Primrose

Welcome to primrose! Primrose is a Python framework for executing in-memory workflows–typically machine learning—defined by directed acyclic graphs (DAG) via configuration files.

Contents:

• Overview
  • Primrose at a glance
  • Table of Contents
  • Introduction
  • Getting Started
  • Next
  • License
  • Contributing
• Getting Started
  • Your first primrose job
  • A Machine Learning Job: Clustering
  • A Machine Learning Job with Model Caching: Classification
  • Next
• DAG Configurations
• Configuration Files
  • Section Order
  • Node Definitions
• Comments
  • YAML
  • JSON
• Configuration File Fragment Substitution
• Environment Variable Substitution
• Configuration File Validation
  • When Validation Passes
  • When Validation Fails
  • Next
• Configuration Metadata
• traverser key
• section_registry key
• section_run key
  • section_run / traverser compatability checking
• data_object key
  • section by section development workflow
  • Next
• Primrose Command Line Interface
  • Getting help
  • primrose run
  • primrose create_project
  • primrose validate
  • primrose plot
  • Other commands
  • Specifying your primrose node registry
  • Next
• Developer Notes
  • necessary_config(node_config)
  • run(data_object)
  • Complete example: sleep
  • Registering Your Classes
  • Conditional Pathing
  • Next
• DataObject
• Saving Data
  • Specifying a key
• Getting Data
  • Specifying form of data
• Other methods
  • Next
• Notifications
  • DAG Notifications
  • Error Notifications
  • Environment Variables

Indices and tables

• Index

• Module Index

• Search Page