Summary ---IN PROGRESS PLEASE DO NOT EDIT---
- Summary ---IN PROGRESS PLEASE DO NOT EDIT---
- Test Sets
- Initial Tests
- Modification Tests
- Monitor Modifications
Discovered for many inputs into ppStack (i.e., deep and even refstacks), that the PSF in the output convolved stack grows with the number of inputs beyond the initial target PSF. Had been missed due to convolved stacks really only being used for magic, but was noticed with some M31 analysis and deepstacks of MD04. Memory use has also been an issue with the deepstacks and simultaneously being investigated.
Example: some MD04 skycell.055, ~100 input warps
- FWHM input average from log = 4.48+-0.81 (pixels)
- largest input has FWHM_MAJ of ~7.2
- target FWHM from log 6.88, Kron~6.29 from convolution part of log
- convolved stack FWHM_MAJ=14.36
- for comparison, unconvolved stack FWHM_MAJ=4.32
 | left to right: input warp FWHM_MAJ~6.8, unconvolved stack, convolved stack |
More specific examples:
- MD04.refstack.20111012 skycell.055-i ~92 inputs
Input FWHMs : 4.010976 +/- 0.334657 fit ext: 1.439042 sec for 225 of 225 sources fit psf: 0.395260 sec for 172 of 225 sources apresid: 0.022497 +/- 0.005780 (systematic error) from statistics of 225 psf stars (172 used) chisq vs flux fit term 0: 0.000000 +/- 0.000000 chisq vs flux fit term 1: 0.000000 +/- 0.000000 try model PS_MODEL_PS1_V1, ap-fit: 0.022497 +/- 0.005780 : sky bias: 0.000000 Target seeing FWHM: 4.706741 --> output convolved PSF, end of log fwhm (psf): 10.825166,10.697605 (moments): 0.000000,0.000000
- MD04.deeptest.20111012 skycell.055-i ~300 inputs
Input FWHMs : 4.835018 +/- 0.719810 fit ext: 1.385216 sec for 225 of 225 sources fit psf: 0.441030 sec for 171 of 225 sources apresid: 0.055969 +/- 0.001243 (systematic error) from statistics of 225 psf stars (171 used) chisq vs flux fit term 0: 0.000000 +/- 0.000000 chisq vs flux fit term 1: 0.000000 +/- 0.000000 try model PS_MODEL_PS1_V1, ap-fit: 0.055969 +/- 0.001243 : sky bias: 0.000000 Target seeing FWHM: 6.630503 --> output convolved PSF, end of log fwhm (psf): 14.636683,14.511685 (moments): 0.000000,0.000000
Status
- ...
- 6/19 restarting to look at target PSF, merged trunk (r34040) to pptack_test branch again, so tests (target PSF) now will be using that
- 6/20 added simple PSF model option to use mean FWHM + X for Target PSF for testing
- 6/21 finally brought page up to date with notes.. reset testing directories for target PSF tests and re-ran initial simtest set.
- 6/28 revised sim run based on results/plots from 6/21.
Other Issues/Problems
Will add to as things are found/discussed:
- (fixed) hard limit on number of inputs
- was shell descriptors limit of 1024 easily being reached with image,mask,weight,psf,cmf files for each input. Raised to 65535 now by default in shells.
- (in progress) selection of target PSF FWHM (for convolved stack)
- worse input PSF not necessarily best choice as could end up being rejected and only a few inputs may be that large (not representative)
- (on hold) heavy RAM use
- 10-20GB typical for 100+ inputs -- tests run unoptimized, will need to look at effect
- (on hold) long run-time for 20+ inputs
- 400 inputs order 12-14hr (use method similar to psphotStack?) -- tests run unoptimized, will need to look at effect
- (on hold) stacks (nightly and deeper) seeing more of the fault 5 stack events
- for MD nightly where inputs have strangely small <2 pixel input PSF.
- Bill has code for quality setting but not turned on, more of an issue possibly with the PSF determination for edge skycells with possible poor extrapolation from center of the field.
- Not really a ppStack issue but needs to deal with and not fault.
- convolved stacks have different background than unconvolved stacks?
Code/Branch Modification
Branch for testing -- http://svn.pan-starrs.ifa.hawaii.edu/trac/ipp/browser/branches/meh_branches/ppstack_test
- originally r32161 from last summer -- needed to update --
- many outputs turned on (runs slower, makes many many files)
- ppStackMatch.c -- need add a TESTING2 for reading in already processed conv images..
- ppSubMatchPSFs.c has TESTING but not ifdef at start, is a .DEBUG. used to trigger?
Now all modification tests using updated branch from trunk/ipp-20120531 (r34040, 6/19/2012) -- Target PSF for conv stacks (some modifications still not added below yet, mostly TESTING?)
- ppStackTarget.c -- Gene found bug with meanVariance after merge, so added
- simple model PSF as target PSF set by input FWHM for new tag ipp-20120626
- ppStackOptions.c/h -- added clippedMean/Stdev floats
- ppStack.h -- ppStackPSF for all option not just option Mask
- ppStackPrepare.c -- setting the option->clippedMean/stdev
- ppStackPSF.c -- use the clippedMean and stdev if options for it
- recipes/ppStack.config -- added keyword for +Nsig option
All modification tests now using updated branch from ipp-20120216 tag (r33407, 3/04/2012) -- testing modifications, some added to trunk already
- pmSubtractionMatch.c:
- has isfinite(sysError) rather than skyError compared to 0 - fixed +trunk
- sbrk in memCheck issue on OSX - not sure how to fix, added note +trunk
- pmSubtraction.c:
- TESTING for output conv1/2_* has problem with index now -- what supposed to be? fix removed index and add comment +trunk
- comment ConvolveStampThread footprint as stamp index is wrong -- fix +trunk
- ppStackMatch.c
- TESTING w/ and w/o using previous conv images or output samples.. need to split the two -- more explicit TESTING_REUSE_CONV case added - fix +trunk
- Bill found MASK.BAD/VAL issue in psphotStack also will be here in BG model, no clear problem noticed here -- fix local (should merge from trunk to pick up other fixes)
- ppStackSetup.c -
- also conflict issue with names ZP is active, MATCH.ZERO.POINTS inactive for recipes/ppStack.config -- leaving, not sure how fix
- ppStackSources.c -
- logic issue with ZP and photometry option, if ZP false then segfault - false not work, need to be OR? -- need to trace out
- also TESTING mix again - outputs and fake added mag offsets -- wait
- TESTING section for start_#.fits not possible now with order change in ppStackPrepare.c -- local and turned off now
- if PHOTOMETRY and many rejected from poor detections or in past more commonly magic masking on edges, then Abort case triggered since minMatches WRT all inputs and not numGoodImages -- fix local
- ppStackConvolve.c calls with normal index but ends up odd index in kernel name - bug?
- .kernel index odd, why not seq in pmFPAfileActivateSingle - psMod/camera/pmFPAfileIO.c -- SelectSingle in pmFPAfile.c -- could use date stamp or is threading the reason?
