Donated on 4/23/2015

The data are binary attack-point vectors and their clave-direction class(es) according to the partido-alto-based paradigm.

Dataset Characteristics


Subject Area


Associated Tasks


Feature Type


# Instances


# Features


Dataset Information

Additional Information

The data consist of 16 binary inputs and one 'four-bit' one-hot classification output. The 16-bit inputs are binary-valued attack-point vectors. 1 indicates the substantial presence (0, absence) of an onset (note start) in a certain time window during one bar of 4/4 time music (not limited to percussion, hence *onset* vectors without duration) quantized to 16th-note subdivisions. Each vector has 16 positions in which there may be or not be an onset. The output classes (left to right: neutral, reverse clave, forward clave, and incoherent) were determined through the music-theoretic/ethnomusicological portion of the my dissertation studies, based on both double-blind listening tests and informal interviews with with four professional master-musicians, as well as decades of studying the music. Future uploads (subject only to formatting) can include an additional column of fuzzy descriptors (of the degree of match to the output class).

Has Missing Values?


Variable Information

In terms of divisive rhythm counting, the first 16 attributes (input bits) correspond to a significant onset at: 1 e & a 2 e & a 3 e & a 4 e & a of one bar of 4/4 time. The last four are the output classes (3 - neutral, 2 - reverse clave, 1 - forward clave, 0 - incoherent) in one-hot (one-up) encoding.

0 citations


Mehmet Vurka


By using the UCI Machine Learning Repository, you acknowledge and accept the cookies and privacy practices used by the UCI Machine Learning Repository.

Read Policy