Data must be introduced as a classification square matrix. In each row, different elements must be separated by comma ','. If there is a gold standard, his or her classification is expected to appear by rows. On the other hand, if there are marginal frequencies set beforehand, they are expected to appear by rows.