music-information-retrieval

Is your feature request related to a problem? Please describe.
Using the CLI to output parameters, those audio features who are multidimensional like amplitudeSpectrumand mfcc will not get corresponding label names in the first row of the generated csv-file. This makes it complicated to import using csv-importers like pandas in Python.
It's also unclear which bin each column represents.

The current implementation of NESMusicDatabase does not handle the training-validation-test splits provided in the original dataset. To avoid changing the base Dataset class too much, we could add a subset method and achieve something like the following.

nes = muspy.NESMusicDatabase("data/nes/")

training_set = nes.subset("training")  # also a Dataset object
validation_set =

frames[-1] = np.append(frames[-1], np.array([0]*(frame_length - len(frames[0]))))

TypeError: can't multiply sequence by non-int of type 'float'