Processing of SweetSpot WIYN images using LSST DM Stack.
Requires the definition of the environment variable DR1BASE and the existence of SweetSpot stacks in ${DR1BASE}/stacks.
-
Create LSST-formatted FITS with image, weight map, and mask.
sh make_wiyn_dr1_lsst.sh
-
Ingest files
sh ingest_wiyn_dr1_images.sh
-
processCcd files
sh run_processCcd.sh
-
coadd files
sh run_make_all_skymaps.sh
sh run_all_coadds.sh` -
Photometry coadds
sh run_all_photometry.sh
-
Run lightcurves on calexps based on coadd detections
. run_all_calexp_lightcurves.sh
-
Run subtractions
python sub_wiyn_dr1.py
-
Run lightcurves on subtractions
python wiyn_forcedPhotExternalCatalog.py
Notes:
- Generate "dr1_dataid.list" after ingest step with
- sqlite3 registry.sqlite3 .separator ',' .output dr1_dataid.csv select field, seq, filter, night, expnum from stack; Then reprocess with cat dr1_dataid.csv | awk -F , '{printf "--id field=%s seq=%s filter=%s night=%s\n", $1, $2, $3, $4}' > dr1_dataid.list
- Generate "dr1_coadd.list" using cat dr1_dataid.csv | awk -F , '{print $2, $4, $3, $1}' | grep ' H ' | sort -k 4 | uniq -f 3 | awk '{print $4, NR-1, $1, $2}' > dr1_coadd.list
- Generate "dr1_coadd.list" for images actually successfully processed to calexp stage: ls ${REPO}/rerun/processCcd/calexp/bkgd*| cut -d / -f 11 | sed -e 's/bkgd-(.)([A-F])([JHKS])([0-9]{8})([0-9][0-9]*).fits/\2 \3 \4 \5 \1/' | sort -k 5 | uniq -f 4 | awk '{print $5, NR-1, $1, $2, $3}' > dr1_coadd_actual.list
The simplest job would just be to run the above commands in a single-node, serial job.
The slightly more efficient job spreads out the processing of the processCcd step, which is the the most expensive step.
- processCcd
sbatch run_processCcd_each_field_jobarray.sl
- Run subtractions
sbatch run_subtractions.sl
- Run lightcurves
sbatch run_forced_photometry.sl
Want to run LSST DM Stack on NERSC. The install from Twinkles is here:
https://github.com/mwvgroup/wiyn_lsst.git
The git issue discussing this install is LSSTDESC/Twinkles#242.