Databricks Machine Learning Associate Question 76
Select 3You are tasked with creating a feature store table in Databricks to standardize and share features for a machine learning project. Which of the following steps are required to successfully create a feature store table?
- A
Define a feature table schema using a Spark DataFrame.
- B
Use the
FeatureStoreClient.create_tablemethod to register the feature table. - C
Manually write the feature table data to a Delta table without using the Feature Store API.
- D
Provide a unique name for the feature table in the format 'database_name.table_name'.
- E
Train a machine learning model before creating the feature table.
Show answer and explanation
Correct answers: A, B, D
Explanation
Creating a feature store table in Databricks involves defining the schema using a Spark DataFrame, registering the table with the Feature Store API via the FeatureStoreClient.create_table method, and providing a unique name for the table. Writing data manually without the Feature Store API or training a model beforehand is not required.
- A. Correct.
Correct. A Spark DataFrame is required to define the schema and data for the feature table before registering it in the feature store.
- B. Correct.
Correct. The
FeatureStoreClient.create_tablemethod is used to register the feature table and associate it with the feature store. - C. Incorrect.
Incorrect. Writing data manually to a Delta table without the Feature Store API does not integrate the table with Databricks Feature Store.
- D. Correct.
Correct. A unique name in the format 'database_name.table_name' is required to register the feature table in the feature store.
- E. Incorrect.
Incorrect. Training a machine learning model is not a prerequisite for creating a feature table in the feature store.