch columns to a specific processor, while other use -ing form, Fitting a more powerful model.
Last sentence of this notebook, there is an extra ‘s’:
seen that gradient boosting methods can outperforms the basic linear approach.