hdpmodel.py 文件源码

def lda_e_step(doc_word_ids, doc_word_counts, alpha, beta, max_iter=100):
    gamma = np.ones(len(alpha))
    expElogtheta = np.exp(dirichlet_expectation(gamma))
    betad = beta[:, doc_word_ids]
    phinorm = np.dot(expElogtheta, betad) + 1e-100
    counts = np.array(doc_word_counts)
    for _ in xrange(max_iter):
        lastgamma = gamma

        gamma = alpha + expElogtheta * np.dot(counts / phinorm, betad.T)
        Elogtheta = dirichlet_expectation(gamma)
        expElogtheta = np.exp(Elogtheta)
        phinorm = np.dot(expElogtheta, betad) + 1e-100
        meanchange = np.mean(abs(gamma - lastgamma))
        if (meanchange < meanchangethresh):
            break

    likelihood = np.sum(counts * np.log(phinorm))
    likelihood += np.sum((alpha - gamma) * Elogtheta)
    likelihood += np.sum(sp.gammaln(gamma) - sp.gammaln(alpha))
    likelihood += sp.gammaln(np.sum(alpha)) - sp.gammaln(np.sum(gamma))

    return (likelihood, gamma)