Benchmarking

This document outlines the current approach to benchmarking within PMMoTo using the pytest_benchmark plugin. Please note that this is a work in progress, and improvements are ongoing.

Overview

Benchmarking is essential for evaluating the performance of the PMMoTo framework. The pytest_benchmark plugin provides a simple way to measure the performance of your code.

Usage

To run benchmarks, you can create benchmark tests in your test files. Here is a basic example:

import pytest

@pytest.mark.benchmark
def test_my_function(benchmark):
    # ...existing code...
    result = benchmark(my_function, *args, **kwargs)
    # ...existing code...

Running Benchmarks

To execute your benchmarks, run the following command:

pytest benchmark/

Interpreting Results

The results will include various metrics such as:

Mean: The average time taken for the benchmarked function.
Standard Deviation: The variability of the benchmark results.
Minimum and Maximum: The fastest and slowest recorded times.

Future Improvements

As this benchmarking framework is still under development, future enhancements may include:

More detailed reporting options.
Integration with continuous integration tools for automated performance testing.
Additional benchmarks for various components of PMMoTo.

Conclusion

Benchmarking is a critical aspect of maintaining performance standards in PMMoTo. We encourage contributions and feedback as we continue to refine this process.