WISE Software Interface Specification; QAoutput_icl06; (WSDC D-I137) Written by F. Masci, version 1.0, 01/16/2009 Interface Name: QA metadata for dynamic (on-orbit) responsivity generation Type of Interface: QA metadata written to ASCII file in IPAC table format Read By: WISE QA Application Framework ___________ H. Brandenburg D. Kirkpatrick C. Gelino Written By: COMPFLAT ________________________________ F. Masci DESCRIPTION: The dynamic responsivity (flat-field) generation pipeline COMPFLAT shall generate a table storing metrics and diagnostics to support Quality Assurance (QA) activities downstream. The table shall be in the IPAC standard format. There shall be one table per calibration FITS image product, per band. EXAMPLE TABLE: Below is an example of an output QA metadata table file from COMPFLAT. The metrics were computed from flat-field products generated from simulated frames (May 2008 simulation) and pre-processed with the WSDS pipeline. Specific algorithms are described in the COMPFLAT Subsystem Design Specification (in preparation). See below for some notes and associated graphics files. \ WISE QA metadata for flat-field calibration products \ Generated by compflat, v.1.3 on 2009-01-16 at 16:42:58 \ Definitions of metric identifiers: \ flt: flat-field (responsivity) values \ unc: flat-field uncertainty values \ Metric units pertain to dimensionless responsivity values \ FITS file products represented: \ 1: /wise/fmasci/testdata/flatcal1/outputs/flat-w1-est.fits \ 2: /wise/fmasci/testdata/flatcal1/outputs/flat-w1-unc.fits \ |name |band |hdrname |type |value |comment | |c |i |c |c |c |c | \ \ Flat-field responsivity metrics \ flatf:flt:numframes 1 NUMFRMS i 70 Number of input frames used flatf:flt:NumNaN 1 - i 0 Number of NaN pixels in responsivity map flatf:flt:Min 1 - r 0.999694466590881 flt pixels: Minimum pixel value in flat flatf:flt:Max 1 - r 5.75225257873535 flt pixels: Maximum pixel value in flat flatf:flt:Mean 1 - r 1.00009369850159 flt pixels: Mean pixel value in flat flatf:flt:Median 1 - r 1.00000476837158 flt pixels: Median pixel value in flat flatf:flt:StdDev 1 - r 0.0134721351787448 flt pixels: Standard (RMS) Deviation from mean (unbiased estimate) in flat flatf:flt:Mode 1 - r 0.999998986721039 flt pixels: Mode pixel value (fuzzy) in flat flatf:flt:Med16ptile 1 - r 4.63724136352539e-05 flt pixels: Robust sigma from Median - 16%-tile flatf:flt:84-16ptile 1 - r 4.72664833068848e-05 flt pixels: Robust sigma from [84%-tile - 16%-tile]/2 flatf:flt:Skewness 1 - r 235.648788353148 flt pixels: Sample skewness flatf:flt:Kurtosis 1 - r 65018.9286265043 flt pixels: Sample kurtosis flatf:flt:JBCoeff 1 - r 181835472936683 flt pixels: Jarque-Bera normality test coefficient; larger JBCoeff => more non-normal flatf:flt:Locount 1 - i 9 flt pixels: Number of low responsivity pixels [< med - 5*sigma = 0.999768435955048] flatf:flt:Hicount 1 - i 3226 flt pixels: Number of high responsivity pixels [> med + 5*sigma = 1.00024104118347] \ \ Flat-field uncertainty metrics \ flatf:unc:Min 1 - r 1.18841280709603e-05 unc pixels: Minimum pixel value in flat uncert flatf:unc:Max 1 - r 1.02813684940338 unc pixels: Maximum pixel value in flat uncert flatf:unc:Mean 1 - r 6.61484882584773e-05 unc pixels: Mean pixel value in flat uncert flatf:unc:Median 1 - r 4.48165665147826e-05 unc pixels: Median pixel value in flat uncert flatf:unc:MeanAccu 1 - r 0.00556510267779231 unc pixels: Mean accuracy in responsivity; = 100* % flatf:unc:MedianAccu 1 - r 0.00448162481188774 unc pixels: Median accuracy in responsivity; = 100*MED[uncert/responsivity] % NOTES: * We don't expect all metrics to be propagated to the final WISE QA Application Framework. Some of these will serve to (i) perform routine health/sanity checks, (ii) select appropriate products for the instrumental calibration pipeline, and (iii) tune parameters and optimize generation from on-orbit data. FILE NAME: Since COMPFLAT generates one QA meta-table per band, a suggested filename is: --meta.tbl ANCILLARY QA PRODUCTS: The following plots and graphics are currently being generated. From COMPCAL pipeline: * Histogram of full responsivity map. * Histogram of pixel responsivity accuracy: 100 * [uncertainty / responsivity] % Suggested for QA Application Framework: * JPEGs of on-orbit flat-field image products per band. * JPEGs of flat-field status mask images. These store the locations of thresholded low/high responsivity values.