discretionary_policy: fix noprint option

Output argument and return statement were missing

added bug optimal policy labels

Can you possibly clarify the purpose of this commit? I’m struggling to understand what it really achieves and the relation to the noprint option.

assigned to @sebastien

1. Within discretionary_policy_1.m we missed a required output. That would always crash discretionary_policy if the steady state could not be computed.
2. discretionary_policy_1.m calls stoch_simul, which will provide an error message in case of an error code. But if noprint was passed, it would return to discretionary_policy_1.m, which would try to output steady state values that could not be computed. Instead, we need to return to the caller function if there was an error code.

mentioned in commit 51e2ff30

merged