Path dataPath = new Path(dataStr);
FileSystem ifs = dataPath.getFileSystem(conf);
FSDataInputStream input = ifs.open(dataPath);
Scanner scanner = new Scanner(input);
DataConverter converter = new DataConverter(dataset);
int nbInstances = dataset.nbInstances();
int id = 0;
while (scanner.hasNextLine()) {
if (id % 1000 == 0) {
log.info(String.format("progress : %d / %d", id, nbInstances));