IPP Software Navigation Tools IPP Links Communication Pan-STARRS Links

Changes between Initial Version and Version 1 of fullforce_intro


Ignore:
Timestamp:
Dec 10, 2013, 3:07:59 PM (13 years ago)
Author:
bills
Comment:

--

Legend:

Unmodified
Added
Removed
Modified
  • fullforce_intro

    v1 v1  
     1
     2== This page discusses the processing stages and database tables to be used to manage IPP full force photometry ==
     3
     4Notes on the new "full force" stages of the pipeline
     5
     6The current terminis of staticsky processing is the skycal results. The
     7granulatity of these results are one per filter per skycell per processing
     8version.
     9
     10The new full force stages of the pipeline will take these results
     11back to the input exposures and measure psf and extended source photometry
     12using the staticsky measurements as calibrated by skycal.
     13
     14The input paramters for the full force photometry is a 'cff' file.
     15For efficiency these will be generated during skycal processing.
     16An optional processing stage is outlined here which may be used to
     17generate cff files post skycal (in case algorithms change of tweaks to
     18the processing are necessary).
     19
     20After full force has been run for all of the  warps for a given skycell
     21and filter, the results are then summarized though a TBD process.
     22
     23
     24Tables
     25======
     26
     27{{{
     28fullForceWarpRun
     29    ffw_id
     30    warp_id
     31    skycal_id           (sets sky_id and through stack_id, filter,
     32                         tess_id, skycell_id)
     33    input_path_base     path that leads to cff file
     34                        Note: this is independent of the method
     35                        used to generate the data.
     36                        The expectation is that the cffs will be generated
     37                        during skycal processing. The data has been
     38                        read and is available. zero point is done.
     39
     40    workdir
     41    reduction           as usual specifies the recipe
     42    label
     43    data_group
     44    dist_group
     45}}}
     46
     47
     48This structure has to a separate run for each skycell. This will lead to a
     49Nstack (~2Million) * Ninputs (~12) runs for a processing version (PVn)
     50
     51Another option would be to structure like diffRun with a input table.
     52In that case skycal_id, and input_path_base would move there.
     53
     54At this level we're really doing the processing on the skycell, so the
     55grouping by exposure might be less interesting. Avoiding the extra
     56table simplifies things.
     57
     58Is the 'Warp' in the table name needed? Would we ever do fullforce on
     59a diff or stack? I'm hesitant to use the stage, stage_id style structure
     60(addRun, distribution) since that makes everything more complicated.
     61The code sharing ends up being somewhat illusory.
     62
     63
     64----
     65
     66We might as well anticipate that we will need to generate cffs as a
     67post skycal step. We don't want to require this because if we can generate
     68it at the skycal stage we will save an awful lot of I/O.
     69
     70
     71{{{
     72cffGenerateRun
     73    cffg_id
     74    skycal_id
     75    label
     76    reduction
     77    etc
     78
     79}}}
     80
     81When processed reads the skycal cmf and generates a cff file. Ehh why bother.
     82Why not just create a new skycalRun? Because the skycalRun might have already
     83been used to build psps, etc and so we probably shouldn't rerun calibration.
     84
     85===========
     86Once all of the fullforce warp processing is done we need to somehow collect
     87the results
     88
     89
     90{{{
     91ffullForceSummaryRun
     92    ffs_id
     93    skycal_id       (sets stack_id, filter, skycell_id)
     94
     95    The inputs are defined in the secondary table
     96  fullForceResultsInput
     97    ffw_id  (all inputs are linked back to the same skycal_id insured by query)
     98
     99}}}
     100
     101When processed this thing goes and generates a summary based on all of the
     102results. (Kind of like addstar and dvo)
     103
     104XXX: At this stage should we be combining the 5 filter results for each
     105skycell?
     106
     107TOOL
     108----
     109
     110{{{
     111
     112fftool  - full force tool
     113
     114modes
     115-definebyquery      define fullforce runs
     116    usual args to select warps
     117   
     118    args to select input from skycal run or cffgenerate run
     119    -select_skycal_label
     120    -select_skycal_id
     121
     122    -select_cffgenerate_label   if neither of these are specified assume
     123    -select cffg_id             input is in the skycal results
     124
     125    -set_ arguments for all other parameters
     126-addresult
     127-revert
     128-updateffrun
     129
     130
     131-definecffgenerate  define a cffGenerateRun which will make a
     132                    cff file for a stack from the skycal outputs
     133
     134    -select_skycal_label
     135    -select_skycal_id
     136    -set_ arguments for all parameters
     137
     138-addcffgresult
     139
     140
     141}}}