85804dcc5009e23e843d7e8a78b0cca0d4f726bf,pyglmnet.py,glm,qu,#glm#Any#,28

Before Change

            eps = np.spacing(1)
            q = np.log(1 + eps + np.exp(z))
        elif self.distr == "normal":
            q = z
        elif self.distr == "binomial":
            q = expit(z)
        elif self.distr == "multinomial":
            q = softmax(z)
        return q

    // Define the conditional intensity function
    def lmb(self, beta0, beta, x):
        z = beta0 + np.dot(x, beta)

After Change

    def qu(self, z):
        eps = np.spacing(1)
        qu = dict(poisson=np.log(1 + eps + np.exp(z)),
                  normal=z, binomial=expit(z),
                  multinomial=softmax(z))
        return qu[self.distr]

    // Define the conditional intensity function
    def lmb(self, beta0, beta, x):
        z = beta0 + np.dot(x, beta)

Instances