def cluster(data, k, mode = 'full'):
d = data.shape[1]
gm = GM(d, k, mode)
gmm = GMM(gm, 'random')
em = RegularizedEM(pcnt = pcnt, pval = pval)
em.train(data, gmm, maxiter = 20)
return gm, gmm.bic(data)
def _generate_data(nframes, d, k, mode = 'diag'):
N.random.seed(0)
w, mu, va = GM.gen_param(d, k, mode, spread = 1.5)
gm = GM.fromvalues(w, mu, va)
# Sample nframes frames from the model
data = gm.sample(nframes)
#++++++++++++++++++++++++++++++++++++++++++
# Approximate the models with classical EM
#++++++++++++++++++++++++++++++++++++++++++
emiter = 5
# Init the model
lgm = GM(d, k, mode)
gmm = GMM(lgm, 'kmean')
gmm.init(data)
gm0 = copy.copy(gmm.gm)
# The actual EM, with likelihood computation
like = N.zeros(emiter)
for i in range(emiter):
g, tgd = gmm.sufficient_statistics(data)
like[i] = N.sum(N.log(N.sum(tgd, 1)), axis = 0)
gmm.update_em(data, g)
return data, gm