I guess it should be k1 = np.log(2 * np.pi) * x.shape[0] + np.log(np.linalg.det(g['sigma'])) otherwise there will be errors
I guess it should be k1 = np.log(2 * np.pi) * x.shape[0] + np.log(np.linalg.det(g['sigma']))
otherwise there will be errors