These are Heather's notes (and status) for SAS37 / ipptopsps

build of minis

This was done with minimal problems. Addstar ff/ sky had problems initially, Gene put in fixes for that. The faults that I have in the minis (only a handful of faults) are related to cluster / gpc1 problems. Rather than sort them out, I left those few exposures out. It's not complete (to ~10 exposures).

The labels and associated minis:

stage	label	mini
cam	SAS.20141118	SAS.20141118.cam
sky	SAS.20141118.rerun.V3	SAS.20141118.sky.rerun.V3
fullforce	SAS.20141118.rerun.V3	SAS.20141118.ff.rerun.V3
diff	SAS.20141118	SAS.20141118.diff

merge of SAS

Merge had problems, I retried 3 time with parallel then hacked it to not be parallel (and then it worked). The merged dvodb names have .V5 because that's the 5th total attempt to do this.

Name	minis	order
SAS.20141118.diff.V5	SAS.20141118.diff	1
SAS.20141118.V5	SAS.20141118.sky.rerun.V3	1
SAS.20141118.V5	SAS.20141118.cam	2
SAS.20141118.V5	SAS.20141118.ff.rerun.V3	3

V5 had an incorrect dvomerge command, so I redid them all as V6 - the minis are the same:

Name	minis	order
SAS.20141118.diff.V6	SAS.20141118.diff	1
SAS.20141118.V6	SAS.20141118.sky.rerun.V3	1
SAS.20141118.V6	SAS.20141118.cam	2
SAS.20141118.V6	SAS.20141118.ff.rerun.V3	3

calibration of SAS

See SAS37 calibration wiki

SAS37CalibrationHeatherNotes

creation of datastores

SAS37

dsprodtool --add PSPS_SAS37_IN --type IPP-PSPS --description 'PSPS SAS37 IN'
dsprodtool --add PSPS_SAS37_OB --type IPP-PSPS --description 'PSPS SAS37 OB'
dsprodtool --add PSPS_SAS37_P2 --type IPP-PSPS --description 'PSPS SAS37 P2'
dsprodtool --add PSPS_SAS37_ST --type IPP-PSPS --description 'PSPS SAS37 ST'
dsprodtool --add PSPS_SAS37_FO --type IPP-PSPS --description 'PSPS SAS37 FO'
dsprodtool --add PSPS_SAS37_FW --type IPP-PSPS --description 'PSPS SAS37 FW'

SAS37Diff

dsprodtool --add PSPS_SAS37_DIFF_IN --type IPP-PSPS --description 'PSPS SAS37 Diff IN'
dsprodtool --add PSPS_SAS37_DIFF_DO --type IPP-PSPS --description 'PSPS SAS37 Diff DO'
dsprodtool --add PSPS_SAS37_DIFF_DF --type IPP-PSPS --description 'PSPS SAS37 Diff DF'

For the next iteration, ie, we found bugs in ST and P2 which caused C/T problems;

dsprodtool --add PSPS_SAS37_P2_V2 --type IPP_PSPS --description 'second iteration of batches for SAS37'
dsprodtool --add PSPS_SAS37_ST_V2 --type IPP_PSPS --description 'second iteration of batches for SAS37'

status of SAS37 ready for psps

dvodb	datastore	batches	comments
SAS37 / SAS.20141118.V5	PSPS_SAS37_IN	1	IN batch
SAS37 / SAS.20141118.V5	PSPS_SAS37_OB		currently loading
SAS37 / SAS.20141118.V5	PSPS_SAS37_P2
SAS37 / SAS.20141118.V5	PSPS_SAS37_FW
SAS37 / SAS.20141118.V5	PSPS_SAS37_FO
SAS37 / SAS.20141118.V5	PSPS_SAS37_ST
SAS37Diff / SAS.20141118.diff.V5	PSPS_SAS37_DIFF_IN	1	IN Batch
SAS37Diff / SAS.20141118.diff.V5	PSPS_SAS37_DIFF_DO
SAS37Diff / SAS.20141118.diff.V5	PSPS_SAS37_DIFF_DF

queue of SAS37

located in '~heather/queueLoaders/SAS.loaders.newschema/createSAS37.pl', I have queued up the following:

• 851 OB
• 620 P2
• 612 ST
• 206 (??) DO
• 23224 DF

ipptopsps progress

I am nearly done with FW/FO stuff, the other batches have been done and tested.

