README.md

Create new applications

Adding a new application outside of the core library

1. Define your computation as in example_application.py.
2. Modify the main json configuration file to include the application and provide a link to the local python file.
3. In the configuration file define the job in "jobs" section providing the path to the python file in "path" and the "file.module" name in "type". The file will be added to the tar file that contains the executables. In the "users" section add the new application and assign a probability of executing it.

{
  "project_name": "test",
  "connector": "NO_LINK",
  ...
  "jobs": [
    {
      "type": "VQE",
      "qubit_min": 2,
      "qubit_max": 4,
      "walltime": 3,
      "nthreads": 2,
      "lsf/jsrun_opt": "-nnodes=4"
    },
    {
      "type": "custom1.Custom1",
      "path": "tests/data/apps/custom1.py",
      "qubit_min": 2,
      "qubit_max": 4,
      "walltime": 3,
      "nthreads": 2,
      "slurm/schedmd_opt": "--cpus-per-task=4" 
    }
  ],
  "users": [
    {
      "user": "user0",
      "weight": 1,
      "computations": {
        "VQE": 0.50,
        "custom1.Custom1" : 0.50
      }
    }
  ]
} 

Adding a new application as part of the core library

1. Define your computation as in example_application.py. This is the src script of your application you want to run. Place this in the qstone/apps folder.
2. In qstone.apps.computation_registry add {"yourcomputationname" : ComputationClass}.
3. To use your computation simply use "yourcomputationname" when defining the configuration JSON for the generator.