Chris Morrison [X] (Inactive) added a comment - 01/Feb/19 11:54 AM Pessimistic matcher is performing as expected with no failures. The regression in squash values comes from differences in the matched pattern resulting in slightly different shift/rotations and therefore calculated distances. When these distances are clipped and feed to the WCS fitter the matches are then slightly different resulting in a bit more slop in the fit WCS that is apparent in the drp metrics. The solution found is to add a single iteration to the "final verify" step, using the first round of verified matches to refit the shift/rotation matrix resulting in a more stable set of matches regardless of the initial matched pattern. Below I'll post results from the 3 different validation datasets.

Chris Morrison [X] (Inactive) added a comment - 01/Feb/19 11:56 AM validation_data_cfht: Median value of the astrometric scatter - all magnitudes: 56.896 marcsec Astrometric scatter (median) - snr > 100.0 : 12.9 marcsec No stars found that are 199.0 arcmin-- 201.0 arcmin apart. ================================================================= r band metric measurements ================================================================= validate_drp.PF1_minimum_gri - LPM- 17 , p. 21 , http: //ls.st/lpm-17 The maximum fraction of magnitudes deviating by more than PA2_minimum_gri from mean. Failed minimum 40.74 % <= 20 % Failed design 40.74 % <= 10 % Failed stretch 56.3 % <= 5 % validate_drp.PA1 - LPM- 17 , p. 21 , http: //ls.st/lpm-17 The maximum rms of the unresolved source magnitude distribution around the mean value. Failed design 16.99 mmag <= 5 mmag Failed minimum 16.99 mmag <= 8 mmag Failed stretch 16.99 mmag <= 3 mmag validate_drp.PA2_minimum_gri - LPM- 17 , p. 21 , http: //ls.st/lpm-17 At most PF1_minimum_gri % of magnitudes may deviate by more than this magnitude from the mean. Failed minimum 24.44 mmag <= 15 mmag Failed design 35.4 mmag <= 15 mmag Failed stretch 50.32 mmag <= 10 mmag validate_drp.AD1_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 No more than AF1 of the astrometric distances will deviate by more than this from the median (see AM1, AF1). Passed design 14.61 marcsec <= 20 marcsec Passed minimum 8.062 marcsec <= 40 marcsec Failed stretch 20.7 marcsec <= 10 marcsec validate_drp.AD2_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 No more than AF2 of the astrometric distances will deviate by more than this from the median (see AM2, AF2). Passed design 16.16 marcsec <= 20 marcsec Passed minimum 9.359 marcsec <= 40 marcsec Failed stretch 22.36 marcsec <= 10 marcsec validate_drp.AD3_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 No more than AF3 of the astrometric distances will deviate by more than this from the median (see AM3, AF3). Skipped AD3.design no measurement Skipped AD3.minimum no measurement Skipped AD3.stretch no measurement validate_drp.AF1_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum fraction of astrometric distances which deviate by more than AD1 milliarcsec (see AM1). Passed design 5.315 % <= 10 % Passed minimum 0.991 % <= 20 % Failed stretch 15.81 % <= 5 % validate_drp.AF2_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum fraction of astrometric distances which deviate by more than AD2 milliarcsec (see AM2). Passed design 6.658 % <= 10 % Passed minimum 1.198 % <= 20 % Failed stretch 18.84 % <= 5 % validate_drp.AF3_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum fraction of astrometric distances which deviate by more than AD3 milliarcsec (see AM3). Skipped AF3.design no measurement Skipped AF3.minimum no measurement Skipped AF3.stretch no measurement validate_drp.AM1 - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum rms of the astrometric distance distribution for stellar pairs with separations of D= 5 arcmin (repeatability). Passed design 7.914 marcsec <= 10 marcsec Passed minimum 7.914 marcsec <= 20 marcsec