User Guide

This guide provides comprehensive documentation for using mice-py effectively in your data analysis workflows.

• Understanding Missing Data
  • Checking for Missing Data
  • Visualizing Missing Patterns
  • Types of Missing Data
  • When to Use MICE
  • Next Steps
• MICE Overview
  • What is MICE?
  • How MICE Works
  • Basic Usage
  • Controlling the Imputation Process
  • Accessing Results
  • When MICE Works Well
  • Limitations of MICE
  • The Imputation Model vs Analysis Model
  • Typical Workflow
  • Next Steps
• Imputation Methods
  • Overview of Methods
  • PMM: Predictive Mean Matching
  • CART: Classification and Regression Trees
  • Random Forest
  • MIDAS: Multiple Imputation with Distant Average Substitution
  • Sample: Random Sampling
  • Choosing a Method
  • Using Different Methods for Different Variables
  • Comparing Methods
  • Research Findings
  • Next Steps
• Predictor Matrices
  • What is a Predictor Matrix?
  • Default Behavior
  • When to Customize
  • Creating Custom Predictor Matrices
  • Common Patterns
  • Quickpred: Automatic Predictor Selection
  • Monotone Patterns
  • Practical Examples
  • Checking Your Predictor Matrix
  • Common Issues
  • Tips and Best Practices
  • Testing Your Choices
  • Next Steps
• Convergence Diagnostics
  • What is Convergence?
  • Chain Statistics
  • Visualizing Convergence
  • Numerical Assessment
  • What to Do If Not Converged
  • How Many Iterations?
  • Convergence by Variable
  • Other Diagnostic Checks
  • Common Convergence Issues
  • Best Practices
  • Quick Convergence Checklist
  • Next Steps
• Pooling Analysis
  • Why Pool Results?
  • Rubin’s Rules
  • Using mice-py for Pooling
  • Understanding the Output
  • Fraction of Missing Information (FMI)
  • Formula Syntax
  • Advanced Pooling
  • Interpreting Pooled Results
  • How Many Imputations?
  • Common Pitfalls
  • Checking Results
  • Tips for Better Pooling
  • Next Steps
• Best Practices
  • Basic Setup
  • Recommended Parameters
  • Method Selection
  • Always Check Convergence
  • Compare Distributions
  • Proper Pooling
  • Common Mistakes
  • Variable Selection
  • Predictor Matrix
  • Performance Tips
  • Essential Checklist
  • See Also

Overview

The user guide covers all aspects of using mice-py, from understanding missing data to producing final pooled results. Each section builds on the previous ones, but you can also jump directly to topics of interest.

Understanding Missing Data

Learn about missing data mechanisms, patterns, and why they matter for your analysis.

MICE Overview

Understand how the MICE algorithm works and when to use it.

Imputation Methods

Detailed guide to all available imputation methods and how to choose the right one.

Predictor Matrices

Learn how to control which variables predict which other variables during imputation.

Convergence Diagnostics

Check if your imputation has converged and what to do if it hasn’t.

Pooling Analysis

Combine results from multiple imputed datasets using Rubin’s rules.

Best Practices

Tips, tricks, and common pitfalls to avoid when using MICE.

Getting Help

If you need help:

1. Check the relevant section in this user guide

2. Look at the Examples for practical examples

3. Consult the API Reference for detailed API documentation

4. Review the Theory & Background for theoretical background

5. Open an issue on GitHub