Details
-
Type:
RFC
-
Status: Implemented
-
Resolution: Done
-
Component/s: DM
-
Labels:None
Description
RFC for DM-7677: recommended changes to lsstSimMapper.paf file
Request to modify some templates and dataset names to make the obs_* Mapper classes more consistent.
DM-7049 was intended to make the Mapper datasets more consistent by centralizing datasets which could be shared across Mappers.
The latest HscMapper.paf was used as a model. Datasets were moved to datasets.yaml and exposures.yaml in daf_butlerUtils.
- Datasets in the individual Mapper.paf which matched the shared ones in butlerUtils were deleted. To avoid changing any datasets deletion was only done on those matched in both name and content. Those which matched in name but not content are listed (1) below.
- Datasets were also flagged when it appeared that they could be shareable, but in fact were not in HscMapper at all. These are candidates to be renamed and moved to daf_butlerUtils. Some of these could also be datasets no longer in use.
Please also see RFC-231 for datasets which are specifically connected to Coadd and Difference imaging. They were omitted from (2).
-------------------------------------
1. Datasets with same name but different content in lsstSimMapper.paf
The difference is shown between the existing paf dataset and the
proposed change to the "ref": the dataset is defined in butlerUtils.
datasets.deepCoadd_calexp_background
|
template deepCoadd-results/%(filter)s/%(tract)d/%(patch)s/b
|
kgd-%(filter)s-%(tract)d-%(patch)s.fits} }
|
ref:template deepCoadd-results/%(filter)s/%(tract)d/%(patch)s/d
|
et_bkgd-%(filter)s-%(tract)d-%(patch)s.fits
|
datasets.deepCoadd_forced_metadata
|
template deepCoadd_forced_metadata/%(filter)s/%(tract)d/%(p
|
atch)s.boost
|
ref:template deepCoadd_forcedPhotCoadd_metadata/%(filter)s/%(tr
|
act)d/%(patch)s.boost
|
datasets.deepCoadd_forced_config
|
python lsst.meas.base.forcedCoadd.ForcedCoaddMeasurementC
|
onfig
|
ref:python lsst.meas.base.forcedPhotCoadd.ForcedPhotCoaddConf
|
ig
|
template config/deepCoadd_forced.py
|
ref:template config/forcedPhotCoadd.py
|
datasets.deepCoadd_forced_src
|
template deepCoadd-results/%(filter)s/%(tract)d/%(patch)s/f
|
orced-%(filter)s-%(tract)d-%(patch)s.fits
|
ref:template deepCoadd-results/%(filter)s/%(tract)d/%(patch)s/f
|
orced_src-%(filter)s-%(tract)d-%(patch)s.fits
|
datasets.deep_safeClipAssembleCoadd_metadata
|
template deep_safeClipAssembleCoadd_metadata/%(filter)s/%(t
|
ract)d/%(patch)s.boost
|
ref:template deep_assembleCoadd_metadata/%(filter)s/%(tract)d/%
|
(patch)s.boost
|
datasets.forced_src_schema
|
template schema/forced.fits
|
ref:template schema/forced_src_schema.fits
|
-------------------------------------
2. Datasets which should be examined to see if they should be
removed, renamed, or moved to dafButlerUtils.
Note mappers with the same dataset name are shown:
l=lsstSim,d=decam,s=suprimecam,S=sdss,m=megacam,t=test
datasets.badSourceHist: Slt
|
datasets.badSource_schema: Slt
|
datasets.badSource: Slt
|
datasets.cal_ref_cat: lt
|
datasets.invalidSource_schema: Slt
|
datasets.invalidSource: Slt
|
datasets.measureCcd_config: lt
|
datasets.measureCcd_metadata: lt
|
datasets.modelfits: mlt
|
datasets.modelfits_schema: mlt
|
datasets.Mosaic_metadata: lt
|
datasets.multiband_config: smSl
|
datasets.object_schema: Slt
|
datasets.object: Slt
|
datasets.refcat: lt
|
datasets.sourceAssoc_config: Slt
|
datasets.sourceAssoc_metadata: Slt
|
datasets.sourceHist: Slt
|
datasets.source_schema: Slt
|
datasets.source: Slt
|
exposures.dcrModel: l
|
Attachments
Issue Links
Activity
This is an incorrect persistable setting in lsstSim:
forcedPhotCoaddConfig is the correct class for deepCoadd_forced_config in lsstSim Template may be wrong too, but it is somewhat dangerous to correct so I will not.
Perry Gee
added a comment - This is an incorrect persistable setting in lsstSim:
forcedPhotCoaddConfig is the correct class for deepCoadd_forced_config in lsstSim Template may be wrong too, but it is somewhat dangerous to correct so I will not. Sorry for taking so long to respond. I have no familiarity with this obs package nor any plans for it to use meas_mosaic (although I'd highly doubt there are any such intentions seeing as jointcal will be LSST's version). However, the commit message introducing the mosaic-related datasets confirms their association with meas_mosaic:
https://github.com/lsst/obs_lsstSim/commit/2afb17daa100266e109fd4116343f09a03dd2eaf
and it was your commit:
https://github.com/lsst/obs_lsstSim/commit/73ea51022c2c85bc68b3711c493d406942e1d339
that removed the config dataset.
Perry Gee
added a comment - Implemented changes to this Mapper.paf from RFC-233 
Lauren MacArthur, there is a dataset called Mosaic_metadata (no config) in this mapper, but it does not appear to be related to the MosaicTask in meas_mosaic. Can you comment on whether this dataset is in use?