Extending the ALD

For basic extension, add a conditional statement to helpers.create_and_launch_job, i.e.:
if args["job_system"] == "slurm":

    # Do what is currently implemented

elif args["job_system"] == "your_new_scheduler_name":

    # Use current implementation as template and modify for your scheduler

else:

    print "ERROR: Unrecognized scheduler: ", args["job_system"]
    exit()
and make similar edits to helpers.wait_for_job and helpers.wait_for_jobs. Search for “srun” in *.py files to check for compatibility with your system. When running, set HPC_SYSTEM=your_new_scheduler_name

To add support for cluster-based active learning, additional utilities/new*sh files will need to be added and properly selected for in cluster.py

Adding a new labeling method

To add a new labeling method, edit the following scripts:

main.py
qm_driver.py
verify_config.py

Additionally, create a new Python script to be read by qm_driver.py to perform the relabeling. Follow the same code structure as existing *_driver.py scripts e.g. vasp_driver.py. Update verify_config.py and main.py so that your new driver’s required variables are defined, briefly explained, and assigned default values.

To commit the changes, you must provide a minimum working example using your labeling method and a successful run through the current example list for acceptance.

Adding a new MD method

To add a new MD method, edit the following scripts:

main.py
run_md.py
verify_config.py

Create a new Python script to be read by run_md.py to perform the MD. Follow the same code structure as other *_run_md.py scripts. Update verify_config.py and main.py so that all required variables for your new driver are defined, briefly explained, and assigned default values.

To commit the changes, provide a minimum working example using your MD method and ensure it passes the current example list.