Profile Hopsworks Feature Views and generate EDA reports.
npx skills add https://github.com/logicalclocks/hopsworks-api --skill hops-eda
Please help me install this Agent Skill. Skill: hops-eda Source: https://github.com/logicalclocks/hopsworks-api/tree/main/skills/ml/hops-eda Command: npx skills add https://github.com/logicalclocks/hopsworks-api --skill hops-eda
This Skill streamlines exploratory data analysis (EDA) for Hopsworks Feature Views, enabling users to efficiently profile feature-view training data, check for data leakage, and generate EDA reports.
Run the hops-eda skill with a feature view 'example-fv', by specifying the feature view name and optionally a version and time range:
python3 ~/scripts/fv-eda.py example-fv
High-intent search queries and answers about installing and using this skill.
You can perform exploratory data analysis on a Hopsworks Feature View by running a Python script that profiles feature data, identifies data leakage, and generates reports to streamline machine learning workflows.
Data leakage detection in feature views identifies unintended information in training data that biases machine learning models. Checking for leakage before training ensures reliable model evaluation and prevents inflated performance metrics.
Yes, exploratory data analysis for Hopsworks supports both Polars and PySpark, allowing you to profile feature data and check for data leakage across different dataset sizes and processing requirements.
To generate an EDA report for a Hopsworks feature view version, execute the analysis script by specifying the feature view name along with optional version and time range parameters to profile the targeted training data.
You should run data profiling before training a machine learning model to analyze feature characteristics, identify potential data issues, and check for data leakage. This ensures clean inputs and improves model reliability.
logicalclocksTrain ML models with feature views and register them in the Hopsworks model registry.
andikarachmanDetect temporal, direct, and group data leakage in machine-learning feature sets.
logicalclocksCreate and schedule feature pipelines in the Hopsworks Feature Store using Python.
Mekann2904Perform exploratory data analysis on over 200 scientific file formats and generate markdown reports.