Closed ataymano closed 2 years ago
@olgavrou, I got tests failures here: https://github.com/VowpalWabbit/estimators/blob/82e5bc198dbab569e04afbfca926b0adfc462f1f/estimators/test/test_bandits.py#L123 after replacing default output (when there is no data or something like 0/0) from 0 to None. "floats are non comparable to None" => all examples that are pushed to estimators have p_pred ==0. Most likely it means that the test is not testing what it is supposed to test and passed by accident before. Do you see what can be wrong there?