Feast Ride Sharing Use Case

Provides an end-to-end workflow using Feast in HPE Ezmeral Unified Analytics Software to generate training data and perform online model inference for the ride-sharing driver satisfaction model.

Prerequisites

  • Sign in to HPE Ezmeral Unified Analytics Software.

About this task

Use Feast to generate training data and perform online model inference for the ride-sharing driver satisfaction model.

In this tutorial, you will:

  1. Deploy a local feature store with a Parquet file offline store and SQLite online store.
  2. Build a training dataset using time series features from Parquet files.
  3. Read the latest features from the offline store for batch scoring.
  4. Ingest batch features ("materialization") and streaming features into the online store.
  5. Read the latest features from the online store for real-time inference.
  6. Explore the Feast web interface to see Data Sources, Entities, Feature Views, Feature Services, and Datasets that are defined through feature definitions.

Procedure

  1. Connect to the notebook server. See Creating and Managing Notebook Servers.
  2. In the <username>/Feast folder, open the ride-sharing-example.ipynb file.
    NOTE

    If you do not see the Feast folder in the <username> folder, copy the folder from the shared/ezua-tutorials/current-release/Data-Science/ directory into the <username> folder. The shared directory is accessible to all users. Editing or running examples from the shared directory is not advised. The <username> directory is specific to you and cannot be accessed by other users.

    If the Feast folder is not available in the shared/ezua-tutorials/current-release/Data-Science/ directory:

    1. Go to GitHub repository for tutorials.
    2. Clone the repository.
    3. Navigate to ezua-tutorials/Data-Science.
    4. Navigate back to the shared directory.
    5. Copy the /Feast folder from the ezua-tutorials/Data-Science repository into the shared directory.
    6. Copy the /Feast folder from the shared folder to the <username> directory.
  3. Validate the ride-sharing-example.ipynb file, definitions.py file, and the data folder are available in the /<username>/Feast directory.
  4. Validate the driver_stats.parquet file is available in the <username>/Feast/data directory.
  5. Select the first cell of the ride-sharing-example.ipynb notebook, and click Run Selected Cell and All Below.

Results

  1. Click the Tools & Frameworks icon on the left navigation bar. Navigate to the Feast tile under the Data Science tab and click Open.
  2. Explore the Feast web interface to see Data Sources, Entities, Feature Views, Feature Services, and Datasets that are defined through feature definitions.