BaseOptimizer

class openff.bespokefit.optimizers.BaseOptimizer[source]

Bases: ABC

This is the abstract basic BaseOptimizer class that each optimizer should use.

Methods

`__init__`()
`deregister_target`(target_name)	Remove a registered target from the optimizer.
`description`()	Returns a friendly description of the optimizer.
`get_registered_targets`()	Internal method to get the registered targets for this specific optimizer.
`is_available`()	This method should check that installation requirements are met before trying to run the optimizer.
`name`()	Returns the friendly name of the optimizer.
`optimize`(schema, initial_force_field[, ...])	This is the main function of the optimizer which is called when the optimizer is put in a workflow.
`prepare`(schema, initial_force_field, ...)	Prepares the optimization by creating any required inputs and setting up the required environment.
`provenance`()	This function should detail the programs with the version information called during the running od the optimizer.
`register_target`(target_type[, replace])	Take a target and register it with the optimizer under an alias name which is used to call the target.

classmethod deregister_target(target_name: str)[source]

Remove a registered target from the optimizer.

Parameters: target_name (str) – The name of the target that should be removed.
Raises: TargetRegisterError – If no target is registered under the name to be removed.

abstract classmethod description() → str[source]: Returns a friendly description of the optimizer.

classmethod get_registered_targets() → Dict[str, Type[BaseTargetSchema]][source]: Internal method to get the registered targets for this specific optimizer.

abstract classmethod is_available() → bool[source]: This method should check that installation requirements are met before trying to run the optimizer.

abstract classmethod name() → str[source]: Returns the friendly name of the optimizer.

classmethod optimize(schema: OptimizationStageSchema, initial_force_field: ForceField, root_directory: Optional[str] = None) → OptimizationStageResults[source]

This is the main function of the optimizer which is called when the optimizer is put in a workflow.

It should loop over the targets and assert they are registered and then dispatch compute and optimization.

classmethod prepare(schema: OptimizationStageSchema, initial_force_field: ForceField, root_directory: str)[source]: Prepares the optimization by creating any required inputs and setting up the required environment.

abstract classmethod provenance() → Dict[source]

This function should detail the programs with the version information called during the running od the optimizer.

Returns: A dictionary containing the information about the optimizer called.

classmethod register_target(target_type: Type[BaseTargetSchema], replace: bool = False) → None[source]

Take a target and register it with the optimizer under an alias name which is used to call the target.

Parameters

target_type – The type of target to be registered with the optimizer.
replace – If the alias is already registered replaced with the new target data with no exception.

Raises

TargetRegisterError – If the target has already been registered.

abstract classmethod name() → str[source]: Returns the friendly name of the optimizer.

abstract classmethod description() → str[source]: Returns a friendly description of the optimizer.