In Example 4.12, we applied a principal component analysis (PCA) to the iris data, but refrained from classifying the flowers based on their feature vectors \(\boldsymbol{x}\). Implement a 1-nearest neighbor algorithm, using a training set of 50 randomly chosen data pairs \((\boldsymbol{x}, y)\) from the iris data set. How many of the remaining 100 flowers are correctly classified? Now classify these entries with an off-the-shelf multi-logit classifier, e.g., such as can be found in the sklearn and statsmodels packages.