Preparing data for predictive modeling

In this section, you will prepare the data to ready it for modeling with XGBoost. To start preparing to make a model, you’ll split our data into testing and training sets. As in Chapter 2, you will use the train_test_split function from Scikit-learn.

Follow these steps:

1. Split the housing data into training and test DataFrames: You can reuse the code from Chapter 2. To do this, swap housingX and housingy for irisdata. You can do both X and y in one line, by providing four variable names to store the results. We’ve chosen X_train, X_test, y_train, y_test. No other changes are needed from this step in Chapter 2. Now, you have 80% of the data in the training set, reserving 20% for testing the model. random_state=17 is again an arbitrary value to seed the random selection of which columns go into which set; by using the same value as we do here, you’ll have the same rows in your data, and our outputs will match:

   from...