Failed stretch 7.914 marcsec <= 5 marcsec validate_drp.AM2 - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum rms of the astrometric distance distribution for stellar pairs with separations of D= 20 arcmin (repeatability). Passed design 8.937 marcsec <= 10 marcsec Passed minimum 8.937 marcsec <= 20 marcsec Failed stretch 8.937 marcsec <= 5 marcsec validate_drp.AM3 - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum rms of the astrometric distance distribution for stellar pairs with separations of D= 200 arcmin (repeatability). Skipped AM3.design no measurement Skipped AM3.minimum no measurement Skipped AM3.stretch no measurement validate_drp.TE1 - LPM- 17 , p. 31 , http: //ls.st/lpm-17 The averaged E1, E2, and Ex residual PSF ellipticity correlations must have median absolute value less than this for separations < 1 arcmin. Failed design 0.0001075 <= 2e- 05 Failed minimum 0.0001075 <= 3e- 05 Failed stretch 0.0001075 <= 1e- 05 validate_drp.TE2 - LPM- 17 , p. 31 , http: //ls.st/lpm-17 The averaged E1, E2, and Ex residual PSF ellipticity correlations must have median absolute value less than this for separations > 5 arcmin. Passed design 9 .242e- 06 <= 5e- 05 Passed minimum 9 .242e- 06 <= 3e- 05 Passed stretch 9 .242e- 06 <= 1e- 05 Wrote plot: Cfht_output_r_validate_drp.AM1_D_5_arcmin_17.0_21.5_mag.png Wrote plot: Cfht_output_r_validate_drp.AM2_D_20_arcmin_17.0_21.5_mag.png Skipping validate_drp.AM3 -- no measurement ================================================================= r band summary ================================================================= Failed minimum 8 of 23 failed Failed design 12 of 23 failed Failed stretch 18 of 23 failed ================================================================================================================================== design level summary ================================================================= FAILED ( 12 / 23 measurements) ================================================================= validation_data_decam: Median value of the astrometric scatter - all magnitudes: 53.643 marcsec Astrometric scatter (median) - snr > 100.0 : 11.9 marcsec No stars found that are 199.0 arcmin-- 201.0 arcmin apart. ================================================================= z band metric measurements ================================================================= validate_drp.PF1_minimum_uzy - LPM- 17 , p. 21 , http: //ls.st/lpm-17 The maximum fraction of magnitudes deviating by more than PA2_minimum_uzy from mean. Failed minimum 69.55 % <= 20 % Failed design 69.55 % <= 10 % Failed stretch 81.68 % <= 5 % validate_drp.PA1 - LPM- 17 , p. 21 , http: //ls.st/lpm-17 The maximum rms of the unresolved source magnitude distribution around the mean value. Failed design 52.44 mmag <= 7.5 mmag Failed minimum 52.44 mmag <= 12 mmag Failed stretch 52.44 mmag <= 4.5 mmag validate_drp.PA2_minimum_uzy - LPM- 17 , p. 21 , http: //ls.st/lpm-17 At most PF1_minimum_uzy % of magnitudes may deviate by more than this magnitude from the mean. Failed minimum 49.74 mmag <= 22.5 mmag Failed design 57.75 mmag <= 22.5 mmag Failed stretch 73.76 mmag <= 15 mmag validate_drp.AD1_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 No more than AF1 of the astrometric distances will deviate by more than this from the median (see AM1, AF1). Passed design 16.62 marcsec <= 20 marcsec Passed minimum 9.969 marcsec <= 40 marcsec Failed stretch 22.28 marcsec <= 10 marcsec validate_drp.AD2_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 No more than AF2 of the astrometric distances will deviate by more than this from the median (see AM2, AF2). Passed design 17.09 marcsec <= 20 marcsec Passed minimum 10.41 marcsec <= 40 marcsec Failed stretch 24.44 marcsec <= 10 marcsec validate_drp.AD3_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 No more than AF3 of the astrometric distances will deviate by more than this from