Thompson sampling get_satisfaction_from_true_distribution()

[fedeotto]

Hello, in the code you provide for Thompson sampling I don't understand the following:

    def get_satisfaction_from_true_distribution(self):
        s = np.random.normal(self.mu, self.sigma)
        self.n += 1
        self.sum_satisfaction += s
        return s

In the first line we're just returning a value sampled from the true distribution of Restaurant satisfaction, but then why do you write:

        self.n += 1
        self.sum_satisfaction += s

n is not defined in the class variables, raising an error.. and the same holds for sum_satisfaction..