libkalibera

Overview

libkalibera contains reimplementations of the statistical computations for benchmarking evaluation from the following two papers:

• Rigorous benchmarking in reasonable time. Tomas Kalibera, Richard Jones
• Quantifying performance changes with effect size confidence intervals. Tomas Kalibera, Richard Jones

libkalibera started off as a pure Python module, but support for other languages has since been added. Currently supported languages are:

• Python (Edd Barrett, Carl Friedrich Bolz)
• Ruby (Chris Seaton)

Research Deploying LibKalibera

The following academic papers used libkalibera:

• Approaches to Interpreter Composition. Edd Barrett, Carl Friedrich Bolz, Laurence Tratt (to appear). Preprint available.
• Dynamically Composing Languages in a Modular Way: Supporting C Extensions for Dynamic Languages. Matthias Grimmer, Chris Seaton, Thomas Wuerthinger, Hanspeter Moessenboeck. Modularity 2015 (to appear).

Download

Up to date versions can be found in the GitHub repository.