Usage Guide

This guide describes both basic and advanced Modin usage, including usage examples, details regarding Modin configuration settings, as well as tips and tricks on how you can further optimize the performance of your workload with Modin.

• Modin Configuration Settings
  • Public API
    • EnvironmentVariable
      • EnvironmentVariable.get()
      • EnvironmentVariable.get_help()
      • EnvironmentVariable.get_value_source()
      • EnvironmentVariable.once()
      • EnvironmentVariable.put()
      • EnvironmentVariable.subscribe()
  • Modin Configs List
  • Usage Guide
• Modin Usage Examples
  • Tutorials
  • Data Science Benchmarks
  • Modin Spreadsheets
  • Modin with scikit-learn
• Advanced Usage
  • Modin engines
  • Additional APIs
  • DataFrame partitioning API
  • Modin Spreadsheet API
  • Progress Bar
  • Dataframe Algebra
  • SQL on Modin Dataframes
  • Distributed XGBoost on Modin
  • Logging with Modin
  • Batch Pipeline API
  • Fuzzydata Testing
• Optimization Notes
  • Range-partitioning in Modin
  • Understanding Modin’s partitioning mechanism
    • How Modin partitions a dataframe
    • Full-axis functions
    • How to tune partitioning
      • Configure Modin’s default partitioning scheme
      • Manually trigger repartitioning
  • Avoid iterating over Modin DataFrame
  • Use Modin’s Dataframe Algebra API to implement custom parallel functions
  • Avoid mixing pandas and Modin DataFrames
  • Operation-specific optimizations
    • merge
• Benchmarking Modin
  • Summary
  • Modin’s execution and benchmark mode
  • A caveat about benchmark mode
  • How to benchmark complex workflows
  • Appendix: System Information
• Third Party Library Integrations
  • Modin Interoperability by Library
  • A Deeper Dive
    • seaborn
    • plotly
    • matplotlib
    • altair
    • bokeh
    • sklearn
    • Hugging Face
    • Tensorflow
    • NLTK
    • XGBoost
    • statsmodels
  • Appendix: System Information