IPP Software Navigation Tools IPP Links Communication Pan-STARRS Links

Changes between Version 75 and Version 76 of ppStack_testing_201111


Ignore:
Timestamp:
Mar 19, 2012, 2:52:36 PM (14 years ago)
Author:
Mark Huber
Comment:

--

Legend:

Unmodified
Added
Removed
Modified

  • ppStack_testing_201111

    v75 v76  
    8686
    8787
    88 
     88All testing currently being dumped into /data/ipp059.0/mhuber/testppstack_deep201111/ which has reached >3TB and will need to probably clean up some soon.
    8989
    9090
    … …  
    137137== MD04: 20 input case: ==
    138138
    139  * initially run for a memory test, but added extra outputs
    140  * input FWHM 4.49+/-0.48 (ave?)
     139 * initially run for a memory test, but added extra outputs and some missing
     140 * input FWHM 4.49+/-0.48 (mean)
    141141 * target seeing FWHM: 5.06, kron ~4.0
    142142 *
    … …  
    184184== Sim: 20 input case: 5 groups ==
    185185
     186Simulated inputs:
     187
     188 * basic set of 4 images of growing FWHM and repeated as 4 groups for 20 inputs
     189 * input FWHM 4.57+/-0.72 (mean)
     190 * target seeing FWHM: 5.70 -- set by worst input, Kron ~3.1 from log
     191 * some wildly varying kron values
     192
    186193|| '''filename''' || '''sim FWHM''' || '''sim FWHM_MAJ''' || '''.log Input FWHM  (pixel)''' ||  '''.log kron (pixel)''' ||
    187194||  ||  ||  ||  ||  ||
    … …  
    192199||  ||  ||  ||  ||  || 
    193200
    194  * basic set of 4 images of growing FWHM and repeated as 4 groups for 20 inputs
    195  * input FWHM 4.57+/-0.72 (ave?)
    196  * target seeing FWHM: 5.70 -- set by worst input, kron ~3.1
    197  * some wildly varying kron values
    198 
     201Additional and final outputs:
    199202
    200203|| '''filename''' || '''FWHM_MAJ (pixel)''' || '''IQ_FW1 (pixel)''' ||
    … …  
    222225== Sim: 20 input case w/ similar inputs: 10 groups ==
    223226
     227 * similar set, just created with 2 types with similar FWHM
     228 * input FWHM 4.07+/-0.19 (ave)
     229 * target seeing FWHM: 4.28 -- set by worst input, kron ~2.26
     230
    224231|| '''filename''' || '''sim FWHM''' || '''sim FWHM_MAJ''' || '''.log Input FWHM  (pixel)''' || '''.log kron (pixel)''' ||
    225232||  ||  ||  ||  ||  ||
    … …  
    227234|| _001 || 1.1 || 4.25(1) || 4.26(1) || 2.28(4) || 
    228235
    229  * similar set, just created with 2 types with similar FWHM
    230  * input FWHM 4.07+/-0.19 (ave)
    231  * target seeing FWHM: 4.28 -- set by worst input, kron ~2.26
    232  * s
     236
     237 * 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?)
     238 * small difference in the fake_and conv. so 20 input case w/ range of inputs better
    233239
    234240|| '''filename''' || '''FWHM_MAJ (pixel)''' || '''IQ_FW1 (pixel)''' ||
    … …  
    245251||  ||  ||  || 
    246252
    247  * odd 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?)
    248  * small difference in the fake_and conv. so 20 input case w/ range of inputs better
    249253
    250254
    251255== Sim: 100 input case: 25 groups ==
    252256
     257 * same set as before, just more groups
     258 * input FWHM 4.56+/-0.71 (mean)
     259 * target seeing FWHM: 5.72, kron ~3.02
     260 * an odd run mix-up in this test? 
    253261
    254262|| '''filename''' || '''sim FWHM''' || '''sim FWHM_MAJ''' || '''.log Input FWHM  (pixel)''' ||  '''.log kron (pixel)''' ||
    … …  
    260268||  ||  ||  ||  ||  || 
    261269
    262  * same set as before, just more groups
    263  * input FWHM 4.56+/-0.71 (ave?)
    264  * target seeing FWHM: 5.72, kron ~3.02
    265  * an odd run mix-up in this test?
     270
     271 * the x/y in stack is stack.cmf versus running psphot manually on stack, IQ_FW1 is missing from stack.cmf
     272 * stack.000/stack.0.conv.cmf has 6.26/5.37
     273 * odd the fake_ target image is much smaller than the target seeing FWHM chosen -- while the convolved images are larger than the target still.
    266274
    267275
    … …  
    279287||  ||  ||  ||
    280288
    281  * the x/y in stack is stack.cmf versus running psphot manually on stack, IQ_FW1 is missing from stack.cmf
    282  * stack.000/stack.0.conv.cmf has 6.26/5.37
    283  * odd the fake_ target image is much smaller than the target seeing FWHM chosen -- while the convolved images are larger than the target still.
    284289
    285290----
    … …  
    289294== SYS.ERR/target variance ==
    290295
    291 setting the SYS.ERR=0 to not mix an N scaled model flux fraction into the variance (SYS.ERR is applied to both stamps)
     296Setting the SYS.ERR=0 to not mix an N scaled model flux fraction into the variance (SYS.ERR is applied to both stamps)
    292297 * run with current ops tag (20120216), need to run with TESTING turned on or update the branch since there is still some blowup
    293  * 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/calc'd
     298 * 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
    294299
    295300|| '''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)''' ||
    … …  
    305310
    306311
    307 a sys error on the real image may still be useful as a tuning parameter, but found to still degrade the ability to achieve a good convolution with 0.1
     312A sys error on the real image may still be useful as a tuning parameter, but found to still degrade the ability to achieve a good convolution with 0.1
    308313 * sim 100 inputs, target~5.7, conv output ~7.8
    309  * selected stamps no significant different to if sys.err=0 -- better to test on MD04 sample however, need hardcopy plot of stamps chi2 vs flus
    310 
    311 seems to be slightly greater sensitivity for rejecting input images
    312 
    313 
    314 to setup stack to use SYS.ERR=0, set SYS.ERR in the STACK metedata in recipes/ppSub.config
    315 
    316 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. next test improvement..
     314 * selected stamps no significant different to if sys.err=0 -- better to test on MD04 sample however, need hardcopy plot of stamps chi2 vs flux ('''TODO''')
     315
     316
     317To setup stack to use SYS.ERR=0, set SYS.ERR in the STACK metedata in recipes/ppSub.config
     318
     319
     320How 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. next test improvement..
     321May have slightly greater sensitivity for rejecting input images on the worse quality ('''QUANTIFY''').
    317322 * could also just keep SYS.ERR applied to real image --
    318323
    319324
    320 need to run larger FPA skycell sample, multiple filters of 100, 400 to test stability
     325Need to run larger FPA skycell sample, multiple filters of 100 - 400 to test stability
    321326{{{
    322327mem400_stack_test.fixtest_s055g/mem400.log:      Target seeing FWHM: 10.172999                   
    … …  
    350355
    351356
    352 suspect this is also an issue (minor) with the single convolution diffims as well
    353 
    354 leads into larger devel need of splitting/carrying the variances through and not dumping into a weight variable so early.
     357
     358Suspect SYS.ERR is also an issue (minor) with the single convolution diffims as well, so add back to test [wiki:ppSub_testing_201201].
     359
     360Leads into larger development need of splitting/carrying the variances through and not dumping into a weight variable so early.
    355361
    356362
    … …  
    406412
    407413
    408