TaskRunner

class lsst.pipe.base.TaskRunner(TaskClass, parsedCmd, doReturnResults=False)

Bases: object

Run a command-line task, using multiprocessing if requested.

Parameters:
TaskClass : lsst.pipe.base.Task subclass

The class of the task to run.

parsedCmd : argparse.Namespace

The parsed command-line arguments, as returned by the task’s argument parser’s parse_args method.

Warning

Do not store parsedCmd, as this instance is pickled (if multiprocessing) and parsedCmd may contain non-picklable elements. It certainly contains more data than we need to send to each instance of the task.

doReturnResults : bool, optional

Should run return the collected result from each invocation of the task? This is only intended for unit tests and similar use. It can easily exhaust memory (if the task returns enough data and you call it enough times) and it will fail when using multiprocessing if the returned data cannot be pickled.

Note that even if doReturnResults is False a struct with a single member “exitStatus” is returned, with value 0 or 1 to be returned to the unix shell.

Raises:
ImportError

If multiprocessing is requested (and the task supports it) but the multiprocessing library cannot be imported.

Notes

Each command-line task (subclass of lsst.pipe.base.CmdLineTask) has a task runner. By default it is this class, but some tasks require a subclass. See the manual Creating a command-line task for more information. See CmdLineTask.parseAndRun to see how a task runner is used.

You may use this task runner for your command-line task if your task has a runDataRef method that takes exactly one argument: a butler data reference. Otherwise you must provide a task-specific subclass of this runner for your task’s RunnerClass that overrides TaskRunner.getTargetList and possibly TaskRunner.__call__. See TaskRunner.getTargetList for details.

This design matches the common pattern for command-line tasks: the runDataRef method takes a single data reference, of some suitable name. Additional arguments are rare, and if present, require a subclass of TaskRunner that calls these additional arguments by name.

Instances of this class must be picklable in order to be compatible with multiprocessing. If multiprocessing is requested (parsedCmd.numProcesses > 1) then runDataRef calls prepareForMultiProcessing to jettison optional non-picklable elements. If your task runner is not compatible with multiprocessing then indicate this in your task by setting class variable canMultiprocess=False.

Due to a python bug, handling a KeyboardInterrupt properly requires specifying a timeout. This timeout (in sec) can be specified as the timeout element in the output from ArgumentParser (the parsedCmd), if available, otherwise we use TaskRunner.TIMEOUT.

By default, we disable “implicit” threading – ie, as provided by underlying numerical libraries such as MKL or BLAS. This is designed to avoid thread contention both when a single command line task spawns multiple processes and when multiple users are running on a shared system. Users can override this behaviour by setting the LSST_ALLOW_IMPLICIT_THREADS environment variable.

Attributes Summary

TIMEOUT Default timeout (seconds) for multiprocessing.

Methods Summary

__call__(args) Run the Task on a single target.
getTargetList(parsedCmd, **kwargs) Get a list of (dataRef, kwargs) for TaskRunner.__call__.
makeTask([parsedCmd, args]) Create a Task instance.
precall(parsedCmd) Hook for code that should run exactly once, before multiprocessing.
prepareForMultiProcessing() Prepare this instance for multiprocessing
run(parsedCmd) Run the task on all targets.
runTask(task, dataRef, kwargs) Make the actual call to runDataRef for this task.

Attributes Documentation

TIMEOUT = 2592000

Default timeout (seconds) for multiprocessing.

Methods Documentation

__call__(args)

Run the Task on a single target.

Parameters:
args

Arguments for Task.runDataRef()

Returns:
struct : lsst.pipe.base.Struct

Contains these fields if doReturnResults is True:

  • dataRef: the provided data reference.
  • metadata: task metadata after execution of run.
  • result: result returned by task run, or None if the task fails.
  • exitStatus: 0 if the task completed successfully, 1 otherwise.

If doReturnResults is False the struct contains:

  • exitStatus: 0 if the task completed successfully, 1 otherwise.

Notes

This default implementation assumes that the args is a tuple containing a data reference and a dict of keyword arguments.

