Create categorical dataset
A = CAT2DATA(DATA,LABELS)
This routine creates a dataset of categorical data, either given as integers, or as a set of string arrays. If given the dataset is labeled by LABELS. In case DATA is a cell array of size [1,K], it is assumed that every cell contains a string array of M named categories, in which M is the number of objects. In case DATA is a cell array of size [M,1], it is assumed that every cell contains a string array of K named categories, one for every feature. Alternatively DATA can be a [M,K] cell array containing a single string per cell.
The routine FEAT2LAB may be used to convert a categorical feature into a class label.
A categorical dataset A can be combined with a dataset B representing a set of numeric features for the same objects by [A B].