Models such as KNN and SVM or even decision trees typically need their data to be in numeric format. This is a problem with nominal ones since it cannot be fed directly to the model. Some sort of data preparation technique is needed. A dataframe that contains data about features of two different animals needs to be prepared to be fed into an SVM model. All other columns are numeric except for the column called "label" that determines if the record is a 'Dog' or a 'Cat. Write the code snippet that fix the problem and elaborate on why this step is necessary