# ask-ai

R Lucas

02/27/2024, 7:38 AM
Is asset check a replacement of the current Dagster type to validate an asset output ?


02/27/2024, 7:38 AM
No, asset checks are not a replacement for Dagster's type system used to validate asset outputs. Asset checks are a complementary feature that allows you to define additional data quality checks on your assets. According to the documentation, each Asset Check verifies some property of a Dagster data asset, such as ensuring that there are no null values in a particular column or that the data conforms to a specific format. Dagster's type system is used to ensure that the outputs of computations match expected types, which is a form of compile-time validation. Asset checks, on the other hand, are runtime validations that can be used to enforce data quality constraints and can be executed either inline with the data pipeline or independently. Asset checks provide a way to communicate data contracts, execute and track computations that verify data properties, and they can be integrated with tools like dbt tests. They are part of a broader system that includes asset definitions, materializations, and check executions, which together provide a reliable picture of what to expect from an asset. In summary, asset checks are an additional layer of validation that works alongside Dagster's type system to ensure data quality and integrity.