Model Parameters Pipeline

A Python package for applying sequential data transformations according to the Model Parameters specification developed by Big Life Lab.

Supported Transformations

• Center – Subtract a specified value from a variable

• Dummy – Create binary indicator variables for categorical values

• Interaction – Multiply variables together to create interaction terms

• Restricted Cubic Splines (RCS) – Create spline basis functions with specified knots

• Logistic Regression – Apply logistic regression coefficients and the sigmoid function

Contents

• Introduction
  • Overview
  • What is the Model Parameters Specification?
  • Supported Transformations
  • Installation
  • Basic Usage
  • Using DataFrames for Input Data
  • Processing Multiple Datasets
  • Method Chaining
  • Restricting File Access with sandbox_path
  • Working with Results
  • Real-World Example: HTNPoRT Model
  • Next Steps
  • Additional Resources
• Adding a New Transformation Step
  • Overview
  • Step 1: Register the Step in _STEP_DISPATCH
  • Step 2: Create the Step Function
  • Step 3: Add Unit Tests
  • Reference Documentation
  • Checklist
  • Common Patterns
  • Getting Help
• Model Parameters Step Tests
  • Directory Structure
  • Adding a New Unit Test
  • Model Parameters Steps Reference
  • Regenerating Expected Output
• API Reference
  • Pipeline
  • Utilities
  • Transformation Steps