Design a feature store that can handle multiple data sources and provide efficient access to features for machine learning models. **Example:** Consider a scenario where you have two data sources: a customer database and a transaction log. You want to store customer demographics and transaction history as features in your feature store. **Constraints:** The feature store should be able to handle millions of rows of data and provide features in a format suitable for model training.
Test Cases
Test Case 1
Input:
{"customer_database": {"name": "customer_database", "features": [[1, 2, 3], [4, 5, 6]]}, "transaction_log": {"name": "transaction_log", "features": [[7, 8, 9], [10, 11, 12]]}}Expected:
{"customer_database": [[1, 2, 3], [4, 5, 6]], "transaction_log": [[7, 8, 9], [10, 11, 12]]}Test Case 2
Input:
{"customer_database": {"name": "customer_database", "features": [[13, 14, 15], [16, 17, 18]]}, "transaction_log": {"name": "transaction_log", "features": [[19, 20, 21], [22, 23, 24]]}}Expected:
{"customer_database": [[13, 14, 15], [16, 17, 18]], "transaction_log": [[19, 20, 21], [22, 23, 24]]}+ 3 hidden test cases