Example batches that I tested:

• 9196 --> DO Batch
• 9200 --> DF Batch
• 9065 --> ST Batch
• 9067 --> P2 Batch
• 9192 --> OB Batch

I opened each batch in topcat and checked that the columns were mostly sensible and investigated 0's or -999s as necessary

DO Batch

• DiffObject
  • skycellID = -1 (not sure what it should be)
  • ra/dec err = -999
  • qfPerfect not by filter (H: need to fix)
  • ng/nr/ni/nz/ny = 0 (H: need to investigate)

DF Batch

• DiffMeta
  • magSat = 0. (???)
• DiffToImage
  • has 60 chips... (H: needs to select only the chips , see how Stacks do this)
• DiffDetection
  • telluricExt = 0 (this is in the dvo)
  • momentM3*/M4* = -999 (not in cmf)
  • appFillF = -999 (not in cmf)

P2 Batch

• FrameMeta (ok!)
• ImageMeta
  • magSat =0
  • completMag =0
  • psfModelID = -999
• Detection
  • posAngle = -999
  • extNSigma = -999 (H: I put this in here, but do we calculate it for camera?)
  • zp changes in a funny pattern (it looks like each chip is cut into 4 quarters, each quarter a different zp)
• psfMajorFWHM = -999
• psfMinorFWHM = -999
• psfCore = -999
• apFlux/apFluxErr/appFillF = 0

I am rechecking with P2 batch B00136491, this uses new addstar / new dvo / new calibration (compared to the above example, which was Gene's SAS35)

Numbers look much better:

• FrameMeta = ok!
• ImageMeta
  • magSat =0
  • completMag =0
  • psfModelID = -999
• Detection
  • posAngle is filled
  • extNSigma = -999
  • zp doesn't look so funky
  • psfMajorFWHM/psfMinorFWHM are filled properly
  • psfCore filled properly
• apFlux/ApFluxErr/appFillF filled properly

Here is a plot of zp for this exposure, I think it looks ok:

OB Batch

• ObjectThin
  • nStackObjectRows = 0 (not sure if this is just for this batch or for all)
• MeanObject (ok!)

• recheck with B00136451
  • ObjectThin
    • nstackobjectrows = 0
  • MeanObject- looks good!

ST Batch

(this batch was a lucky batch that only had one filter (near the edge) - so I only checked z filter)

• StackMeta
  • magSat, completMag, astroScat,photoScat, nAstro, nPhoRef, calibmodnum = -999
• StackObjectThin
  • bestDetection = -999
• StackObjectAttributes
  • bestDetection = -999
  • zExtNSigma = -999
• StackModelFitExp
  • stackDetectRowID = -999
  • bestDetection = -999
  • zexpRadiusErr, zExpAbErr, zExpPhiErr, zExpCf = -999
• StackModelFitDev
  • stackDetectRowID = -999
  • bestDetection = -999
  • zDevRadiusErr, zDevAbErr, zDevPhiErr, zDevCf = -999
• StackModelFitSer
  • stackDetectRowID = -999
  • bestDetection = -999
  • zSerRadiusErr, zSerAbErr, zSerPhiErr, zSerCf = -999
• StackPetrosian
  • stackDetectRowID = -999
  • bestDetection = -999
• StackApFlx
  • stackDetectRowID = -999
  • bestDetection = -999
• StackApFlxEGUnc
  • stackDetectRowID = -999
  • bestDetection = -999
  • apertures 6-10 unfilled
• StackApFlxEGCon6
  • stackDetectRowID = -999
  • bestDetection = -999
  • apertures 6-10 unfilled
• StackApFlxEGCon8
  • stackDetectRowID = -999
  • bestDetection = -999
  • apertures 6-10 unfilled

stackDetectRowID is clearly something I missed, that's easy to fix.

best Detection, is it set?

apertures 6-10 unfilled strikes me as odd (I need to investigate) the errors in the models - are these set?

• I investigated a SAS37 ST batch last night, and found the following:
  • StackMeta
    • magSat, completMag, astroScat,photoScat, nAstro, nPhoRef, calibmodnum = -999
  • StackObjectThin
    • bestDetection = -999
  • StackObjectAttributes
    • bestDetection = -999
    • ()ExtNSigma = -999
  • StackModelFitExp
    • bestDetection = -999
    • zexpRadiusErr, zExpAbErr, zExpPhiErr, zExpCf = -999
  • StackModelFitDev
    • bestDetection = -999
    • zDevRadiusErr, zDevAbErr, zDevPhiErr, zDevCf = -999
  • StackModelFitSer
    • bestDetection = -999
    • zSerRadiusErr, zSerAbErr, zSerPhiErr, zSerCf = -999
  • StackPetrosian
    • bestDetection = -999
  • StackApFlx
    • bestDetection = -999
  • StackApFlxEGUnc
    • bestDetection = -999
  • StackApFlxEGCon6
    • bestDetection = -999
  • StackApFlxEGCon8
    • bestDetection = -999

BestDetection is not yet set (I think this is OK) Other parameters are not yet set in cmf (I think) StackMeta- some params are missing (has always been the case)

Things I still need to fix

• stacks
  • number of apertures (from 10 to 9)
  • aperture metadata
• fullforce detections
• fullforce objects