Example command-line task that computes simple statistics on an image. More...

Inheritance diagram for lsst.pipe.tasks.exampleCmdLineTask.ExampleCmdLineTask:

Public Member Functions
def	__init__

def	run
	Compute a few statistics on the image plane of an exposure. More...

Static Public Attributes
	ConfigClass = ExampleCmdLineConfig

Private Member Functions
def	_getConfigName
	Get the name prefix for the task config's dataset type, or None to prevent persisting the config. More...

def	_getMetadataName
	Get the name prefix for the task metadata's dataset type, or None to prevent persisting metadata. More...

Static Private Attributes
string	_DefaultName = "exampleTask"

Detailed Description

Example command-line task that computes simple statistics on an image.

Description

Example command-line task that computes simple statistics on an image.

This task was written as an example for the documents How to Write a Task and How to Write a Command-Line Task. The task reads in a "calexp" (a calibrated science exposure), computes statistics on the image plane, and logs and returns the statistics. In addition, if debugging is enabled, it displays the image in ds9.

The image statistics are computed using a subtask, in order to show how to call subtasks and how to retarget (replace) them with variant subtasks.

The main method is run.

Configuration parameters

See ExampleCmdLineConfig

Debug variables

This task supports the following debug variables:

display: If True then display the exposure in ds9

To enable debugging, see Using lsstDebug to control debugging output.

A complete example of using ExampleCmdLineTask

This code is in examples/exampleCmdLineTask.py, and can be run as follows:

 examples/exampleCmdLineTask.py $OBS_TEST_DIR/data/input --id
 # that will process all data; you can also try any combination of these flags:
 --id filter=g
 --config doFail=True --doraise
 --show config data

Definition at line 51 of file exampleCmdLineTask.py.

Constructor & Destructor Documentation

def lsst.pipe.tasks.exampleCmdLineTask.ExampleCmdLineTask.__init__	(	self,
		args,
		kwargs
	)

Construct an ExampleCmdLineTask

Call the parent class constructor and make the "stats" subtask from the config field of the same name.

Definition at line 104 of file exampleCmdLineTask.py.

 
     def __init__(self, *args, **kwargs):
         """Construct an ExampleCmdLineTask
 
         Call the parent class constructor and make the "stats" subtask from the config field of the same name.
         """
         pipeBase.CmdLineTask.__init__(self, *args, **kwargs)
         self.makeSubtask("stats")

lsst.pipe.tasks.exampleCmdLineTask.ExampleCmdLineTask.__init__

def __init__

Definition: exampleCmdLineTask.py:104

Member Function Documentation

def lsst.pipe.tasks.exampleCmdLineTask.ExampleCmdLineTask._getConfigName ( self )

private

Get the name prefix for the task config's dataset type, or None to prevent persisting the config.

This override returns None to avoid persisting metadata for this trivial task.

However, if the method returns a name, then the full name of the dataset type will be <name>_config. The default CmdLineTask._getConfigName returns _DefaultName, which for this task would result in a dataset name of "exampleTask_config".

Normally you can use the default CmdLineTask._getConfigName, but here are two reasons why you might want to override it:

If you do not want your task to write its config, then have the override return None. That is done for this example task, because I didn't want to clutter up the repository with config information for a trivial task.
If the default name would not be unique. An example is MakeSkyMapTask: it makes a sky map (sky pixelization for a coadd) for any of several different types of coadd, such as deep or goodSeeing. As such, the name of the persisted config must include the coadd type in order to be unique.

Normally if you override _getConfigName then you override _getMetadataName to match.

Definition at line 142 of file exampleCmdLineTask.py.

 
     def _getConfigName(self):
         """!Get the name prefix for the task config's dataset type, or None to prevent persisting the config
 
         This override returns None to avoid persisting metadata for this trivial task.
 
         However, if the method returns a name, then the full name of the dataset type will be <name>_config.
         The default CmdLineTask._getConfigName returns _DefaultName,
         which for this task would result in a dataset name of "exampleTask_config".
 
         Normally you can use the default CmdLineTask._getConfigName, but here are two reasons
         why you might want to override it:
         - If you do not want your task to write its config, then have the override return None.
           That is done for this example task, because I didn't want to clutter up the
           repository with config information for a trivial task.
         - If the default name would not be unique. An example is
           \ref lsst.pipe.tasks.makeSkyMap.MakeSkyMapTask "MakeSkyMapTask": it makes a
           \ref lsst.skymap.SkyMap "sky map" (sky pixelization for a coadd)
           for any of several different types of coadd, such as deep or goodSeeing.
           As such, the name of the persisted config must include the coadd type in order to be unique.
 
         Normally if you override _getConfigName then you override _getMetadataName to match.
         """
         return None

lsst.pipe.tasks.exampleCmdLineTask.ExampleCmdLineTask._getConfigName

def _getConfigName

Get the name prefix for the task config's dataset type, or None to prevent persisting the config...

Definition: exampleCmdLineTask.py:142

def lsst.pipe.tasks.exampleCmdLineTask.ExampleCmdLineTask._getMetadataName ( self )

private

Get the name prefix for the task metadata's dataset type, or None to prevent persisting metadata.

This override returns None to avoid persisting metadata for this trivial task.

However, if the method returns a name, then the full name of the dataset type will be <name>_metadata. The default CmdLineTask._getConfigName returns _DefaultName, which for this task would result in a dataset name of "exampleTask_metadata".

See the description of _getConfigName for reasons to override this method.

Definition at line 166 of file exampleCmdLineTask.py.

 
     def _getMetadataName(self):
         """!Get the name prefix for the task metadata's dataset type, or None to prevent persisting metadata
 
         This override returns None to avoid persisting metadata for this trivial task.
 
         However, if the method returns a name, then the full name of the dataset type will be <name>_metadata.
         The default CmdLineTask._getConfigName returns _DefaultName,
         which for this task would result in a dataset name of "exampleTask_metadata".
 
         See the description of _getConfigName for reasons to override this method.
         """
         return None

def lsst.pipe.tasks.exampleCmdLineTask.ExampleCmdLineTask.run	(	self,
		dataRef
	)

Compute a few statistics on the image plane of an exposure.

Parameters

dataRef: data reference for a calibrated science exposure ("calexp")

Returns

a pipeBase Struct containing:

mean: mean of image plane
meanErr: uncertainty in mean
stdDev: standard deviation of image plane
stdDevErr: uncertainty in standard deviation

Definition at line 113 of file exampleCmdLineTask.py.

 
     def run(self, dataRef):
         """!Compute a few statistics on the image plane of an exposure
         
         @param dataRef: data reference for a calibrated science exposure ("calexp")
         @return a pipeBase Struct containing:
         - mean: mean of image plane
         - meanErr: uncertainty in mean
         - stdDev: standard deviation of image plane
         - stdDevErr: uncertainty in standard deviation
         """
         self.log.info("Processing data ID %s" % (dataRef.dataId,))
         if self.config.doFail:
             raise pipeBase.TaskError("Raising TaskError by request (config.doFail=True)")
 
         # Unpersist the raw exposure pointed to by the data reference
         rawExp = dataRef.get("raw")
         maskedImage = rawExp.getMaskedImage()
 
         # Support extra debug output.
         # - 
         import lsstDebug
         display = lsstDebug.Info(__name__).display
         if display:
             frame = 1
             mtv(rawExp, frame=frame, title="exposure")
 
         # return the pipe_base Struct that is returned by self.stats.run
         return self.stats.run(maskedImage)