Manually Run a PrimitiveΒΆ
TODO: Describe how you can manually run a primitive: create an instance, call methods. Document temporary directory, random seed, and static files.
Version
- This version: 2022.5.6.dev0
- Other versions
Table of Contents
Getting Started
- Get a Docker Image
- Get Datasets
- Get and Run a Pipeline
- Core Package Installation
- Primitives Installation
Guides
Advanced Topics
- Metadata
- Metalearning Database
- Advanced Primitives
- Advanced Pipelines
- Structure of Pipeline Predictions
- Data Preparation and Scoring Pipelines
HOWTOs
- Load Datasets and Problems
- Wrap a Sklearn-compatible Primitive
- Write a Good Primitive
- Reproduce Pipeline Run
- Primitive with Static Files
- Register a Custom Metric
- Register a Custom Problem Type
- Write Tests for Primitives
- Debug a Primitive or Pipeline
- Create a Primitive Based on an Existing Primitive
- Primitives Installation
- Register a Primitive
- Blocklist Primitives
- Manually Run a Primitive
- Build and Extend a Docker Image
- Use Jupyter/Colab
Contribute
Reference