TimingMetricTask¶

class lsst.ap.verify.measurements.TimingMetricTask(config=None, name=None, parentTask=None, log=None)¶

Bases: lsst.verify.compatibility.MetricTask

A Task that measures a timing metric using metadata produced by the lsst.pipe.base.timeMethod decorator.

Parameters:	args kwargs Constructor parameters are the same as for `lsst.verify.compatibility.MetricTask`.

Methods Summary

`adaptArgsAndRun`(inputData, inputDataIds, …)	Compute a metric from in-memory data.
`addStandardMetadata`(measurement, outputDataId)	Add data ID-specific metadata required for all metrics.
`emptyMetadata`()	Empty (clear) the metadata for this Task and all sub-Tasks.
`getAllSchemaCatalogs`()	Get schema catalogs for all tasks in the hierarchy, combining the results into a single dict.
`getFullMetadata`()	Get metadata for all tasks.
`getFullName`()	Get the task name as a hierarchical name including parent task names.
`getInputDatasetTypes`(config)	Return input dataset types for this task.
`getName`()	Get the name of the task.
`getOutputMetricName`(config)	Identify the metric calculated by this `MetricTask`.
`getSchemaCatalogs`()	Get the schemas generated by this task.
`getTaskDict`()	Get a dictionary of all tasks as a shallow copy.
`makeField`(doc)	Make a `lsst.pex.config.ConfigurableField` for this task.
`makeSubtask`(name, **keyArgs)	Create a subtask as a new instance as the `name` attribute of this task.
`run`(metadata)	Compute a wall-clock measurement from metadata provided by `lsst.pipe.base.timeMethod`.
`timer`(name[, logLevel])	Context manager to log performance data for an arbitrary block of code.

Methods Documentation

adaptArgsAndRun(inputData, inputDataIds, outputDataId)¶

Compute a metric from in-memory data.

Parameters:	inputData : `dict` from `str` to any Dictionary whose keys are the names of input parameters and values are Python-domain data objects (or lists of objects) retrieved from data butler. Accepting lists of objects is strongly recommended; this allows metrics to vary their granularity up to the granularity of the input data without the need for extensive code changes. Input objects may be `None` to represent missing data. inputDataIds : `dict` from `str` to `list` of dataId Dictionary whose keys are the names of input parameters and values are data IDs (or lists of data IDs) that the task consumes for corresponding dataset type. Data IDs are guaranteed to match data objects in `inputData`. outputDataId : `dict` from `str` to dataId Dictionary containing a single key, `"measurement"`, which maps to a single data ID for the measurement. The data ID must have the same granularity as the metric.
Returns:	struct : `lsst.pipe.base.Struct` A `Struct` containing at least the following component: `measurement`: the value of the metric identified by `getOutputMetricName`, computed from `inputData` (`lsst.verify.Measurement` or `None`). The measurement is guaranteed to contain not only the value of the metric, but also any mandatory supplementary information.
Raises:	lsst.verify.MetricComputationError Raised if an algorithmic or system error prevents calculation of the metric. Examples include corrupted input data or unavoidable exceptions raised by analysis code. The `MetricComputationError` should be chained to a more specific exception describing the root cause. Not having enough data for a metric to be applicable is not an error, and should not trigger this exception.

Notes

This implementation calls run on the contents of inputData, followed by calling addStandardMetadata on the result before returning it. Any subclass that overrides this method must also call addStandardMetadata on its measurement before returning it.

adaptArgsAndRun and run should assume they take multiple input datasets, regardless of the expected metric granularity. This rule may be broken if it is impossible for more than one copy of a dataset to exist.

All input data must be treated as optional. This maximizes the MetricTask’s usefulness for incomplete pipeline runs or runs with optional processing steps. If a metric cannot be calculated because the necessary inputs are missing, the MetricTask must return None in place of the measurement.

Examples

Consider a metric that characterizes PSF variations across the entire field of view, given processed images. Then, if run has the signature run(images):

inputData = {'images': [image1, image2, ...]}
inputDataIds = {'images': [{'visit': 42, 'ccd': 1},
                           {'visit': 42, 'ccd': 2},
                           ...]}
outputDataId = {'measurement': {'visit': 42}}
result = task.adaptArgsAndRun(
    inputData, inputDataIds, outputDataId)

addStandardMetadata(measurement, outputDataId)¶

Add data ID-specific metadata required for all metrics.

This method currently does not add any metadata, but may do so in the future.

Parameters:	measurement : `lsst.verify.Measurement` The `Measurement` that the metadata are added to. outputDataId : `dataId` The data ID to which the measurement applies, at the appropriate level of granularity.

Notes

This method must be called by any subclass that overrides adaptArgsAndRun, but should be ignored otherwise. It should not be overridden by subclasses.

