wanna know if the n_features of sequencing data is subjective

[CeciliaPYY]

What you written in your README, about Supported Data Types, is that the n_features of imdb data is 1, and that of music mfcc data is 13, I wanna know if the number of features about one's own dataset is subjective or decided by some kind of result of machine learning methods, such as random forest???

Many thanks!!!