Warning

If you override this method and wish to return something when doReturnResults is False, then it must be picklable to support multiprocessing and it should be small enough that pickling and unpickling do not add excessive overhead.

static getTargetList(parsedCmd, **kwargs)

Get a list of (dataRef, kwargs) for TaskRunner.__call__.

Parameters:
parsedCmd : argparse.Namespace

The parsed command object returned by lsst.pipe.base.ArgumentParser.parse_args.

kwargs

Any additional keyword arguments. In the default TaskRunner this is an empty dict, but having it simplifies overriding TaskRunner for tasks whose runDataRef method takes additional arguments (see case (1) below).

Notes

The default implementation of TaskRunner.getTargetList and TaskRunner.__call__ works for any command-line task whose runDataRef method takes exactly one argument: a data reference. Otherwise you must provide a variant of TaskRunner that overrides TaskRunner.getTargetList and possibly TaskRunner.__call__. There are two cases.

Case 1

If your command-line task has a runDataRef method that takes one data reference followed by additional arguments, then you need only override TaskRunner.getTargetList to return the additional arguments as an argument dict. To make this easier, your overridden version of getTargetList may call TaskRunner.getTargetList with the extra arguments as keyword arguments. For example, the following adds an argument dict containing a single key: “calExpList”, whose value is the list of data IDs for the calexp ID argument:

def getTargetList(parsedCmd):
    return TaskRunner.getTargetList(
        parsedCmd,
        calExpList=parsedCmd.calexp.idList
    )

It is equivalent to this slightly longer version:

@staticmethod
def getTargetList(parsedCmd):
    argDict = dict(calExpList=parsedCmd.calexp.idList)
    return [(dataId, argDict) for dataId in parsedCmd.id.idList]

Case 2

If your task does not meet condition (1) then you must override both TaskRunner.getTargetList and TaskRunner.__call__. You may do this however you see fit, so long as TaskRunner.getTargetList returns a list, each of whose elements is sent to TaskRunner.__call__, which runs your task.

makeTask(parsedCmd=None, args=None)

Create a Task instance.

Parameters:
parsedCmd

Parsed command-line options (used for extra task args by some task runners).

args

Args tuple passed to TaskRunner.__call__ (used for extra task arguments by some task runners).

Notes

makeTask can be called with either the parsedCmd argument or args argument set to None, but it must construct identical Task instances in either case.

Subclasses may ignore this method entirely if they reimplement both TaskRunner.precall and TaskRunner.__call__.

precall(parsedCmd)

Hook for code that should run exactly once, before multiprocessing.

Notes

Must return True if TaskRunner.__call__ should subsequently be called.

Warning

Implementations must take care to ensure that no unpicklable attributes are added to the TaskRunner itself, for compatibility with multiprocessing.

The default implementation writes package versions, schemas and configs, or compares them to existing files on disk if present.

prepareForMultiProcessing()

Prepare this instance for multiprocessing

Optional non-picklable elements are removed.

This is only called if the task is run under multiprocessing.

run(parsedCmd)

Run the task on all targets.

Parameters:
parsedCmd : argparse.Namespace

Parsed command argparse.Namespace.

Returns:
resultList : list

A list of results returned by TaskRunner.__call__, or an empty list if TaskRunner.__call__ is not called (e.g. if TaskRunner.precall returns False). See TaskRunner.__call__ for details.

Notes

The task is run under multiprocessing if TaskRunner.numProcesses is more than 1; otherwise processing is serial.

runTask(task, dataRef, kwargs)

Make the actual call to runDataRef for this task.

Parameters:
task : lsst.pipe.base.CmdLineTask class

The class of the task to run.

dataRef

Butler data reference that contains the data the task will process.

kwargs

Any additional keyword arguments. See TaskRunner.getTargetList above.

Notes

The default implementation of TaskRunner.runTask works for any command-line task which has a runDataRef method that takes a data reference and an optional set of additional keyword arguments. This method returns the results generated by the task’s runDataRef method.