n_jobs and random_state? #30811Description
Discussed in #30809
Originally posted by adosar February 11, 2025
In Controlling randomness, the guide is discussing how to properly control randomness either for an estimator or CV or when using both. However, there is no mention if random_state and n_jobs > 1 interact in any unexpected way.
Lets consider a typical use case where a user cross validates a RandomForestClassifier with KFold:
estimator = RandomForestClassifer(random_state=np.random.RandomState(1)) # Recommended to pass RandomState instance.
kfold = KFold(shuffle=True, random_state=42) # Recommended to pass int.
cross_val_score(estimator, n_jobs=-1, ..., cv=kfold)Since n_jobs=-1 this means that multiple cores will be used for cross validation (e.g. 1 core per fold).
Would the same state be used for the different folds, since during multiprocessing the estimator and hence the rng passed to it, is copied via fork?