lsst.ap.verify

The lsst.ap.verify package provides an executable python program for pipeline verification. It runs the alert production pipeline (encapsulated in the lsst.ap.pipe package), computes lsst.verify metrics on both the pipeline’s state and its output, and works with the SQuaSH system to allow their monitoring and analysis.

ap_verify is designed to work with small, standardized datasets that can be interchanged to test the Stack’s performance under different conditions. To ensure consistent results, it runs the entire AP pipeline as a single unit, from data ingestion to source association. It produces measurements, using the lsst.verify framework, that can be used for both monitoring stack development and debugging failure cases.

Using lsst.ap.verify

• Running ap_verify from the command line
• The dataset framework
• Error-handling in ap_verify and failed runs
• ap_verify command-line reference
• ap_verify configuration file reference

Contributing

lsst.ap.verify is developed at https://github.com/lsst/ap_verify. You can find Jira issues for this module under the ap_verify component.

• Adding new metrics to ap_verify

Python API reference

lsst.ap.verify Package

Functions

runApVerify([cmdLine])	Execute the AP pipeline while handling metrics.
runIngestion([cmdLine])	Ingest a dataset, but do not process it.

Classes

DatasetIngestConfig

Settings and defaults for DatasetIngestTask.

lsst.ap.verify.measurements Package

Functions

measureFromButlerRepo(butler, dataId)

Create measurements from a butler repository.

Classes

TimingMetricConfig	Information that distinguishes one timing metric from another.
TimingMetricTask([config, name, parentTask, log])	A Task that measures a timing metric using metadata produced by the lsst.pipe.base.timeMethod decorator.