H2O AutoDoc Python Documentation

This document provides an overview of the H2O AutoDoc Python package’s usages and API.

Requirements

• Python 3.6 or 3.7

• H2O AutoDoc Python client

Have Questions?

If you have questions about using H2O AutoDoc, post them on the H2O.ai Community Slack workspace in the #autodoc channel. If you have not signed up for the H2O.ai Community Slack workspace, you can do so here: https://www.h2o.ai/community/.

H2O AutoDoc Python Module

• Introduction to H2O AutoDoc
  • What is H2O AutoDoc
  • Supported Software Details
  • Dataset Dependent Sections
  • Supported File Types
  • Change Log
• Installing H2O AutoDoc
  • H2O AutoDoc Requirements
  • H2O AutoDoc Setup & Installation
  • H2O AutoDoc License Key
• Creating & Configuring H2O AutoDoc
  • Building an H2O Model
  • Generate a Default H2O AutoDoc
  • Set the H2O AutoDoc File Type
  • Model Interpretation Dataset
  • Partial Dependence Features
  • Specify ICE Records
  • Enable/Disable Shapley Values
  • Provide Additional Testsets
  • Provide Alternative Models
• H2O AutoDoc API Docs
  • Configure H2O AutoDoc
  • Render the H2O AutoDoc Report

Steam

• Using H2O AutoDoc with Steam
  • Introduction
  • Configure H2O AutoDoc
    • Connecting to H2O-3 Cluster
    • Connecting to Sparkling Water Cluster
    • Build H2O Model on connected cluster (H2O-3 / Sparkling Water)
    • Generate and Download a Default H2O AutoDoc
    • Set the H2O AutoDoc File Type
    • Model Interpretation Dataset
    • Partial Dependence Features
    • Specify ICE Records
    • Enable/Disable Shapley Values
    • Saving AutoDoc to HDFS