Python Training dataset

[jagathg]

import numpy as np
import scipy.io.wavfile
from scikits.talkbox.features import mfcc
sr1,x1=scipy.io.wavfile.read("filename.wav")
ceps1,mspec1,spec1=mfcc(x1)
I can obtain MFCC values as above. Now, I have 100 songs with 2 different music genres(electronic,jazz). And, I want to create a dataset with using these songs. I suppose I will use numpy arrays to design that dataset. How can I create that dataset and put them label(electronic or jazz)? I am using python2.7 on ubuntu