Reproducibility is the ability to obtain the same result by replaying the same process: same data, same code, same library versions and same random seeds. It is achieved by pinning dependencies, versioning data and code, and seeding random generators.
Example
RandomForestClassifier(n_estimators=100, random_state=42)