A dataset derived from the Iris dataset used for photonic classification experiment. Generated via sampling of the normal distributions defined by principal components of the original Iris dataset. Four features and three classes, each of the three classes has 5000 samples.