the median (see AM3, AF3). Skipped AD3.design no measurement Skipped AD3.minimum no measurement Skipped AD3.stretch no measurement validate_drp.AF1_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum fraction of astrometric distances which deviate by more than AD1 milliarcsec (see AM1). Passed design 6.439 % <= 10 % Passed minimum 0.5036 % <= 20 % Failed stretch 19.86 % <= 5 % validate_drp.AF2_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum fraction of astrometric distances which deviate by more than AD2 milliarcsec (see AM2). Passed design 7.503 % <= 10 % Passed minimum 0.8336 % <= 20 % Failed stretch 21.04 % <= 5 % validate_drp.AF3_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum fraction of astrometric distances which deviate by more than AD3 milliarcsec (see AM3). Skipped AF3.design no measurement Skipped AF3.minimum no measurement Skipped AF3.stretch no measurement validate_drp.AM1 - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum rms of the astrometric distance distribution for stellar pairs with separations of D= 5 arcmin (repeatability). Passed design 9.493 marcsec <= 10 marcsec Passed minimum 9.493 marcsec <= 20 marcsec Failed stretch 9.493 marcsec <= 5 marcsec validate_drp.AM2 - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum rms of the astrometric distance distribution for stellar pairs with separations of D= 20 arcmin (repeatability). Failed design 10.28 marcsec <= 10 marcsec Passed minimum 10.28 marcsec <= 20 marcsec Failed stretch 10.28 marcsec <= 5 marcsec validate_drp.AM3 - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum rms of the astrometric distance distribution for stellar pairs with separations of D= 200 arcmin (repeatability). Skipped AM3.design no measurement Skipped AM3.minimum no measurement Skipped AM3.stretch no measurement validate_drp.TE1 - LPM- 17 , p. 31 , http: //ls.st/lpm-17 The averaged E1, E2, and Ex residual PSF ellipticity correlations must have median absolute value less than this for separations < 1 arcmin. Failed design 3 .507e- 05 <= 2e- 05 Failed minimum 3 .507e- 05 <= 3e- 05 Failed stretch 3 .507e- 05 <= 1e- 05 validate_drp.TE2 - LPM- 17 , p. 31 , http: //ls.st/lpm-17 The averaged E1, E2, and Ex residual PSF ellipticity correlations must have median absolute value less than this for separations > 5 arcmin. Passed design 4 .878e- 06 <= 5e- 05 Passed minimum 4 .878e- 06 <= 3e- 05 Passed stretch 4 .878e- 06 <= 1e- 05 Wrote plot: Decam_output_z_validate_drp.AM1_D_5_arcmin_17.0_21.5_mag.png Wrote plot: Decam_output_z_validate_drp.AM2_D_20_arcmin_17.0_21.5_mag.png Skipping validate_drp.AM3 -- no measurement ================================================================= z band summary ================================================================= Failed minimum 9 of 23 failed Failed design 13 of 23 failed Failed stretch 19 of 23 failed ================================================================================================================================== design level summary ================================================================= FAILED ( 13 / 23 measurements) =================================================================

Chris Morrison [X] (Inactive) added a comment - 01/Feb/19 9:35 PM - edited validation_data_hsc: ================================================================= HSC-R band metric measurements ================================================================= validate_drp.PF1_minimum_gri - LPM- 17 , p. 21 , http: //ls.st/lpm-17 The maximum fraction of magnitudes deviating by more than PA2_minimum_gri from mean. Failed minimum 30.9 % <= 20 % Failed design 30.9 % <= 10 % Failed stretch 48.38 % <= 5 % validate_drp.PA1 - LPM- 17 , p. 21 , http: //ls.st/lpm-17 The maximum rms of the unresolved source magnitude distribution around the mean value. Failed design 14.29 mmag <= 5 mmag Failed minimum 14.29 mmag <= 8 mmag Failed stretch 14.29 