- pmSubtractionStamps.c
- the stamp.dat output condition case is trouble, if no sources in a stamp will segfault -- set to 0 again.. -- wait
- recipes/ppSub.config
- set SYS.ERR in the STACK metedata -- ALL stacks will use this then -- fix +trunk +ipp-20120216 tag (3/16 UT)
- ppStackPrepare.c
- order logic for reject masking images changed, since target PSF uses all "good" images, all rejections should be done before but zero point/transpanency calcs for each image were done after. Moved above target PSF section, this makes the TESTING section in ppStackSources.c for start_#.fits images not possible since uses the target PSF to make. -- fix local
- pmReadoutFake.c
- typecast S32/F32 conflict resulting in 0 value wrongly setting stamp sizes when calling pmReadoutFakeFromSources -- fix local +trunk+20120404 tag included
- psLib/somewhere
- needs to be fixed if -Db option used that it fails like others in that a config option not provided and not silently go back and use one from default config files
- ippTools/src/stacktool.c,
- more options for definebyquery likely necessary to dump problem input images (ellipticity immediately comes to mind..) - added locally only, added to trunk/tag too
All testing currently being dumped into /data/ipp059.0/mhuber/testppstack_deep201111/ which has reached >3TB and will need to probably clean up some soon (>2TB in 400 stk grizy test)
- moved to underused disk /data/ipp060.0/mhuber/testppstack_deep201111 (june 2012)
Test Sets
More easily setup than for testing ppSub, need a large number of various inputs and some number of similar inputs.
MD04 (GR0)
Sample of 20 and 100 drawn from the the MD04.V3 GR0 deepstack set for memory testing as well.
- FWHM in ppStack logs, cut and put into histogram below. Need to compare to psphot FWHM_MAJ/IQ_FW1?
Batch scripts for running in /data/ipp059.0/mhuber/testppstack_deep201111/MDdeeptest
Simulated (simtest)
A small variety set of four with many groups to make 20 and 100 inputs sets
Simtest setup (mostly default):
- default pixelscale ~0.26"/pixel, exptime 20s, filter r, 1k x 1k
- PSF.MODEL PS_MODEL_GAUSS
| filename | sim FWHM | FWHM_MAJ (pixel) | IQ_FW1 (pixel) |
| image.0 | 1.0 | 3.82 | 2.98 |
| image.1 | 1.1 | 4.19 | 3.28 |
| image.2 | 1.2 | 4.58 | 3.62 |
| image.3 | 1.5 | 5.69 | 4.49 |
| _and repeat_ | |||
Batch scripts for running in various /data/ipp059.0/mhuber/testppstack_deep201111/ppstack_simdeep* (needs cleanup)
Initial Tests
- do clean simulated images suffer the same problem?
- can use for more detailed testing since faster to run than MD04 skycell and not particular to GPC1 data
- is it related to the range of inputs and number
- is 20 inputs enough to reveal problem, if 100 then will be tedious
- what are the output convolved images being stacked, what is the target the images are being convolved to
- need to understand what the code does, so first test is to look at each output, add others and step through the code...
- tests indicate convolution of each input seem to be larger than expected -- appears to be also with the range of input FWHM
MD04: 20 input case:
- initially run for a memory test, but added extra outputs and some missing
- input FWHM 4.49+/-0.48 (mean)
- target seeing FWHM: 5.06, kron ~4.0
| filename | FWHM_MAJ (pixel) | IQ_FW1 (pixel) | .log Input FWHM (pixel) | .log kron (pixel) |
| real_000 | 5.38 | 4.20 | 5.40 | 4.25 |
| real_002 | 5.08 | 4.02 | 5.03 | 4.24/33 |
| real_009 | 4.27 | 3.17 | 4.19 | 3.42/33 |
| fake_000 | 5.10 | 4.09 | ||
| fake_002 | 5.13 | 4.11 | ||
| fake_009 | 5.10 | 4.09 | ||
| conv.0 | x | x | ||
| conv.2 | 8.41 | 6.37 | ||
| conv.9 | 7.91 | 6.02 | ||
| comb.initial | x | x | ||
| stack | 8.39 | 6.06 (also initially missing from header) | ||
| stack.unconv | 4.02 | 3.55 | ||
MD04: 100 input case:
- was run for memory monitoring, will need to be rerun for real/fake/conv outputs later
- input FWHM 4.84+/-0.81 (ave?)
- target seeing FWHM: 6.88, kron ~6.29
| filename | FWHM_MAJ (pixel) | IQ_FW1 (pixel) | .log Input FWHM (pixel) | .log kron (pixel) |
| real_001 | x | x | 5.40 | 4.39/35 |
| real_009 | x | x | 4.65 | 3.63/62 |
| fake_001 | x | x | ||
| fake_009 | x | x | ||
| conv.0 | x | x | ||
| conv.9 | x | x | ||
| comb.initial | x | x | ||
| stack | 14.36 | 9.75 (was missing from header) | ||
| stack.unconv | 4.31 | 4.15 | ||
Sim: 20 input case: 5 groups
Simulated inputs:
- basic set of 4 images of growing FWHM and repeated as 4 groups for 20 inputs
- input FWHM 4.57+/-0.72 (mean)
- target seeing FWHM: 5.70 -- set by worst input, Kron ~3.1 from log
- some wildly varying kron values
| filename | sim FWHM | sim FWHM_MAJ | .log Input FWHM (pixel) | .log kron (pixel) |
| _000 | 1.0 | 3.82 etc | 3.82(1) | 1.99/92,88/78,05/05,91/81 |
| _001 | 1.1 | 4.19(1) | 2.19/09 | |
| _002 | 1.2 | 4.58(1) | 2.55/48 | |
| _003 | 1.5 | 5.69(1) | 3.11/00 | |
Additional and final outputs:
| filename | FWHM_MAJ (pixel) | IQ_FW1 (pixel) |
| real_000 | 3.88 | 3.00 |
| real_003 | 5.82 | 4.54 |
| fake_000 | 5.10 | 3.69 |
| fake_003 | 5.04 | 3.68 |
| conv.1 | 6.51 | 5.10 |
| conv.3 | 6.60 | 5.19 |
| comb.initial | 6.57 | 5.58 |
| stack | 6.57 | 5.41 |
| stack.unconv | 4.33 | 3.49 |
filename descriptions:
- real_ = input warps
- fake_ = target image to convolve the real to (bit more than just that, selected target PSF sources, background, scaling)
- conv. = convolved real image
- comb.initial = initial stack w/o rejecting input images (that all?)
- stack = final convolved stack
Sim: 20 input case w/ similar inputs: 10 groups
- similar set, just created with 2 types with similar FWHM
- input FWHM 4.07+/-0.19 (ave)
- target seeing FWHM: 4.28 -- set by worst input, kron ~2.26
| filename | sim FWHM | sim FWHM_MAJ | .log Input FWHM (pixel) | .log kron (pixel) |
| _000 | 1.0 | 3.87(2) | 3.88(1) | 2.1(1) |
| _001 | 1.1 | 4.25(1) | 4.26(1) | 2.28(4) |
- oddly final stack is ~20% larger than conv and initial stack -- need to check -- seems so, may be another issue to watch for (really poor convolved input?)
