Metrics is a Python library for making rEproducible sofTware peRformance analysIs in perfeCt Simplicity.
You can find its source code on GitHub, read its documentation on Read the Docs and install it with pip:
pip3 install crillab-metrics
This repository contains various use cases of Metrics, that are distributed for demonstration purposes. You may access them through the following table of contents:
- Analysis of different solver competitions
- Analysis of the performance of COP solvers (using various types of files)
- Analysis of the performance of CSP solvers (using XML data files)
- Analysis of the performance of MIP solvers (using raw output files)
- Analysis of the performance of PB solvers (using raw output files)
- Analysis of the performance of SAT solvers (campaign in custom CSV files)
- Analysis of the performance of subgraph solvers (campaign in reverse CSV format)