Within this pipeline, begin by incorporating a SimpleImputer transformation using the mean strategy. Please note that this strategy employs the "mean" instead of the "median" as used in the previous assignment. You might be wondering why we are introducing this step, given that there are no missing data in the numerical columns at the moment. While this holds true for the current dataset, we cannot always guarantee that incoming data will be free of missing values. Therefore, it is advisable to prepare for this possibility as a best practice. Next, apply the custom Assignment3Transformer class to the data. Finally, add a StandardScaler transformation into the pipeline.