- small difference in the fake_and conv. so 20 input case w/ range of inputs better
| filename | FWHM_MAJ (pixel) | IQ_FW1 (pixel) |
| real_000 | 3.87 | 3.00 |
| real_001 | 4.24 | 3.32 |
| fake_000 | 4.27 | 3.22 |
| fake_001 | 4.27 | 3.21 |
| conv.0 | 4.49 | 3.52 |
| conv.1 | 4.56 | 3.58 |
| comb.initial | 4.54 | 3.77 |
| stack | 5.37 | 4.36 |
| stack.unconv | 3.99 | 3.17 |
Sim: 100 input case: 25 groups
- same set as before, just more groups
- input FWHM 4.56+/-0.71 (mean)
- target seeing FWHM: 5.72, kron ~3.02
- an odd run mix-up in this test?
| filename | sim FWHM | sim FWHM_MAJ | .log Input FWHM (pixel) | .log kron (pixel) |
| _000 | 1.0 | 3.82(1) | 1.99/92,88/78,05/05,91/81 | |
| _001 | 1.1 | 4.19(1) | 2.19/09 | |
| _002 | 1.2 | 4.58(1) | 2.55/48 | |
| _003 | 1.5 | 5.69(1) | 3.11/00 | |
- the x/y in stack is stack.cmf versus running psphot manually on stack, IQ_FW1 is missing from stack.cmf
- stack.000/stack.0.conv.cmf has 6.26/5.37
- odd the fake_ target image is much smaller than the target seeing FWHM chosen -- while the convolved images are larger than the target still.
| filename | FWHM_MAJ (pixel) | IQ_FW1 (pixel) |
| real_000 | 3.82 | 2.11/24 |
| real_003 | 5.69 | 3.10/02 |
| fake_000 | 4.86 | 3.64 |
| fake_003 | 4.85 | 3.63 |
| conv.0 | 6.44 | 5.74 |
| conv.3 | 6.61 | 5.63 |
| comb.initial | 8.33 | 6.89 |
| stack | 7.73/8.13 | x/6.72 |
| stack.unconv | 4.40 | 3.60 |
Modification Tests
SYS.ERR/target variance
(lots of previous tracing of stages and such...) SYS.ERR setting history:
- ppSub_vs_Hotpants
- IPP_Progress_Report_20100118
- IPP_Progress_Report_20100111
- IPP_Progress_Report_20100104
- set to 0.1 since USE_WEIGHT off in pmSubtractionEquation.c -- when changed?
- looks like change from eam_branch 2/10/2010 r26893 removed
- put back in 2/13/2011 r30622 -- "psf solution uses weights to get sensible chisq values"
For now, setting the SYS.ERR=0 to not mix an N scaled model flux fraction into the variance (SYS.ERR is applied to both stamps)
- run with current ops tag (20120216), need to run with TESTING turned on or update the branch since there is still some blowup
- may still want SYS.ERR applied relative to real images, will need to look at deep modification for how/when the variance/weights are passed/calculated
| filename | sim 20 FWHM_MAJ (pixel) | sim 100 FWHM_MAJ (pixel) | MD04 20 FWHM_MAJ (pixel) | MD04 100 FWHM_MAJ (pixel) | MD04 400 FWHM_MAJ (pixel) | |
| input/target FWHM | 4.5/5.7 | 4.6/5.7 | 4.5/5.3 | 4.8/6.9 | 5.1/7.1 | |
| real_000 | ||||||
| fake_000 | ||||||
| conv.0 | ||||||
| comb.initial | ||||||
| stack (log/psphot) | 5.6/5.8 | 5.6/5.8 | 5.8/5.8 | 7.8/7.8 | 9.7/9.8 | |
| stack.unconv | 4.3 | 4.3 | 4.0 | 4.2 | 4.6 | |
A sys error on the real image may still be useful as a tuning parameter -- sim set of 100
- initially may have been a mis-run test, re-running shows target~5.7, conv output ~5.6, so no growth
- selected stamps not significantly different to if sys.err=0 -- better to test on MD04 sample however, need hardcopy plot of stamps chi2 vs flux (TODO)
How does SYS.ERR=0 affect the image rejection list - seems to reject some of the worse seeing images, which are what is used to set the TARGET to begin with. (TODO)
- May have slightly greater sensitivity for rejecting input images on the worse quality (QUANTIFY).
- Could also just keep SYS.ERR applied to real image (see above) but may be a different rejection list since before was driven by huge simulated target flux..
Suspect SYS.ERR is also an issue (minor) with the single convolution diffims as well, so add back to test ppSub_testing_201201. Leads into larger development need of splitting/carrying the variances through and not dumping into a weight variable so early (see above).
Need to run larger FPA skycell sample, grizy filters of 100 - 400 to test stability
--> skycell.055 near center mem400_stack_test.fixtest_s055g/mem400.log: Input FWHMs : 6.017093 +/- 0.818257 mem400_stack_test.fixtest_s055i/mem400.log: Input FWHMs : 5.144256 +/- 0.946591 mem400_stack_test.fixtest_s055r/mem400.log: Input FWHMs : 5.681267 +/- 0.910019 mem400_stack_test.fixtest_s055y/mem400.log: Input FWHMs : 5.302593 +/- 0.957232 mem400_stack_test.fixtest_s055z/mem400.log: Input FWHMs : 5.116535 +/- 0.909322 mem400_stack_test.fixtest_s055g/mem400.log: Target seeing FWHM: 10.172999 mem400_stack_test.fixtest_s055i/mem400.log: Target seeing FWHM: 7.128121 mem400_stack_test.fixtest_s055r/mem400.log: Target seeing FWHM: 8.744406 mem400_stack_test.fixtest_s055y/mem400.log: Target seeing FWHM: 8.210619 mem400_stack_test.fixtest_s055z/mem400.log: Target seeing FWHM: 8.907239 --> convolved stack PSFs mem400_stack_test.fixtest_s055g/mem400.log: fwhm (psf): 10.278124,10.079561 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s055i/mem400.log: fwhm (psf): 9.777281,9.695623 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s055r/mem400.log: fwhm (psf): 11.805278,11.348278 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s055y/mem400.log: fwhm (psf): 8.573956,8.533222 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s055z/mem400.log: fwhm (psf): 9.877050,9.810744 (moments): 0.000000,0.000000 --> skycell.058 closer to edge mem400_stack_test.fixtest_s058g/mem400.log: Input FWHMs : 5.537649 +/- 0.747769 mem400_stack_test.fixtest_s058i/mem400.log: Input FWHMs : 5.045243 +/- 0.928610 mem400_stack_test.fixtest_s058r/mem400.log: Input FWHMs : 5.281566 +/- 0.931646 mem400_stack_test.fixtest_s058y/mem400.log: Input FWHMs : 5.341974 +/- 1.066415 mem400_stack_test.fixtest_s058z/mem400.log: Input FWHMs : 5.282053 +/- 0.859332 mem400_stack_test.fixtest_s058g/mem400.log: Target seeing FWHM: 8.574877 mem400_stack_test.fixtest_s058i/mem400.log: Target seeing FWHM: 8.097901 mem400_stack_test.fixtest_s058r/mem400.log: Target seeing FWHM: 7.353693 mem400_stack_test.fixtest_s058y/mem400.log: Target seeing FWHM: 8.535203 mem400_stack_test.fixtest_s058z/mem400.log: Target seeing FWHM: 8.813280 --> convolved stack PSFs mem400_stack_test.fixtest_s058g/mem400.log: fwhm (psf): 8.743460,8.612661 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s058i/mem400.log: fwhm (psf): 9.086349,9.011602 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s058r/mem400.log: fwhm (psf): 7.497458,7.397744 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s058y/mem400.log: fwhm (psf): 8.861631,8.803410 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s058z/mem400.log: fwhm (psf): 9.594481,9.520103 (moments): 0.000000,0.000000
Seems to be a few issues in this larger sample -
- if FWHM cut >7.5 pixels, why is target selected >7.5? Are input images not being rejected? Somewhat part of next section on Target PSF, but need to trace out here. Could split up 400 set into subgroups and run. Looks like rejected are not being used but is growing itself in converting to the PSF.MODEL PS_MODEL_PS1_V1 or just in the envelope itself to include some wacky IQ images.
- s055g - rerun in smaller sub-sets ~100, target is ~9.45 and has similar output. Limiting input FWHM <7.1 selects a target ~8.4. rejection for >7.5 is active. The large target PSF looks to be trying to account for inputs with odd IQ in the envelope calculation and will need to collect stats on the input images for comparison.
