Tests
We use pytest to implement tests. From the repository’s root directory you can do
> pytest .
==================================== test session starts ====================================
platform darwin -- Python 3.11.8, pytest-8.1.1, pluggy-1.4.0
rootdir: /Users/evanberkowitz/physics/supervillain
configfile: pyproject.toml
plugins: anyio-4.3.0, cov-4.1.0
collected 962 items
test/test_compare-reference-observables.py .......................................... [ 4%]
...... [ 4%]
test/test_ensemble-extension.py ................................................ [ 9%]
test/test_equivalence-class-v.py .................................................... [ 15%]
..................................................................................... [ 24%]
..................................................................................... [ 33%]
..................................................................................... [ 41%]
..................................................................................... [ 50%]
........................................ [ 54%]
test/test_gauge-invariance.py ....................................................... [ 60%]
..................................................................................... [ 69%]
..................................................................................... [ 78%]
..................................................................................... [ 87%]
..ssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssss........... [ 95%]
..................................... [ 99%]
test/test_harness.py sx [100%]
================== 888 passed, 73 skipped, 1 xfailed in 290.62s (0:04:50) ===================
Some tests are skipped on purpose (for example, if an action does not have a particular observable implemented) or xfailed purposefully; only actual failures should be of concern. You can get a more detailed report with -v
.
At first the tests may seem very slow. For meaningful tests different ensembles are generated. However, these ensembles are cached, so that the tests are slow to start but go faster as later Monte Carlo is not needed.