Based on your data exploration, begin considering the features you want to include in your model. Limiting your data can be beneficial because it reduces complexity and can improve model performance by focusing on the most relevant features.

Create lists below for the columns you want to use in your model based on your exploration above. These features will be used in the column transformer. The list names must match exactly.

numeric$\_$columns: This is your continuous numerical data that MUST include absences$\_$G$1,$ absences$\_$G$2,$ absences$\_$G$3,$ G$1,$ and G$2$ for use in your custom transformer, in addition to any other numerical columns you want to select. Note: The fact that a column is labeled as an integer or float does not necessarily indicate that it contains continuous data.

categorical$\_$columns: Include at least one categorical column.

ordinal$\_$columns: Include at least one ordinal column.