You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
It looks like other than the notebooks, none of the tutorials (demos) are currently tested and thus likely to break. Currently at least the demo_2d_tracer.py is broken (missing x, y = SpatialCoordinate(mesh) line. I know some of them are (more or less) duplicated in the examples/ but that's not a reliable way to ensure they are still runnable. As I don't wont to add significantly to the CI time - and adding if os.getenv('THETIS_REGRESSION_TEST'): logic would be confusing/ugly - maybe we can preprocess the scripts before running in CI, or reversely strip out the logic when building the documentation.
The text was updated successfully, but these errors were encountered:
In the mesh adaptation packages, @ddundo added some neat functionality for setting parameters when running as tests that avoids the THETIS_REGRESSION_TEST approach. Instead, the modifications are defined in the test_demos.py script. See (for example), https://github.com/mesh-adaptation/animate/blob/main/test/test_demos.py.
It looks like other than the notebooks, none of the tutorials (demos) are currently tested and thus likely to break. Currently at least the
demo_2d_tracer.py
is broken (missingx, y = SpatialCoordinate(mesh)
line. I know some of them are (more or less) duplicated in the examples/ but that's not a reliable way to ensure they are still runnable. As I don't wont to add significantly to the CI time - and addingif os.getenv('THETIS_REGRESSION_TEST'):
logic would be confusing/ugly - maybe we can preprocess the scripts before running in CI, or reversely strip out the logic when building the documentation.The text was updated successfully, but these errors were encountered: