Skip to content

Latest commit

 

History

History
42 lines (34 loc) · 2.33 KB

README.md

File metadata and controls

42 lines (34 loc) · 2.33 KB
Raw

ADCIRC Model Test Suite

The included suite of tests has been developed to test changes to the ADCIRC model and ensure solutions are consistent across releases. This suite is run any time a change is made to the model code and must pass before the code is allowed to be integrated into the upstream repository.

These also encompass a nice set of examples for users to get acquainted with setting up and running the model.

Running the Suite

Compiling for the suite

The automated test scripts may require the following ADCIRC executables to be compiled:

  1. adcirc
  2. padcirc
  3. adcswan
  4. padcswan
  5. adcprep

Depending on the modules used within the test suite, you may not need to compile all of these, though some tests may require certain features are enabled, such as netCDF or GRIB2 support.

Running Tests

The test suite uses a python test apparatus. This runner relies on a yaml file to define the testing metadata. The default test list is test_list.yaml and is located in the root directory fo this repository. It is possible to leverage your own tests by modifying this file or creating a new one. The test suite produces a series of plots as well to aid diagnosing any issues that may arise.

To run the full set of tests, the following command is used:

python3 test_runner/test_runner.py --all --bin <path/to/adcirc/build> --test-yaml test_list.yaml --tolerance 0.00001 --test-root .

You can also run a single test by specifying the test name:

python3 test_runner/test_runner.py --test <name_of_test> --bin <path/to/adcirc/build> --test-yaml test_list.yaml --tolerance 0.00001 --test-root .

Note that the default testing tolerance is 0.00001. This can be adjusted by changing the --tolerance flag. It may be useful to do so depending on your build settings and compiler.

Submitting a new case

New cases are definitely welcomed. Anyone looking to submit a new case should follow one of the other directories as an example of how a case should be constructed. Cases should exercise a feature or combination of features that is not currently part of the test suite. Cases should be able to be run in under 5 minutes in serial to avoid the suite becoming too cumbersome to run. Cases should be submitted either to [email protected] or via Pull Request with a description of the case and what it aims to test.