Summary of MOPS_IASC20130319_tests130422

Comparison of MOPS diffim discoveries to IASC visual chip inspection for OSS.20130319

Related summary pages:

• MD_WSdiffim_photcheck_test20121004
• ppSub_testing_201201
• MOPS_diff_false_positives

MOPS/IASC sample exposures to compare to (162.OSS.O.Q.w ps1_14_2124 visit 1-4, quad) stage _id:

| exp_id | exp_name    | dateobs             | comment                           |
+--------+-------------+---------------------+-----------------------------------+
| 590994 | o6370g0351o | 2013-03-19 11:37:26 | 162.OSS.O.Q.w ps1_14_2124 visit 1 | 
| 591016 | o6370g0373o | 2013-03-19 11:59:29 | 162.OSS.O.Q.w ps1_14_2124 visit 2 | 
| 591038 | o6370g0395o | 2013-03-19 12:21:39 | 162.OSS.O.Q.w ps1_14_2124 visit 3 | 
| 591060 | o6370g0417o | 2013-03-19 12:43:52 | 162.OSS.O.Q.w ps1_14_2124 visit 4 |

| exp_name    | comment                           | chip_id | warp_id |
+-------------+-----------------------------------+---------+---------+
| o6370g0351o | 162.OSS.O.Q.w ps1_14_2124 visit 1 |  803498 |  759973 | 
| o6370g0373o | 162.OSS.O.Q.w ps1_14_2124 visit 2 |  803520 |  759994 | 
| o6370g0395o | 162.OSS.O.Q.w ps1_14_2124 visit 3 |  803542 |  760016 | 
| o6370g0417o | 162.OSS.O.Q.w ps1_14_2124 visit 4 |  803564 |  760040 | 

| diff_id | warp1  | warp2  | data_group   |
+---------+--------+--------+--------------+
|  406579 | 759973 | 759994 | OSS.20130319 |  -- v1-v2 pos, v2-v1 neg
|  406688 | 760016 | 760040 | OSS.20130319 |  -- v3-v4 pos, v4-v3 neg

Related MOPS czar pages: ​http://mopshq2.ifa.hawaii.edu/czar/psmops_ps1_mdrm152/tracklets.html?sort=d&limit=100&nn=56369&minv=0.1&sm=162.OSS.O&run_id=831

Datastore products:

pub.791526.diff.406579	2013-03-19T14:01:22Z	IPP-MOPS	162.OSS.O.Q.w ps1_14_2124 visit 1 	o6370g0351o	o6370g0373o
pub.791568.diff.406688	2013-03-19T15:17:53Z	IPP-MOPS	162.OSS.O.Q.w ps1_14_2124 visit 3 	o6370g0395o	o6370g0417o

3330813323.gpc1_IPP-MOPS-TEST_IPP-MOPS-TEST.791526.pos.mops
3330814731.gpc1_IPP-MOPS-TEST_IPP-MOPS-TEST.791526.neg.mops

3331155060.gpc1_IPP-MOPS-TEST_IPP-MOPS-TEST.791568.pos.mops
3331158431.gpc1_IPP-MOPS-TEST_IPP-MOPS-TEST.791568.neg.mops

---

MOPS to IASC object comparison

Missed IASC objects within ~2" in the diffims: ~30% !!

o6370g0351o	o6370g0373o
3330813323.gpc1_IPP-MOPS-TEST_IPP-MOPS-TEST.791526.pos.mops = 147/514
3330814731.gpc1_IPP-MOPS-TEST_IPP-MOPS-TEST.791526.neg.mops = 168/523

o6370g0395o	o6370g0417o
3331155060.gpc1_IPP-MOPS-TEST_IPP-MOPS-TEST.791568.pos.mops = 170/519
3331158431.gpc1_IPP-MOPS-TEST_IPP-MOPS-TEST.791568.neg.mops = 169/505

Need MOPS diffim object list to reverse compare..

IASC Objects Missed in Diffims (Manual Inspection)

Cases for missing matches:

• S/N too low for detection in diffim -- really two cases and split as
  • (faint) source just too faint to accomodate the S/N hit in normal diffims (will need additional match to SMF for photerr measurement)
  • (conv) convolution fail and S/N degraded (could be dual convolution or wrong choice in convolving input or just bad convolution)
• edge or masked -- really two cases and split as
  • (edge) edges and masks have growth factor from convolution
  • (mask) non-target image has mask so diffim has mask (why WS diffim should be better in addition to S/N)

Object	diffim	mode	skycell	ra dec	case type	comment
TVPS480	791526.pos	1	1278.021	185.470925 -3.000194	faint	not visual in diffim, smudge in warp. SMF detect MAG_SIG~0.15, CAL_PSF=22.5
TVPS485	791526.pos	1	1278.011	185.551467 -3.201447	faint	in overlap and no detection, SMF
TVPS48U		1	1278.042	185.025058 -2.092311	edge, faint	SMF
B1795		1	1278.042	184.876229 -2.043800	edge
TVPS48V		1	1278.042	185.086067 -2.171717	faint	SMF MAG_SIG~0.09 CAL_PSF=22.0
K11UO3A	791526.pos	1	1279.019	186.089088 -3.404686	edge and mask	cell edge and mask

---

Re-run tests

diffim stats

For full OSS.20130319 diffim exposure/skycell sample:

• convolve input = 1393
• convolve template/reference/input 2 = 968
• dual convolution = 179

  -- need cam stage stat summary
        Input FWHM: 4.949776
        Reference FWHM: 5.265106
  

Only IASC exposures --

• convolve input = 52, 51 (v1-v2, v3-v4)
• convolve template/reference/input 2 = 2, 0
• dual convolution = 3, 6

Dual convolution typically lower good_frac and on edges of FPA -- add samples?

add rerun details with dual convolution and no-dual convolution (issue is typically no-dual case has different kernel params)?