IMP logo
IMP Reference Guide  2.22.0
The Integrative Modeling Platform
IMP.parallel.Context Class Reference

A collection of tasks that run in the same environment. More...

Detailed Description

A collection of tasks that run in the same environment.

Context objects are typically created by calling Manager::get_context().

Note
This class is only available in Python.

Definition at line 330 of file parallel/__init__.py.

Public Member Functions

def __init__
 Constructor. More...
 
def add_task
 Add a task to this context. More...
 
def get_results_unordered
 Run all of the tasks on available workers, and return results. More...
 

Constructor & Destructor Documentation

def IMP.parallel.Context.__init__ (   self,
  manager,
  startup = None 
)

Constructor.

Definition at line 334 of file parallel/__init__.py.

Member Function Documentation

def IMP.parallel.Context.add_task (   self,
  task 
)

Add a task to this context.

Tasks are any Python callable object that can be pickled (e.g. a function or a class that implements the __call__ method). When the task is run on the worker its arguments are the return value from this context's startup function.

Definition at line 341 of file parallel/__init__.py.

def IMP.parallel.Context.get_results_unordered (   self)

Run all of the tasks on available workers, and return results.

If there are more tasks than workers, subsequent tasks are started only once a running task completes: each worker only runs a single task at a time. As each task completes, the return value(s) from the task callable are returned from this method, as a Python generator. Note that the results are returned in the order that tasks complete, which may not be the same as the order they were submitted in.

Exceptions
NoMoreWorkersErrorthere are no workers available to run the tasks (or they all failed during execution).
RemoteErrora worker encountered an unhandled exception.
NetworkErrorthe manager lost (or was unable to establish) communication with any worker.

Definition at line 357 of file parallel/__init__.py.


The documentation for this class was generated from the following file: