wrong code to calculate the expected utility function with respect to the initial state distribution

[siyaoZHAO]

Hi, great work but there is problem about the code in file dilo.py.
Lines 137~143 is to calculate the expected utility function with respect to the initial state distribution. The coefficient before the expected value is (1 - lambda), however in the formula of the paper the coefficient is (1 - discount)*beta. And there also should not be a divergence type conditional check in the code.

Is that wrong in this code?