Data transformation and exploratory data analysis are two iterative steps. The objective is to improve the data quality to create a more accurate model. In order to transform your data, you need to understand it first. So, in real life, you can explore and transform iteratively until you are fine with your data.

For simplicity, we'll cover data transformation in this chapter and data exploration in the next chapter.

Data mining experts usually spend a lot of time preparing data before they start modeling. Preparing data is not as glamorous as creating predictive models but it has a great impact in the model performance. So, be patient and spend time to create a good dataset.

When we execute a transformation in a variable, Rattle doesn't modify the original variable. Rattle creates a new variable with a prefix that indicates the performed transformation and the name of the original variable. An example can be seen in the following screenshot:

We see the list of variables...