ci_lsstcam provides test scripts to run the LSST Science
Pipelines on LSSTCam data. As of Aug 20 2025, this package only runs at the
USDF where the test data is available at /sdf/group/rubin/shared/data/test_data. Work is underway to enable running on Jenkins.
ci_lsstcam requires access to the test data in the testdata_ci_lsstcam_m49 and testdata_ci_imsim directories (as the source of the pretrained models collection) at the USDF.
To run this package at the USDF, follow these steps:
1) Clone this package, ci_builder, and
ci_lsstcam.
2) setup -r ci_builder
3) setup -kr ci_lsstcam
4) From the root of the ci_lsstcam directory run bin/rewrite.sh to rewrite python shebang lines.
5) Run bin/ci_lsstcam_run.py. See available options with --help, as steps may be run individually.
To cleanup after a run, use either bin/ci_lsstcam_run.py --clean or rm -rf DATA/.
Note that there are 20 detector visits across 4 (ugri) bands and 1 patch in testdata_ci_lsstcam. Thus, running with up to -j 20 will speed up visit-level processing. Single-band coadd-level processing will benefit from up to -j 4.