- s055r - appears to have a problem, the convolved images are poor and many are rejected. Target PSF is in some way not representative? 100 input subsets seem to not show this problem. Wacky bug, see pmReadoutFake footprint/stamp radius typecast to not 0.0 in threading.
- s055i - largest growth still from target to final stack (~2 pixels). The target image PSF is ~7 pixels as says, however all the convolved input images are 9-11 pixels. Rerunning this and see if can reproduce, seems so in the pmReadoutFake footprint/stamp radius typecast to not 0.0 in threading. Running sub-groups ~100 doesn't have this problem.
- generally would be helpful to have other stats, at least min/max inputs on the envelope/target PSF calculation.
Related to the s055r problem, there may be a problem with excess image rejection: generally ~20-40% are from the >7.5 pixel cut initially
- Note -- rejected will be cut for input into the unconvolved stacks as well. Even for 058g,i seems a bit higher than would want.
mem400_stack_test.fixtest_s055g/mem400.stats:REJECT_IMAGES S32 115 # Number of images rejected completely mem400_stack_test.fixtest_s055i/mem400.stats:REJECT_IMAGES S32 89 # Number of images rejected completely mem400_stack_test.fixtest_s055r/mem400.stats:REJECT_IMAGES S32 318 # Number of images rejected completely mem400_stack_test.fixtest_s055r_redo/mem400.stats:REJECT_IMAGES S32 317 # Number of images rejected completely mem400_stack_test.fixtest_s055y/mem400.stats:REJECT_IMAGES S32 45 # Number of images rejected completely mem400_stack_test.fixtest_s055z/mem400.stats:REJECT_IMAGES S32 45 # Number of images rejected completely mem400_stack_test.fixtest_s058g/mem400.stats:REJECT_IMAGES S32 101 # Number of images rejected completely mem400_stack_test.fixtest_s058i/mem400.stats:REJECT_IMAGES S32 122 # Number of images rejected completely mem400_stack_test.fixtest_s058r/mem400.stats:REJECT_IMAGES S32 54 # Number of images rejected completely mem400_stack_test.fixtest_s058y/mem400.stats:REJECT_IMAGES S32 61 # Number of images rejected completely mem400_stack_test.fixtest_s058z/mem400.stats:REJECT_IMAGES S32 62 # Number of images rejected completely
- see pmReadoutFake footprint/stamp radius typecast to not 0.0 in threading
pmReadoutFake footprint/stamp radius typecast fixed not 0.0 in threading
Re-run MD04@400 for s055 grizy including the SYS.ERR=0 and typecast fix:
mem400_stack_test.fixtest_s055g/mem400.log: Target seeing FWHM: 10.172999 mem400_stack_test.fixtest_s055i/mem400.log: Target seeing FWHM: 7.128121 mem400_stack_test.fixtest_s055r/mem400.log: Target seeing FWHM: 8.744406 mem400_stack_test.fixtest_s055y/mem400.log: Target seeing FWHM: 8.210619 mem400_stack_test.fixtest_s055z/mem400.log: Target seeing FWHM: 8.907239 mem400_stack_test.fixtest_s055g/mem400.log: fwhm (psf): 10.245927,10.035181 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s055i/mem400.log: fwhm (psf): 9.053534,9.025082 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s055r/mem400.log: fwhm (psf): 9.291595,9.220863 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s055y/mem400.log: fwhm (psf): 8.697776,8.658995 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s055z/mem400.log: fwhm (psf): 9.827286,9.793766 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s055g/mem400.stats:REJECT_IMAGES S32 115 # Number of images rejected completely mem400_stack_test.fixtest_s055i/mem400.stats:REJECT_IMAGES S32 88 # Number of images rejected completely mem400_stack_test.fixtest_s055r/mem400.stats:REJECT_IMAGES S32 71 # Number of images rejected completely mem400_stack_test.fixtest_s055y/mem400.stats:REJECT_IMAGES S32 50 # Number of images rejected completely mem400_stack_test.fixtest_s055z/mem400.stats:REJECT_IMAGES S32 43 # Number of images rejected completely
Fixes r-band issue, others mostly unaffected. Could affect other code/stages tuned params, need to check before adding to trunk+tag -- nothing drastic in samples so added fix.
Looks to be now a question of tracing the target PSF selection process for the g-band and the extra growth on i-band. Becoming an additional branch of investigation related to the IQ of the inputs going into the stacks that the PSF envelope solution is made from and need to trace down to individual inputs.
Break up into groups, then break into smaller sub-sets the groups having issues.
s055g: Target and convolved output PSF are relatively well matched except much greater than other target PSFs. Why.
- appears the normal maximum input FWHM driver are being rejected, ~80/400 in fact..
... Input 186: 6.528622 Input 187: 11.727304 Input 188: 11.179210 ... PSF FWHM for image 187 is too large (11.727304 vs 7.500000 maxFWHM) --- rejected. PSF FWHM for image 188 is too large (11.179210 vs 7.500000 maxFWHM) --- rejected. ... - then likely the PSF envelope trying to fit some odd inputs being an edge skycell, and break into groups to run - ~100 inputs
- 0-100 group okay, Target PSF ~6.7 pixels
- 100-200 group has issue with Target PSF ~10 -- something in this group
Input FWHMs : 6.206581 +/- 0.880895 Target seeing FWHM: 9.458570 - subdivide further -- s2s1 for ~25 - drop the 7-7.5 ones - 15, 19-25,49,51,54-55,58-59,61-62 (remember to add +1) -
Input FWHMs : 5.867408 +/- 0.759102 Target seeing FWHM: 8.387546 --> so it is the conversion of 7.5-> PS1 PSF? except full g gets to ~10.1, and below 7.3 gets to 8.3 so factor somewhere else still --> so could make set of high 7.x -- use the s2 set and add only -- s2s2 - 15, 19-25,49,51,54-55,58-59,61-62 - add +1 to list Input 0: 7.496574 Input 1: 7.430977 Input 2: 7.192523 Input 3: 7.257564 Input 4: 7.222930 Input 5: 7.049495 Input 6: 7.181604 Input 7: 6.598584 Input 8: 5.182988 Input 9: 7.168517 Input 10: 7.095002 Input 11: 7.000916 Input 12: 7.146842 Input 13: 7.231784 Input 14: 7.243643 Input 15: 7.007473 Input 16: 7.409586 Input FWHMs : 7.170876 +/- 0.208919 fit ext: 1.898029 sec for 225 of 225 sources | fit psf: 0.426171 sec for 161 of 225 sources | apresid: 0.120900 +/- 0.005411 (systematic error) from statistics of 225 psf stars (161 used) chisq vs flux fit term 0: 0.000000 +/- 0.000000 chisq vs flux fit term 1: 0.000000 +/- 0.000000 try model PS_MODEL_PS1_V1, ap-fit: 0.120900 +/- 0.005411 : sky bias: 0.000000 Target seeing FWHM: 9.458885 --> so looks like 7.49->9.45 - s2s3 for <7.4 and still 9.45 -- also 7.2, only to 7.1
- s2s4 - clear down to 7.1 -- drop 0,1,2,4,11,12,14
Input 0: 7.049495 Input 1: 6.598584 Input 2: 5.182988 Input 3: 7.095002 Input 4: 7.000916 Input 5: 7.007473 Input FWHMs : 6.655743 +/- 0.743387 fit ext: 1.873564 sec for 225 of 225 sources | fit psf: 0.475994 sec for 172 of 225 sources | apresid: 0.102933 +/- 0.013974 (systematic error) from statistics of 225 psf stars (172 used) chisq vs flux fit term 0: 0.000000 +/- 0.000000 chisq vs flux fit term 1: 0.000000 +/- 0.000000 try model PS_MODEL_PS1_V1, ap-fit: 0.102933 +/- 0.013974 : sky bias: 0.000000 Target seeing FWHM: 8.786390 - 200-345 group also issue ~10
- 300-345? yes ~8.3 -- s2s4 is test case then - try to track down which they are -
- s41 ~7.6 -- how many in/rejected?
