Best Practices for edaflow

  • Always verify your data quality before starting ML workflows

  • Use setup_ml_experiment for reproducible train/val/test splits

  • Compare multiple models before optimizing hyperparameters

  • Use copy-paste-safe param_distributions blocks for supported models

  • Save model artifacts and document your workflow for reproducibility

  • Refer to the User Guide and API Reference for troubleshooting and advanced usage

Best Practices for New Features

  • Use display_facet_grid for multi-category visual analysis

  • Apply scale_features before ML modeling for better results

  • Use group_rare_categories to simplify categorical variables

  • Export figures with export_figure for reproducible reporting

  • Always check external library requirements before using advanced features