R Programming By Example

Training and testing datasets

For us to be able to measure the predictive accuracy of our models, we need to use some observations to validate our results. This means that our data will be split into three different groups:

• Training data
• Testing data
• Predicting data

The predicting data is the data that we don't have complete cases for, specifically these are wards for which the Vote and Proportion variables have NA values. Our final objective is to provide predictions for these ward's Proportion and Vote variables using what we can learn from other wards for which we do have data for these variables, and it's something we'll do toward the end of the chapter.

The data that has complete cases will be split into two parts, training, and testing ...