Personal sandbox code for me to exercise and experiment with github + associated features (e.g. travis-ci, waffle.io)
I'm hoping to find ways to elegantly add performance profiling and regression testing to the Continuous Integration process on github.
The current setup uses python as a (currently non-existent) codebase with a single dependency on numpy.
This sandbox exercises simple pytest tests using travis-ci.
There exists a cProfile plugin "pytest_profiling" that can be installed via
pip install pytest_profiling
However this plugin probably does not provide the flexibility we need to achieve the goals set out. For example, it currently changes the test result to issue a warning and I do not really want that. At this stage of the game however, it is functionally equivalent if not better than the basic profiler plugin written.