roastcoffea documentation

roastcoffea provides comprehensive performance metrics collection for Coffea workflows running on Dask. Track throughput, resource usage, and fine-grained profiling data without modifying your analysis code.

Features

  • Workflow metrics: Throughput, event rates, resource utilization

  • Chunk tracking: Per-chunk performance with @track_metrics decorator

  • Fine-grained profiling: Section-level timing with track_time() and track_memory()

  • Dask Spans integration: Separate processor work from Dask overhead

  • Worker monitoring: Time-series resource tracking

  • Export options: Rich tables, JSON measurements

Getting started

🚀 Quickstart

Quickstart - Install and run your first metrics collection in minutes.

📖 Tutorial

Tutorial - Step through examples covering all collection levels.

💡 Concepts

Core Concepts - Understand what metrics mean and how they’re calculated.

📊 Metrics Reference

Performance Metrics Reference - Complete catalog of available metrics.