extract-test-set

To extract vault price data for pytest, run the extraction script targeting a specific chain and contract address. It generates a dedicated test module and a parquet file containing the raw price dataframe needed for deterministic unit tests.

A parquet file stores the extracted raw price dataframe for a specific vault test case. It provides a static, deterministic dataset that enables isolated and repeatable unit testing of vault price logic without hitting live databases.

Yes, extracting price data for a vault test case requires access to the DEFAULT_UNCLEANED_PRICE_DATABASE. The extraction script queries this database to pull the raw price dataframe for the specified chain and contract address.

Yes, you can generate a pytest module directly from a vault contract address. The Skill uses an ad-hoc script to target the given chain and address, producing both the test module and the corresponding parquet price data file.

The best way to create deterministic unit tests for vault price logic is to extract raw price data into a parquet file. This decouples tests from live database fluctuations by providing a fixed dataframe aligned to the specific vault test case.