- s42 ~6.8 but highest <7.5 is 6.2 and only couple
- s43 ~6.4 and clear
- s44 ~8.35 in/rej~14/6, 7.3 largest -- use this one, lots of large
- so then which ones in this set are causing the large PSF and why? maybe not too unlikely that it would change 7.3->8.3 so do another above
- specific sub test -- mem20_stack_test.fixfwhmincut055g
PSF FWHM for image 1 is too large (7.785680 vs 7.500000 maxFWHM) --- rejected. PSF FWHM for image 5 is too large (7.621037 vs 7.500000 maxFWHM) --- rejected. PSF FWHM for image 9 is too large (7.927819 vs 7.500000 maxFWHM) --- rejected. PSF FWHM for image 11 is too large (7.701270 vs 7.500000 maxFWHM) --- rejected. PSF FWHM for image 12 is too large (9.293368 vs 7.500000 maxFWHM) --- rejected. PSF FWHM for image 13 is too large (9.460286 vs 7.500000 maxFWHM) --- rejected. Input seeing FWHMs: Input 0: 7.312490 -- Input 1: 7.785680 Input 2: 6.733018 Input 3: 7.106098 Input 4: 7.107179 -- Input 5: 7.621037 Input 6: 6.591544 Input 7: 7.348337 Input 8: 6.802173 -- Input 9: 7.927819 Input 10: 7.252710 -- Input 11: 7.701270 -- Input 12: 9.293368 -- Input 13: 9.460286 Input FWHMs : 7.031693 +/- 0.286558 fit ext: 1.920824 sec for 225 of 225 sources fit psf: 0.459095 sec for 168 of 225 sources apresid: 0.102416 +/- 0.002603 (systematic error) from statistics of 225 psf stars (168 used) chisq vs flux fit term 0: 0.000000 +/- 0.000000 chisq vs flux fit term 1: 0.000000 +/- 0.000000 try model PS_MODEL_PS1_V1, ap-fit: 0.102416 +/- 0.002603 : sky bias: 0.000000 Target seeing FWHM: 8.350183
- may indicate only able to happen with large numbers of varying IQ inputs, not with 20
- really need more output details from target PSF selection to know what it is doing..
- -- OOPS -- unfortunately these were cleaned before listings saved so cannot trace filenames w/o rerunning...
s055i: While the FWHM~7.1, the convolved images have FWHM~9-10 generally.
FILENAME FWHM_MAJ IQ_FW1 mem400_deepstack.ppstack.0.conv.im.pht.fpa.cmf 10.47247 8.178617 mem400_deepstack.ppstack.1.conv.im.pht.fpa.cmf 11.10159 8.916221 mem400_deepstack.ppstack.10.conv.im.pht.fpa.cmf 9.585663 8.479208 mem400_deepstack.ppstack.100.conv.im.pht.fpa.cmf 10.89854 8.851029 mem400_deepstack.ppstack.101.conv.im.pht.fpa.cmf 10.68771 8.692747 mem400_deepstack.ppstack.102.conv.im.pht.fpa.cmf 10.64534 8.854232 mem400_deepstack.ppstack.103.conv.im.pht.fpa.cmf 9.786881 9.003357 mem400_deepstack.ppstack.104.conv.im.pht.fpa.cmf 10.20106 8.508514 mem400_deepstack.ppstack.105.conv.im.pht.fpa.cmf 10.36447 8.872278 mem400_deepstack.ppstack.106.conv.im.pht.fpa.cmf 11.17089 8.858175 mem400_deepstack.ppstack.107.conv.im.pht.fpa.cmf 11.02701 9.404055 mem400_deepstack.ppstack.108.conv.im.pht.fpa.cmf 10.61196 8.608229 mem400_deepstack.ppstack.109.conv.im.pht.fpa.cmf 11.64145 8.561131 mem400_deepstack.ppstack.11.conv.im.pht.fpa.cmf 10.83251 8.612485
- Break into sub-groups of different sizes (default is ~100 inputs)
mem400_stack_test.fixtest_s055i_s1sub/mem400.log: Target seeing FWHM: 6.948271 mem400_stack_test.fixtest_s055i_s2sub/mem400.log: Target seeing FWHM: 7.983406 mem400_stack_test.fixtest_s055i_s3sub/mem400.log: Target seeing FWHM: 8.228301 mem400_stack_test.fixtest_s055i_s4sub/mem400.log: Target seeing FWHM: 7.010231 mem400_stack_test.fixtest_s055i_s1sub/mem400.log: fwhm (psf): 7.771870,7.675667 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s055i_s2sub/mem400.log: fwhm (psf): 8.531821,8.457088 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s055i_s3sub/mem400.log: fwhm (psf): 8.927197,8.827012 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s055i_s4sub/mem400.log: fwhm (psf): 8.465529,8.424333 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s055i_s1s200sub/mem400.log: Target seeing FWHM: 7.993505 mem400_stack_test.fixtest_s055i_s1s300sub/mem400.log: Target seeing FWHM: 8.162872 mem400_stack_test.fixtest_s055i_s1s400sub/mem400.log: Target seeing FWHM: 7.128121 mem400_stack_test.fixtest_s055i_s1s200sub/mem400.log: fwhm (psf): 8.951500,8.863860 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s055i_s1s300sub/mem400.log: fwhm (psf): 9.282644,9.198133 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s055i_s1s400sub/mem400.log: fwhm (psf): 9.053215,9.020257 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s055i_s2s200sub/mem400.log: Target seeing FWHM: 7.011785 mem400_stack_test.fixtest_s055i_s2s300sub/mem400.log: Target seeing FWHM: 7.123811 mem400_stack_test.fixtest_s055i_s2s200sub/mem400.log: fwhm (psf): 8.598989,8.531817 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s055i_s2s300sub/mem400.log: fwhm (psf): 8.875980,8.798958 (moments): 0.000000,0.000000
- Appears to choose a much better FWHM for more recent data (s4) but overlap of s2s200 versus what a larger FWHM for s3 shows indicates something is going strange in s4.
