Command-Line Reference

This page describes the command-line arguments and environment variables used by ap_verify.

Signature and syntax

The basic call signature of ap_verify is:

python ap_verify.py --dataset DATASET --output OUTPUTREPO --id DATAID

These three arguments (or replacing --output with --rerun) are mandatory, all others are optional.

Status code

ap_verify returns a status code of 0 if the pipeline ran to completion. If the pipeline fails, the status code will be an interpreter-dependent nonzero value.

Named arguments

Required arguments are --dataset, --id, and exactly one of --output or --rerun.

--id <dataId>

Butler data ID.

The input data ID is required for all ap_verify runs except when using --help or --version.

Specify data ID to process using data ID syntax. For example, --id "visit=12345 ccd=1 filter=g".

Currently this argument is heavily restricted compared to its command line task counterpart. In particular, the dataId must specify exactly one visit and exactly one CCD, and may not be left blank to mean “all data”.

--dataset <dataset_name>

Input dataset designation.

The input dataset is required for all ap_verify runs except when using --help or --version.

The argument is a unique name for the dataset, which can be associated with a repository in the configuration file. See Datasets as Input Arguments for more information on dataset names.

Allowed names can be queried using the --help argument.

-h, --help

Print help.

The help is equivalent to this documentation page, describing command-line arguments.

-j <processes>, --processes <processes>

Number of processes to use.

When processes is larger than 1 the pipeline may use the Python multiprocessing module to parallelize processing of multiple datasets across multiple processors.

Note

This option is provided for forward-compatibility, but is not yet supported by ap_pipe.

--output <output_repo>

Output data repository URI or path.

The output dataset or --rerun is required for all ap_verify runs except when using --help or --version.

The output data repository will be created if it does not exist. The path may be absolute or relative to the current working directory.

--output may not be used with the --rerun argument.

See command-line-task-data-repo-howto for background.

--rerun <output>

Specify output “rerun”.

The rerun or --output is required for all ap_verify runs except when using --help or --version.

For ap_verify, a rerun is an output repository relative to the dataset directory (as determined by --dataset). This is different from command-line task reruns, which have an input repository and chain the rerun to it. An input rerun cannot be specified.

--rerun may not be used with the --output argument.

--silent

Do not report measurements to SQuaSH.

Disables upload of measurements, so that ap_verify can be run for testing purposes by developers.

Note

Ingestion of lsst.verify metrics is not yet supported by SQuaSH, so this flag should always be provided for now.

--version

Print version number.

Since ap_verify is not yet officially part of the Stack, the version number is arbitrary.

Environment variables

The SQUASH_USER, SQUASH_PASSWORD, and SQUASH_URL environment variables are used by the verify framework to configure SQuaSH upload. SQUASH_USER and SQUASH_PASSWORD must be defined in any environment where ap_verify is run unless the --silent flag is used.

SQUASH_USER

User name to use for SQuaSH submissions.

SQUASH_PASSWORD

Unencrypted password for SQUASH_USER.

SQUASH_URL

The location for a SQuaSH REST API. Defaults to the SQuaSH server at lsst.codes.