Benchmark Clojure code with statistically robust timing accounting for JIT warmup and GC.
npx skills add https://github.com/hugoduncan/criterium --skill criterium
Please help me install this Agent Skill. Skill: criterium Source: https://github.com/hugoduncan/criterium/tree/main/skills/criterium Command: npx skills add https://github.com/hugoduncan/criterium --skill criterium
Benchmarking Clojure code can be unreliable without accounting for JVM warmup, garbage collection, and measurement overhead. Criterium provides a principled approach to obtaining stable timing data and meaningful statistics.
Install the library, load the bench namespace, and run a simple benchmark to observe timing results.
High-intent search queries and answers about installing and using this skill.
To benchmark Clojure code accurately, you need a tool that accounts for JVM warmup, garbage collection, and measurement overhead. Criterium uses a three-stage pipeline—collection, analysis, and view—to produce statistically robust, repeatable timing data.
Comparing Clojure implementations requires statistically robust timing that handles warmup and GC. Criterium enables comparative analysis through multiple bench plans and viewers, allowing you to measure and contrast execution times reliably across different code paths.
You can run parameterized benchmarks in Clojure using domain benchmarking and argument generation. Criterium supports arg-gen and test-check generators to execute bench plans across specified input ranges, yielding performance analysis for each parameter.
Yes, effective performance tuning with Criterium requires knowledge of its API, specifically the bench macro, bench plans, and arg-gen. This understanding allows you to leverage its three-stage pipeline for meaningful statistical analysis during code profiling.
You can visualize Clojure profiling data using various viewers in Criterium. The tool's view stage supports distribution analysis and histograms, presenting the processed timing data from the bench plans in formats suitable for human or programmatic consumption.
Clojure benchmark timings are often unstable due to JVM warmup, garbage collection, and measurement overhead. Criterium addresses this by applying a principled approach to timing data gathering, ensuring stable measurements and meaningful statistics for performance tuning.
lespacemanAggregate WebBench results from JSONL into a markdown report.
sorodriguezzExplain methods for defining and managing table statistics in InterSystems SQL.
llmh333Create Go unit tests with table-driven cases and subtests.
joaquimscostaDiagnose Spring Boot 4.x migration gaps in Maven and Gradle projects.
