Machine Learning in Java - Second Edition

Loading data

Before we start the analysis, we will load the data in Weka's Attribute-Relation File Format (ARFF) and print the total number of loaded instances. Each data sample is held within a DataSource object, while the complete dataset, accompanied by meta-information, is handled by the Instances object.

To load the input data, we will use the DataSource object that accepts a variety of file formats and converts them into Instances:

DataSource source = new DataSource("data/zoo.arff"); 
Instances data = source.getDataSet(); System.out.println(data.numInstances() + " instances loaded."); 
System.out.println(data.toString()); 

This will provide the number of loaded instances as output, as follows:

101 instances loaded.

We can also print ...