search

ReactionMechanismGenerator / RMG-tests

Continous Integration Testing Platform for RMG-Py
3 stars 14 forks source link
integration-testing regression-testing

readme

This git repository tracks changes in RMG model generation by retaining a history of RMG generated models in CHEMKIN format.

Normally RMG-tests run automatically all the examples registered in examples folder on Travis platform. For convenience of debugging and running customized jobs (which sometimes violate Travis's restricted rule of memory usage and CPU time), we added the functionality of running RMG-tests locally. To run locally, please follow the instruction below.

Instructions for simultaneous updates to RMG-Py and RMG-database

If you need an RMG-Py pull request to be tested on a certain branch of the RMG-database, add a commit that modifies the file .github/workflows/CI.yml in RMG-Py by changing the line RMG_DB_BRANCH: main to instead state the desired database branch.

Likewise, if you wish to test an RMG-database update using a specific branch of RMG-Py, then add a commit to the modify the file .github/workflows/CI.yml in your RMG-database changing the line RMG_PY_BRANCH: main to instead state the desired RMG-Py branch.

After the tests pass, remove the commits before merging the pull request to main.

Warning

Avoid hyphens in your branch names, on either RMG-database or RMG-Py.

Instructions for Running RMG-tests Locally

  1. Clone the repo down to your local machine: run git clone https://github.com/ReactionMechanismGenerator/RMG-tests.git

  2. Make sure your have installed anaconda (please skip this step if you have already installed). If you are going to use it on a server, you probably are using the global anaconda which means you usually don't have have permission to write into /path/to/anaconda/envs. In this case, you need to install your own local anaconda, so follow the commands below:

    # Set up anaconda
wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh -O miniconda.sh
chmod +x miniconda.sh
./miniconda.sh -b -p $HOME/miniconda
export PATH=$HOME/miniconda/bin:$PATH

# Update conda itself
conda update --yes conda

  3. Modify local_tests/submit_serial.sl by specifying the branch name of RMG-Py and RMG-database. Note that these branches must be already pushed and located on github.com/ReactionMechanismGenerator. Also change the job name to the one you want to run. If you need to test customized jobs, add a new folder with RMG input file which you want to run to examples/rmg/.

  4. Run RMG-tests by 

    cd local_tests
bash submit_serial.sl

    If you are running on a server using SLURM for submitting jobs, we've provided two example submission scripts (serial and parallel). For serial mode of testing:

    For parallel mode of testing:

  5. You'll find the test log in folder RMG-tests/tests/check/${you job name}, and two versions of RMG-generated CHEMKIN models in folder RMG-tests/tests/benchmark/${you job name} and RMG-tests/tests/testmodel/${you job name} for detailed analysis.

  6. We're also testing recording test results into a central mongo database. Please contact rmg_dev@mit.edu to get an authentication file (in config.cfg format) to access to database. Currently it's required to run thermochemistry validation test.