f2fa9bd74d0141c2869d2090ed28fa4f9ffa49e4,GPy/likelihoods/noise_models/binomial_noise.py,Binomial,_moments_match_analytical,#Binomial#Any#Any#Any#,37

Before Change

            mu_hat = v_i/tau_i + data_i*phi/(Z_hat*np.sqrt(tau_i**2 + tau_i))
            sigma2_hat = 1./tau_i - (phi/((tau_i**2+tau_i)*Z_hat))*(z+phi/Z_hat)

        elif isinstance(self.gp_link,gp_transformations.Step):
            Z_hat = None
            mu_hat = None
            sigma2_hat = None

        return Z_hat, mu_hat, sigma2_hat

After Change

            mu_hat = v_i/tau_i + data_i*phi/(Z_hat*np.sqrt(tau_i**2 + tau_i))
            sigma2_hat = 1./tau_i - (phi/((tau_i**2+tau_i)*Z_hat))*(z+phi/Z_hat)

        elif isinstance(self.gp_link,gp_transformations.Heaviside):
            a = data_i*v_i/np.sqrt(tau_i)
            Z_hat = std_norm_cdf(a)
            N = std_norm_pdf(a)
            mu_hat = v_i/tau_i + data_i*N/Z_hat/np.sqrt(tau_i)
            sigma2_hat = (1. - a*N/Z_hat - np.square(N/Z_hat))/tau_i
            if np.any(np.isnan([Z_hat, mu_hat, sigma2_hat])):
                stop

        return Z_hat, mu_hat, sigma2_hat

Instances