PlasmaCalcs.tools.multiprocessing.TaskGroup

class PlasmaCalcs.tools.multiprocessing.TaskGroup(tasks, *, i=None, errors_ok=(), result_missing=RESULT_MISSING)

treats a dict of Tasks as if it was a single Task.

Helps with coarsen. Users probably don’t need to think about TaskGroup directly.

tasks: dict of Tasks errors_ok: tuple of Exception types

the exception types for which to return result_missing instead of crashing, in case of error.

result_missing: any object: result to record for tasks which crash (if errors_ok!=False). Default RESULT_MISSING.

__init__(tasks, *, i=None, errors_ok=(), result_missing=RESULT_MISSING): should set self.data = something related to stuff.

Methods

Attributes

apply_async(pool, *, kw={})

perform the task asynchronously, i.e. pool.apply_async(…)

Equivalent to pool.apply_async(self.f, args=self.args, kwds=self.kw)

kw: dict: kwargs for task will be task.kw, but updated with kw.

E.g. if task.kw = {‘x’: 1}, and kw = {‘y’: 2}, –> task called with x=1, y=2.

enumerate(idx=None): iterate through i in idx, yielding (i, self[i]) pairs.

If idx is None, iterate through all objs in self (see self._enumerate_all).

f(*args__None, **kw): perform all tasks in self, returning the results (as a dict)

get_kw(kw={}): gets self.kw, but updated with values from kw. (doesn’t alter self.kw)

[EFF] if kw is empty, return self.kw instead of making a copy.

classmethod implied(task_info, *, force_kw=True, **kw_init)

return the Task implied from task_info.

task_info: Task, iterable, or callable.: Task –> return task_info, unchanged (unless force_kw and len(kw_init)>0).

if task_info is not an instance of cls but it is a Task, raise TypeError.

iterable –> return Task(*task_info).

callable –> return Task(task_info).
force_kw: bool: applies if kw_init provided and task_info is a Task…

True –> return task_info.new_with(**kw_init)

False –> return task_info, unchanged.

kw_init: pass to Task(…, **kw_init) if task_info is not already a Task.

new_empty(fill=UNSET): return a new dict of the same shape as self, filled with the value fill.

new_with(f=UNSET, args=UNSET, kw=UNSET, i=UNSET): return a new Task, with the given inputs. Use value from self for UNSET inputs.

E.g. Task(f, [5,6]).new_with(f=g, kw=dict(y=7)) –> Task(g, [5,6], dict(y=7))

size(idx=None): return the number of objects in the container, or in idx if provided.