- Break sub-groups into sub-sets
mem400_stack_test.fixtest_s055i_s4s1sub/mem400.log: Target seeing FWHM: 7.462551 mem400_stack_test.fixtest_s055i_s4s2sub/mem400.log: Target seeing FWHM: 5.529995 mem400_stack_test.fixtest_s055i_s4s3sub/mem400.log: Target seeing FWHM: 6.343721 mem400_stack_test.fixtest_s055i_s4s4sub/mem400.log: Target seeing FWHM: 7.060447 mem400_stack_test.fixtest_s055i_s4s1sub/mem400.log: fwhm (psf): 7.691631,7.642853 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s055i_s4s2sub/mem400.log: fwhm (psf): 5.841974,5.706679 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s055i_s4s3sub/mem400.log: fwhm (psf): 6.399483,6.340310 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s055i_s4s4sub/mem400.log: fwhm (psf): 7.731044,7.698786 (moments): 0.000000,0.000000 --> so s4s5.. and some s3s3-4 inputs may be involved? mem400_stack_test.fixtest_s055i_s4s5sub/mem400.log: Target seeing FWHM: 6.130914 mem400_stack_test.fixtest_s055i_s4s5sub/mem400.log: fwhm (psf): 6.409942,6.297189 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s055i_s3s3sub/mem400.log: Target seeing FWHM: 7.966679 mem400_stack_test.fixtest_s055i_s3s4sub/mem400.log: Target seeing FWHM: 7.331856 mem400_stack_test.fixtest_s055i_s3s3sub/mem400.log: fwhm (psf): 8.106666,7.973850 (moments): 0.000000,0.000000 mem400_stack_test.fixtest_s055i_s3s4sub/mem400.log: fwhm (psf): 7.526990,7.406355 (moments): 0.000000,0.000000
- Manually rerun without selected input/exposures:
neb://@HOST@.0/gpc1/condor_MD04.V3_01.haf/o5186g0067o.113047/o5186g0067o.113047.wrp.276921.skycell.055.fits -- fw~1.9 --> Target seeing FWHM drops to ~7.1 (from 8.2) neb://@HOST@.0/gpc1/condor_MD04.V3_01.haf/o5186g0072o.113052/o5186g0072o.113052.wrp.276963.skycell.055.fits -- fw~1.7 neb://@HOST@.0/gpc1/condor_MD04.V3_01.haf/o5186g0060o.113040/o5186g0060o.113040.wrp.276911.skycell.055.fits --> Target seeing FWHM drop to ~8.0
- seems a single odd exposure can drive the target PSF strangely where FWHM~7.1 but sets the convolved images ~9-10.
- o5186 exposures all help push the Target seeing FWHM down when included with larger sample of others.
- not a case of extra growth of the Target to output PSF, but odd value of the Target PSF.
Really looking now issue of types of inputs that drive the target PSF to "strange" values (typically larger for the PSF envelope to match, but also smaller). Can specific cuts be used to restrict odd or largely deviant inputs for the target PSF?
- really an external approach to cut inputs versus internal rejection from large deconvolution likely when using modified Target PSF, so hold here until work through Target PSF choice.
Larger sample test with skycell.02x ipp-20120404 tag (4/9/12)
Rerun MD04@all inputs using the ipp-20120404 tag with SYS.ERR=0, stamp size typecast fix, and minor others (need to note date since various changes been happening in tag -- 4/9/12)
mysql> select stack_id,state,path_base,skycell_id from stackRun join stackSumSkyfile using (stack_id) where data_group like "%deeptest.2012%"; +----------+-------+------------------------------------------------------------------------------------------+-------------+ | stack_id | state | path_base | skycell_id | +----------+-------+------------------------------------------------------------------------------------------+-------------+ | 848697 | new | neb://any/gpc1/MD04.deeptest.20120406/MD04.V3/skycell.021/MD04.V3.skycell.021.stk.848697 | skycell.021 | | 848698 | full | neb://any/gpc1/MD04.deeptest.20120406/MD04.V3/skycell.022/MD04.V3.skycell.022.stk.848698 | skycell.022 | | 848699 | full | neb://any/gpc1/MD04.deeptest.20120406/MD04.V3/skycell.023/MD04.V3.skycell.023.stk.848699 | skycell.023 | | 848700 | full | neb://any/gpc1/MD04.deeptest.20120406/MD04.V3/skycell.024/MD04.V3.skycell.024.stk.848700 | skycell.024 | | 848701 | full | neb://any/gpc1/MD04.deeptest.20120406/MD04.V3/skycell.025/MD04.V3.skycell.025.stk.848701 | skycell.025 | | 848702 | full | neb://any/gpc1/MD04.deeptest.20120406/MD04.V3/skycell.026/MD04.V3.skycell.026.stk.848702 | skycell.026 | | 848703 | full | neb://any/gpc1/MD04.deeptest.20120406/MD04.V3/skycell.027/MD04.V3.skycell.027.stk.848703 | skycell.027 | | 848704 | full | neb://any/gpc1/MD04.deeptest.20120406/MD04.V3/skycell.028/MD04.V3.skycell.028.stk.848704 | skycell.028 |
Run a bit more restrictive then before with an external cap at FWHM major<7.5 from database/camera exposure
stacktool -definebyquery -dbname gpc1 -set_dist_group NULL -set_data_group MD04.deeptest.20120406 -set_label MD04.deeptest.20120406 -set_workdir neb://@HOST@.0/gpc1/MD04.deeptest.20120406 -select_data_group condor_MD04.V3_01_dg -select_filter g.00000 -select_good_frac_min 0.05 -min_num 2 -select_fwhm_major_max 7.5 -set_reduction DEEP_STACK -simple -select_skycell_id skycell.02% -pretend >& run_stacktool.g.md04_deeptest.20120406.fwhm75_n2.log
| Target seeing | g | r | i |
| skycell.021 | fault | fault | fault |
| 022 | 8.89 | 7.94 | 7.67 |
| 023 | 8.25 | 8.30 | 9.02 |
| 024 | 8.21 | 7.74 | 7.36 |
| 025 | 8.30 | 7.91 | 7.88 |
| 026 | 8.44 | 8.76 | 7.63 |
| 027 | 7.75 | 9.06 | 9.43 |
| 028 | 15.80 | 8.65 | 9.61 |
| 029 | 8.50 | 9.42 | 7.98 |
As expected, see possible extreme target PSF with poorly constrained input quality and not just on the edges.
Sample run highlighted another issue with skycell.021 fault due to not enough sources match for transparency measurements that then triggers a full fault. Odd because there are sources in the field even though in a corner. What sets the minimum number of sources?
- ppStack log notes "Selected 0 sources for photometry analysis"
- ppStackSources.c -- modified section options->photometry where minMatches should be in reference to numGoodImages and not the full num images since if many are rejected (near edge) then becomes useless overly conservative match fraction.
skycell.021-g: another issue revealed where many transparency values are NAN but still uses the FWHM for the Target PSF because of order in the code.. really want to do all possible "cuts" before setting Target PSF.
...
Input 293: 7.673504
Input FWHMs : 6.015619 +/- 0.671127
PSF 240 is completely bad --- not including in envelope calculation.
...
Target seeing FWHM: 9.178185
--> final Ninput=7 (tedious to trace out) and undershot the target PSF value, not sure how.
fwhm (psf): 7.569923,7.118117 (moments): 0.000000,0.000000
Change made in ppStackPrepare.c -- move ppStackSourcesTransparency before generate target PSF/ppStackPSF section.
- Note -- this will invalidate the TESTING section that produces the start_xxx.fits images in ppStackSourcesTransparency that uses the defined PSF.TARGET
Target PSF for Convolved Stacks
Want to explore not necessarily setting the convolved "matched/equilized/homogenized" deep-stack target PSF to what generally appears to be driven by the worst input. Have seen cases when the poor inputs could be image rejected and in that case not a good idea to use anyways, and the target PSF re-chosen and stack remade. How can poor imputs be pre-filtered better? Is a most common input based target PSF better? What level of deconvolution is acceptable?
- simple cuts:
- for large input N, a rerun of PSF evelope/target PSF would be of little time cost but rerun of all the image convolutions is terribly expensive.
- for small input N, enter into decision of adding poorer images versus dropping 20-30% of possible inputs.
- instead of a re-run of target PSF with additional image rejection, could choose mean/median PSF or an optimal and use poor deconvolution to reject worse images?
- important to remember, rejecting inputs for convolved stacks is the same for unconvolved stacks (for the pixel rejection method), so excessive rejection not good for either.
- expect the rejecting and target PSF cases to possibly be different for MD (similar good inputs, easy to clip worse seeing tail without significant loss of depth) versus LAP (similar to wildly variable quality inputs, maybe few far outliers or maybe even spread of quality levels).