This method is not responsible for shared metadata like the execution environment (which should be added by this MetricTask’s caller), nor for metadata specific to a particular metric (which should be added when the metric is calculated).

Warning

This method’s signature will change whenever additional data needs to be provided. This is a deliberate restriction to ensure that all subclasses pass in the new data as well.

emptyMetadata()¶: Empty (clear) the metadata for this Task and all sub-Tasks.

getAllSchemaCatalogs()¶

Get schema catalogs for all tasks in the hierarchy, combining the results into a single dict.

Returns:	schemacatalogs : `dict` Keys are butler dataset type, values are a empty catalog (an instance of the appropriate lsst.afw.table Catalog type) for all tasks in the hierarchy, from the top-level task down through all subtasks.

Notes

This method may be called on any task in the hierarchy; it will return the same answer, regardless.

The default implementation should always suffice. If your subtask uses schemas the override Task.getSchemaCatalogs, not this method.

getFullMetadata()¶

Get metadata for all tasks.

Returns:	metadata : `lsst.daf.base.PropertySet` The `PropertySet` keys are the full task name. Values are metadata for the top-level task and all subtasks, sub-subtasks, etc..

Notes

The returned metadata includes timing information (if @timer.timeMethod is used) and any metadata set by the task. The name of each item consists of the full task name with . replaced by :, followed by . and the name of the item, e.g.:

topLevelTaskName:subtaskName:subsubtaskName.itemName

using : in the full task name disambiguates the rare situation that a task has a subtask and a metadata item with the same name.

getFullName()¶

Get the task name as a hierarchical name including parent task names.

Returns:	fullName : `str` The full name consists of the name of the parent task and each subtask separated by periods. For example: The full name of top-level task “top” is simply “top”. The full name of subtask “sub” of top-level task “top” is “top.sub”. The full name of subtask “sub2” of subtask “sub” of top-level task “top” is “top.sub.sub2”.

classmethod getInputDatasetTypes(config)¶

Return input dataset types for this task.

Parameters:	config : `cls.ConfigClass` Configuration for this task.
Returns:	metadata : `dict` from `str` to `str` Dictionary from `"metadata"` to the dataset type of the target task’s metadata.

getName()¶

Get the name of the task.

Returns:	taskName : `str` Name of the task.

See also

Task.getAllSchemaCatalogs

Notes

Warning

Subclasses that use schemas must override this method. The default implemenation returns an empty dict.

This method may be called at any time after the Task is constructed, which means that all task schemas should be computed at construction time, not when data is actually processed. This reflects the philosophy that the schema should not depend on the data.

Returning catalogs rather than just schemas allows us to save e.g. slots for SourceCatalog as well.

getTaskDict()¶

Get a dictionary of all tasks as a shallow copy.

Returns:	taskDict : `dict` Dictionary containing full task name: task object for the top-level task and all subtasks, sub-subtasks, etc..

classmethod makeField(doc)¶

Make a lsst.pex.config.ConfigurableField for this task.

Parameters:	doc : `str` Help text for the field.
Returns:	configurableField : `lsst.pex.config.ConfigurableField` A `ConfigurableField` for this task.

Examples

Provides a convenient way to specify this task is a subtask of another task.

Here is an example of use:

class OtherTaskConfig(lsst.pex.config.Config)
    aSubtask = ATaskClass.makeField("a brief description of what this task does")

makeSubtask(name, **keyArgs)¶

Create a subtask as a new instance as the name attribute of this task.

Parameters:	name : `str` Brief name of the subtask. keyArgs Extra keyword arguments used to construct the task. The following arguments are automatically provided and cannot be overridden: “config”. “parentTask”.

Notes

The subtask must be defined by Task.config.name, an instance of pex_config ConfigurableField or RegistryField.

run(metadata)¶

Compute a wall-clock measurement from metadata provided by lsst.pipe.base.timeMethod.

Parameters:	metadata : iterable of `lsst.daf.base.PropertySet` A collection of metadata objects, one for each unit of science processing to be incorporated into this metric. Its elements may be `None` to represent missing data.
Returns:	result : `lsst.pipe.base.Struct` A `Struct` containing the following component: `measurement`: the total running time of the target method across all elements of `metadata` (`lsst.verify.Measurement` or `None`)
Raises:	MetricComputationError Raised if any of the timing metadata are invalid.

Notes

This method does not return a measurement if any element of metadata is None. The reason for this policy is that if a science processing run was aborted without writing metadata, then any timing measurement cannot be compared to other results anyway. This method also does not return a measurement if no timing information was provided by any of the metadata.

timer(name, logLevel=10000)¶

Context manager to log performance data for an arbitrary block of code.

Parameters:	name : `str` Name of code being timed; data will be logged using item name: `Start` and `End`. logLevel A `lsst.log` level constant.

Navigation

TimingMetricTask¶