ProtocolGraph¶

class openff.evaluator.workflow.ProtocolGraph[source]¶

A graph of connected protocols which may be executed together.

__init__()[source]¶: Initialize self. See help(type(self)) for accurate signature.

Methods

`__init__`()	Initialize self.
`add_protocols`(*protocols[, …])	Adds a set of protocols to the graph.
`execute`([root_directory, …])	Execute the protocol graph in the specified directory, and either using a CalculationBackend, or using a specified set of compute resources.

Attributes

`protocols`	The protocols in this graph.
`root_protocols`	The ids of the protocols in the group which do not take input from the other grouped protocols.

property protocols¶

The protocols in this graph.

Type: dict of str and Protocol

property root_protocols¶

The ids of the protocols in the group which do not take input from the other grouped protocols.

Type: list of str

add_protocols(*protocols, allow_external_dependencies=False)[source]¶

Adds a set of protocols to the graph.

Parameters

protocols (tuple of Protocol) – The protocols to add.
allow_external_dependencies (bool) – If False, an exception will be raised if a protocol has a dependency outside of this graph.

Returns

A mapping between the original protocols and protocols which were merged over the course of adding the new protocols.

Return type

dict of str and str

execute(root_directory='', calculation_backend=None, compute_resources=None, enable_checkpointing=True, safe_exceptions=True)[source]¶

Execute the protocol graph in the specified directory, and either using a CalculationBackend, or using a specified set of compute resources.

Parameters

root_directory (str) – The directory to execute the graph in.
calculation_backend (CalculationBackend, optional.) – The backend to execute the graph on. This parameter is mutually exclusive with compute_resources.
compute_resources (CalculationBackend, optional.) – The compute resources to run using. This parameter is mutually exclusive with calculation_backend.
enable_checkpointing (bool) – If enabled, protocols will not be executed more than once if the output from their previous execution is found.
safe_exceptions (bool) – If true, exceptions will be serialized into the results file rather than directly raised, otherwise, the exception will be raised as normal.

Returns

The paths to the JSON serialized outputs of the executed protocols. If executed using a calculation backend, these will be Future objects which will return the output paths on calling future.result().

Return type

dict of str and str or Future