Purpose of convolved stacks?
- staticsky photometry -- convolved stacks have a similar PSF across the field for photometry and morphology measurements. However, also then goes through another round of (de)convolution/smoothing to match the PSF of all input filters.
- matched PSF over skycell, semi-matched PSF over larger areas? How necessary is that currently?
- specific science reasons for either or?
Test cases:
- Case 0: Greatly reduce the input image FWHM limits and accept possibly fewer inputs into the stacks. For stacks that fail without enough inputs, have a second pass with more relaxed limits. Pass 1 could also require a larger number of minimum inputs (larger than the normal 2-6 inputs currently) with a very restrictive "FWHM" cut (good). Pass 2 more relaxed allowing fewer minimum inputs and/or relaxed cuts (good as going to get).
- affects the unconvolved and well as the convolved however (though 3" seeing images isn't helping either way)
- Case 1: Enhance the already exising simple model target PSF code to interally use the already calculated input FWHM mean (and stdev) to set a simple target PSF.
- is a convolution to a simple model PSF workable
- how much deconvolution is acceptable in the inputs to the stack
- Case 2: Either modify the PSF envelope code to trend towards the more likely target PSF or sub-select the inputs for the the target PSF to be set to.
- open question if the PSF envelope code is behaving as intendent and needs to be tested further
- then similar to case 1 deconvolution acceptable over the field
Test runs (same as defined above ppStack_testing_201111):
- all previous improvements included, like SYS.ERR=0, comes after this stage for the convolution mainly so not important for picking target, but is important for final photometry/property comparisons.
- since comparing between different runs, will want to use same SEED values for ppStack and psphot at least initially. Will also want exactly same input warps copies (i.e., from simtest run and not regenerate for each to avoid any differences even if using same seed there as well).
- be sure to note PSF model choices for the two samples, will want to try match initially, but also look at when not well matched later.
- what is the baseline to use for relative comparison? Unconvolved stacks are made at same time, but can depend on the comparison. Simulated set has known values for its creation.
- will want to include spatial variations, simtest sample should be uniform but MD04 will not be and may have "strange" inputs to note and make use of.
- not sure how to included information rejected images by some IQ or deconvolution limits.
- Simtest sample:
- may want to make a sample distribution of simulated PSF FWHMs, currently sample is flat. Would need to revise the making of the stack.mdc list of inputs, or will just be worse/upper limit of possible stack degrading for including deconvolved images.
- start with sample N=20, not clear what N=100 would help to do here.
- PSF model is _GAUSS for simulated images, start with matched target model and mpdel in psphot. Will want to look at different matched target models.
- MD04 sample:
- for sub-samples of skycell.055: N=20 fairly irregular sample, N=100 similar to full sample
- skycell.055,058 for full test samples for deep stacks: current operating FWHM cut is at 7.5 pixels, generally excluding <10% of all available warps.
 |  | 
|
 |  |
- PSF model choices?
Case 0: Improved Cuts/Limits
Possible preliminary improvement: really can only judge with real data samples Sample summary of past LAP run stacks ~mhuber/lapstack20120510_inputtargetpsffwhm.log
- summary plots tedious but would be helpful to visualize how ofter more restrictive cuts would help
past/current ppStack.config for STACK_THREEPI:
- PSF.INPUT.MAX 12.0 -- anything >12 pixels excluded (3")
- PSF.INPUT.CLIP.NSIGMA NAN -- not used, but would redefine the INPUT.MAX to be clipped_mean + NSIGMA*clipped_stdev
proposed:
- PSF.INPUT.MAX 8.0 (or 2", MD deep stack is set for 7.5)
- PSF.INPUT.CLIP.NSIGMA 2.0 (or 1.0?)
Nightly stacks (primarily MD concerned) have NAN for PSF.INPUT.MAX and so is not set. If the night is decent, should be similar and minor issue (should get stats on though TBT)
Chris quickly looked at the FWHM of the exposures going into LAP stacks and order 90 percentile are <8 pixels. The NSIGMA is more difficult to quantify, so the general question is do we want to just help trim off more of the poorest inputs to keep the convolved target PSF more regular or be much more restrictive to improve the convolved target PSF at the cost of the number of inputs (maybe 10-30% at times for NSIGMA of 1). Any additional exclusion cuts include images that go into making the unconvolved stack as well.
Adding pass cases for limits on the different minimum number of inputs would require some task modifications in LAP (and nightly science).
Should add samples of N inputs with/without the worst inputs for contribution in unconvolved stacks.
Case 1: Initial Manually Set Target as Simple Model PSF (include FWHM)
To explore the implications of (de)convolving to a simple PSF model with simtest and MD data
Key config options:
- recipes/ppStack.config
PSF.MODEL STR PS_MODEL_GAUSS # Model for PSF generation PSF.AUTOSIZE BOOL FALSE # Determine output PSF size from input PSFs? PSF.OUTPUT.FWHM F32 5.0 # Target size for output PSF (if not auto-sized), -1 as FWHM PSF.OUTPUT.NFWHMSIG F32 0.0 # N FWHM-sig to be added to the FWHMmean for PSF -- simtest also has a ppStack.config -- be aware of DECONV.LIMIT, watch rejections PSF.MODEL STR PS_MODEL_GAUSS # Model for PSF generation DECONV.LIMIT F32 1.5 # Deconvolution fraction for rejecting entire image IMAGE.REJ F32 0.3 # Rejected pixel fraction threshold for rejecting entire image
- recipes/psphot.config
PSF_MODEL STR PS_MODEL_GAUSS EXT_MODEL STR PS_MODEL_GAUSS CONSTANT_PHOTOMETRIC_WEIGHTS BOOL FALSE POISSON.ERRORS.PHOT.LMM BOOL TRUE EXTENDED_SOURCE_FITS BOOL FALSE EXTENDED_SOURCE_ANALYSIS BOOL FALSE -- recipes/psphot.config STACKPHOT uses -- so will probably want to change on cmdline PSF_MODEL STR PS_MODEL_PS1_V1 EXT_MODEL STR PS_MODEL_QGAUSS CONSTANT_PHOTOMETRIC_WEIGHTS BOOL TRUE POISSON.ERRORS.PHOT.LMM BOOL TRUE -- simtest has a psphot.config PSF_MODEL STR PS_MODEL_GAUSS PSF_CLUMP_GRID_SCALE F32 1.0 # size of Sx,Sy image pixel
Basic Script:
- ppstack seed=16587045608812410537
- psphot seed=2818454766199170413
- -Db PSF.AUTOSIZE <X> -Df PSF.OUTPUT.FWHM <X>: [T,F,F,F,F,F,F,F], [0.0, 3.8, 4.2, 4.6, 5.0, 5.4, 5.7, -1.0]
