parsl.app.app.python_app¶
-
parsl.app.app.
python_app
(function=None, data_flow_kernel: Optional[parsl.dataflow.dflow.DataFlowKernel] = None, cache: bool = False, executors: Union[List[str], typing_extensions.Literal[all]] = 'all', ignore_for_cache: Optional[List[str]] = None, join: bool = False)[source]¶ Decorator function for making python apps.
- Parameters
function (function) – Do not pass this keyword argument directly. This is needed in order to allow for omitted parenthesis, for example,
@python_app
if using all defaults or@python_app(walltime=120)
. If the decorator is used alone, function will be the actual function being decorated, whereas if it is called with arguments, function will be None. Default is None.data_flow_kernel (DataFlowKernel) – The
DataFlowKernel
responsible for managing this app. This can be omitted only after callingparsl.dataflow.dflow.DataFlowKernelLoader.load()
. Default is None.executors (string or list) – Labels of the executors that this app can execute over. Default is ‘all’.
cache (bool) – Enable caching of the app call. Default is False.
join (bool) – If True, this app will be a join app: the decorated python code must return a Future (rather than a regular value), and and the corresponding AppFuture will complete when that inner future completes.