Automate Hopsworks feature view creation and management with the Python SDK.
npx skills add https://github.com/logicalclocks/hopsworks-api --skill hops-fv
Please help me install this Agent Skill. Skill: hops-fv Source: https://github.com/logicalclocks/hopsworks-api/tree/main/skills/hops/hops-fv Command: npx skills add https://github.com/logicalclocks/hopsworks-api --skill hops-fv
This Skill addresses the complex challenge of managing Hopsworks feature views, simplifying the process of creating, querying, and serving feature data for machine learning projects.
To get started, execute the following command: create_feature_view("my_fv", query=my_query, labels=["is_fraud"], inference_helper_columns=["merchant"]).
High-intent search queries and answers about installing and using this skill.
To create a Hopsworks feature view, use the Python SDK to define a query, select features, and assign labels like `is_fraud`. You can execute this via the `create_feature_view` command, passing the query and specifying inference helper columns for real-time serving.
Automating data transformations for feature views involves using the Hopsworks Python SDK to configure metadata and apply transformation functions directly within the feature view creation process. This streamlines feature engineering and prepares data for both training and real-time serving.
Yes, Hopsworks feature view management handles both online and offline data processing. The Python SDK automates the querying and serving of feature data, ensuring seamless transitions between batch training preparation and real-time model serving requirements.
Yes, you can combine user and transaction data in a single feature view by constructing a unified query in the Hopsworks Python SDK. This allows you to apply transformations and configure metadata for the combined dataset before model training and serving.
Yes, the Hopsworks SDK and a Python environment are required to automate the creation and management of feature views. These dependencies are necessary to handle feature selection, transformations, and the serving of online and offline data.
logicalclocksDeploys ML models and runs online inference in Hopsworks via Python SDK.
robotijnAudit feature store configurations for training-serving skew and data leakage.
logicalclocksQuery Hopsworks feature store tables using Trino SQL.
logicalclocksAutomate batch inference in Hopsworks using Python or PySpark.