ppStack -threads 4 -input stack.000/stack.mdc stack.000/stack -Db TEMP.DELETE F -R PPSTACK.OUTPUT.VARIANCE FITS.TYPE NONE -log test.log -recipe PPSTACK STACK_DEEP -recipe PPSUB STACK -recipe PSPHOT STACK -recipe PPSTATS STACKSTATS -stats test.stats -dumpconfig test.mdc -tracedest test.trace -trace err 10 -trace ppStack 10 -trace ppSub 10 -trace psModules.imcombine 10 -trace psModules.camera 10 -vvv -debug -Db PSF.AUTOSIZE F -Df PSF.OUTPUT.FWHM -1.0 -seed 16587045608812410537 >& run_stackdeep.log ## keep copy of local convolved input warps mkdir tmpconv cp /local/ipp/tmp/stack.*fits tmpconv/ ## normal psphot + STACKPHOT psphot -file stack.000/stack.fits -mask stack.000/stack.mask.fits -variance stack.000/stack.weight.fits stack.fullpht -dumpconfig fullpht_test.mdc -seed 2818454766199170413 -threads 4 -log fullpht_test.log psphot -file stack.000/stack.unconv.fits -mask stack.000/stack.unconv.mask.fits -variance stack.000/stack.unconv.wt.fits stack.unconv.fullpht -seed 2818454766199170413 -threads 4 -log fullpht_unconvtest.log psphot -file stack.000/stack.fits -mask stack.000/stack.mask.fits -variance stack.000/stack.weight.fits stack.stkpht -dumpconfig stkpht_test.mdc -recipe PSPHOT STACKPHOT -D PSF_MODEL PS_MODEL_GAUSS -D EXT_MODEL PS_MODEL_GAUSS -seed 2818454766199170413 -threads 4 -log stkpht_test.log psphot -file stack.000/stack.unconv.fits -mask stack.000/stack.unconv.mask.fits -variance stack.000/stack.unconv.wt.fits stack.unconv.stkpht -recipe PSPHOT STACKPHOT -D PSF_MODEL PS_MODEL_GAUSS -D EXT_MODEL PS_MODEL_GAUSS -seed 2818454766199170413 -threads 4 -log stkpht_unconvtest.log ## (de)convolved input warps psphot -file tmpconv/stack.0.conv.im.fits -mask tmpconv/stack.0.conv.mk.fits -variance tmpconv/stack.0.conv.var.fits stack.0.conv.fullpht -dumpconfig tmpconvstack.0.conv.fullpht.mdc -seed 2818454766199170413 -threads 4 -log fullpht.0.conv_test.log ..(1,2,3).. ## input warps -- should always be same, but should run to keep settings up to date psphot -file image.000/image.0.wrp.fits -mask image.000/image.0.wrp.mask.fits -variance image.000/image.0.wrp.wt.fits image.0.wrp.fullpht -seed 2818454766199170413 -dumpconfig image.0.wrp.fullpht.mdc -threads 4 -log fullpht.0.wrp_test.log ..(1,2,3).. ## header stats to log (easier to read) /home/panstarrs/mhuber/bin/gethead -auh *.pht*.cmf FWHM_MAJ IQ_FW1
NOTE: if wrongly called on command line with a bool -Db PSF.OUTPUT.FWHM 5.0 will silently fail and use the default configuration setting in ppStack.config
Sim set: currently on disk /data/ipp060.0/mhuber
- previous version (r?) in testppstack_deep201111/simstest/simdeep120321
- (need to add details later)
- revised version (6/21) moved to testppstack_deep201111/simstest/simdeep120619/prerun120621
- mostly similar run to revised (6/28), sample plots in main directory and of name format test.<A>.<B>.<page#>.png for delta plots <A>-<B> values (need to add indication which is delta and which is normal)
- revised version (6/28) run summary directories -- plot format similar
- ppstack_simdeep20.testpsfenvel -- normal PSF envelope version (from sim20 set, target PSF should be ~5.7 pixels, ~1.5" or largest input)
- ppstack_simdeep20.testsetpsf10 -- selected target PSF to be 1.0" (3.8 sim pixels)
- ppstack_simdeep20.testsetpsf11 -- selected target PSF to be 1.1" (4.2 sim pixels)
- ppstack_simdeep20.testsetpsf12 -- selected target PSF to be 1.2" (4.6 sim pixels)
- ppstack_simdeep20.testsetpsf13 -- selected target PSF to be 1.3" (5.0 sim pixels)
- ppstack_simdeep20.testsetpsf14 -- selected target PSF to be 1.4" (5.4 sim pixels)
- ppstack_simdeep20.testsetpsf15 -- selected target PSF to be 1.5" (5.7 sim pixels)
- ppstack_simdeep20.testsetpsffwhm -- simple PSF model sized to FWHM of inputs (for default sim20 set, should be ~4.5 pixels, 1.2")
- ppstack_simdeep20.testpsfjunk -- as junk name indicates, junk or misc test runs
- summary comparison plots table -- file format name based on last part of dirname for input vectors of the difference plots (xxxx.v1.v2.xxxx.png)
| note | plots |
| envelope PSF convolved vs unconvolved | 
|
| simple PSF@1.2 (~FWHM) convolved vs unconvolved |
|
| simple PSF@1.2 (~FWHM) vs envelope PSF convolved |
|
| simple PSF@1.5 (~envelope) vs envelope PSF convolved |
|
- completeness type plots table -- red=baseline (simulated input), blue=comparison for completeness
| note | plots |
| envelope PSF convolved vs unconvolved |
|
| envelope PSF vs simple PSF@1.2 convolved |
|
| envelope PSF vs sim20 inputs |
|
| simple PSF@1.2 vs sim20 inputs |
|
MD04 20 input set:
- previous version (cleaned..) : input=4.5, auto selected target 5.3 (5.4 largest input PSF)
- manually set target 4.5, selects target seeing PSF to be 5.3 still.
- manually set target 4.0, selects target seeing PSF to be 4.8. conv PSF grows to 5.3 (5.4 was largest input PSF, related?)
- revised version (6/30) in main directory testppstack_deep201111/MDdeeptest
- mem20_stack_test.fixpsfenvel -- normal PSF envelope version
- mem20_stack_test.fixsetpsffwhm -- simple PSF model sized to FWHM of inputs
need to revise/update MD04
Case 1b: Simple Model FWHM+X Target PSF
If stack quality (TBD) is highly sensitive to deconvolution, could look at simple PSF model set to the FWHM + X, where X could be the input FWHM stdev or stdev/mean.
PSF.OUTPUT.FWHM F32 -1.0 # Target size for output PSF (if not auto-sized), -1 as FWHM PSF.OUTPUT.NFWHMSIG F32 -1.0 # N FWHM-sig to be added to the FWHMmean for PSF, N<0 and then +FWHMstdev/FWHMmean
Case 2: Improved PSF Evelope as Target PSF
- tests to check doing what expecting of it
- improve rejection of extremely deviant inputs
- improve rejection of odd inputs
- generate envelope based on most common/likely versus all inputs
Monitor Modifications
Should watch results of changes in nightly science products.
SYS.ERR=0 @3/15
MD nightly stacks --
- rejection 0-1 inputs similar to before the change
- convol PSF growth before 4.6->7.1 from 3/11, now 4.9->5.3 from 3/19 and growth similar to that seen when manually setting target PSF (related?)
SAS stacks don't use convolved stacks so shouldn't have affect.
LAP not yet running again (3/19) --
Target PSF
More difficult to monitor as little uses the convolved nightly stacks (if anything?). However, if LAP is running then that will be the source to monitor.
Last modified
14 years ago
Last modified on Jul 11, 2012, 1:33:05 PM
Attachments (7)
- md04_100warp_sample_ds9.jpg (87.5 KB ) - added by 14 years ago.
- md04teststk_inputfwhm_histo.png (11.9 KB ) - added by 14 years ago.
- md04deep_g_inputfwhm_histo.png (14.8 KB ) - added by 14 years ago.
- md04deep_i_inputfwhm_histo.png (14.4 KB ) - added by 14 years ago.
- md04deep_r_inputfwhm_histo.png (14.6 KB ) - added by 14 years ago.
- md04deep_y_inputfwhm_histo.png (14.2 KB ) - added by 14 years ago.
- md04deep_z_inputfwhm_histo.png (14.2 KB ) - added by 14 years ago.
Download all attachments as: .zip
Note:
See TracWiki
for help on using the wiki.