mmag <= 3 mmag validate_drp.PA2_minimum_gri - LPM- 17 , p. 21 , http: //ls.st/lpm-17 At most PF1_minimum_gri % of magnitudes may deviate by more than this magnitude from the mean. Failed minimum 19.6 mmag <= 15 mmag Failed design 27.21 mmag <= 15 mmag Failed stretch 34.46 mmag <= 10 mmag validate_drp.AD1_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 No more than AF1 of the astrometric distances will deviate by more than this from the median (see AM1, AF1). Passed design 5.769 marcsec <= 20 marcsec Passed minimum 3.499 marcsec <= 40 marcsec Passed stretch 7.856 marcsec <= 10 marcsec validate_drp.AD2_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 No more than AF2 of the astrometric distances will deviate by more than this from the median (see AM2, AF2). Passed design 5.941 marcsec <= 20 marcsec Passed minimum 3.591 marcsec <= 40 marcsec Passed stretch 8.128 marcsec <= 10 marcsec validate_drp.AD3_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 No more than AF3 of the astrometric distances will deviate by more than this from the median (see AM3, AF3). Skipped AD3.design no measurement Skipped AD3.minimum no measurement Skipped AD3.stretch no measurement validate_drp.AF1_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum fraction of astrometric distances which deviate by more than AD1 milliarcsec (see AM1). Passed design 0.1762 % <= 10 % Passed minimum 0.1112 % <= 20 % Passed stretch 2.443 % <= 5 % validate_drp.AF2_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum fraction of astrometric distances which deviate by more than AD2 milliarcsec (see AM2). Passed design 0.1611 % <= 10 % Passed minimum 0.09087 % <= 20 % Passed stretch 2.672 % <= 5 % validate_drp.AF3_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum fraction of astrometric distances which deviate by more than AD3 milliarcsec (see AM3). Skipped AF3.design no measurement Skipped AF3.minimum no measurement Skipped AF3.stretch no measurement validate_drp.AM1 - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum rms of the astrometric distance distribution for stellar pairs with separations of D= 5 arcmin (repeatability). Passed design 3.646 marcsec <= 10 marcsec Passed minimum 3.646 marcsec <= 20 marcsec Passed stretch 3.646 marcsec <= 5 marcsec validate_drp.AM2 - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum rms of the astrometric distance distribution for stellar pairs with separations of D= 20 arcmin (repeatability). Passed design 3.744 marcsec <= 10 marcsec Passed minimum 3.744 marcsec <= 20 marcsec Passed stretch 3.744 marcsec <= 5 marcsec validate_drp.AM3 - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum rms of the astrometric distance distribution for stellar pairs with separations of D= 200 arcmin (repeatability). Skipped AM3.design no measurement Skipped AM3.minimum no measurement Skipped AM3.stretch no measurement validate_drp.TE1 - LPM- 17 , p. 31 , http: //ls.st/lpm-17 The averaged E1, E2, and Ex residual PSF ellipticity correlations must have median absolute value less than this for separations < 1 arcmin. Passed design 1 .261e- 05 <= 2e- 05 Passed minimum 1 .261e- 05 <= 3e- 05 Failed stretch 1 .261e- 05 <= 1e- 05 validate_drp.TE2 - LPM- 17 , p. 31 , http: //ls.st/lpm-17 The averaged E1, E2, and Ex residual PSF ellipticity correlations must have median absolute value less than this for separations > 5 arcmin. Passed design 3 .482e- 07 <= 5e- 05 Passed minimum 3 .482e- 07 <= 3e- 05 Passed stretch 3 .482e- 07 <= 1e- 05 ================================================================= HSC-I band metric measurements ================================================================= validate_drp.PF1_minimum_gri - LPM- 17 , p. 21 , http: //ls.st/lpm-17 The maximum fraction of magnitudes deviating by more than PA2_minimum_gri from mean. Failed minimum 27.07 % <= 20 % Failed design 27.07 % <= 10 % Failed stretch 42.77 % <= 5 % validate_drp.PA1 - LPM- 17 , p. 21 , http: //ls.st/lpm-17 The maximum rms of the unresolved source magnitude distribution around the mean value. Failed design 12.18 mmag <= 5 mmag Failed minimum 12.18 mmag <= 8 mmag Failed stretch 12.18 mmag <= 3 mmag validate_drp.PA2_minimum_gri - LPM- 17 , p. 21 , http: //ls.st/lpm-17 At most PF1_minimum_gri % of magnitudes may deviate by more than this magnitude from the mean. Failed minimum 18.25 mmag <= 15 mmag Failed design 25.56 mmag <= 15 mmag Failed stretch 33.33 mmag <= 10 mmag validate_drp.AD1_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 No more than AF1 of the astrometric distances will deviate by more than this from the median (see AM1, AF1). Passed design 8.35 marcsec <= 20 marcsec Passed minimum 5.101 marcsec <= 40 marcsec Failed stretch 11.36 marcsec <= 10 marcsec validate_drp.AD2_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 No more than AF2 of the astrometric distances will deviate by more than this from the median (see AM2, AF2). Passed design 8.169 marcsec <= 20 marcsec Passed minimum 4.976 marcsec <= 40 marcsec Failed stretch 11.15 marcsec <= 10 marcsec validate_drp.AD3_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 No more than AF3 of the astrometric distances will deviate by more than this from the median (see AM3, AF3). Skipped AD3.design no measurement Skipped AD3.minimum no measurement Skipped AD3.stretch no measurement validate_drp.AF1_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum fraction of astrometric distances which deviate by more than AD1 milliarcsec (see AM1). Passed design 0.5963 % <= 10 % Passed minimum 0.01295 % <= 20 % Failed stretch 6.894 % <= 5 % validate_drp.AF2_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum fraction of astrometric distances which deviate by more than AD2 milliarcsec (see AM2). Passed design 0.6144 % <= 10 % Passed minimum 0.01471 % <= 20 % Failed stretch 6.57 % <= 5 % validate_drp.AF3_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum fraction of astrometric distances which deviate by more than AD3 milliarcsec (see AM3). Skipped AF3.design no measurement Skipped AF3.minimum no measurement Skipped AF3.stretch no measurement validate_drp.AM1 - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum rms of the astrometric distance distribution for stellar pairs with separations of D= 5 arcmin (repeatability). Passed design 7.096 marcsec <= 10 marcsec Passed minimum 7.096 marcsec <= 20 marcsec Failed stretch 7.096 marcsec <= 5 marcsec validate_drp.AM2 - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum rms of the astrometric distance distribution for stellar pairs with separations of D= 20 arcmin (repeatability). Passed design 7.366 marcsec <= 10 marcsec Passed minimum 7.366 marcsec <= 20 marcsec Failed stretch 7.366 marcsec <= 5 marcsec validate_drp.AM3 - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum rms of the astrometric distance distribution for stellar pairs with separations of D= 200 arcmin (repeatability). Skipped AM3.design no measurement Skipped AM3.minimum no measurement Skipped AM3.stretch no measurement validate_drp.TE1 - LPM- 17 , p. 31 , http: //ls.st/lpm-17 The averaged E1, E2, and Ex residual PSF ellipticity correlations must have median absolute value less than this for separations < 1 arcmin. Passed design 1 .787e- 06 <= 2e- 05 Passed minimum 1 .787e- 06 <= 3e- 05 Passed stretch 1 .787e- 06 <= 1e- 05 validate_drp.TE2 - LPM- 17 , p. 31 , http: //ls.st/lpm-17 The averaged E1, E2, and Ex residual PSF ellipticity correlations must have median absolute value less than this for separations > 5 arcmin. Passed design 5 .421e- 07 <= 5e- 05 Passed minimum 5 .421e- 07 <= 3e- 05 Passed stretch 5 .421e- 07 <= 1e- 05 ================================================================= HSC-Y band metric measurements ================================================================= validate_drp.PF1_design_uzy - LPM- 17 , p. 21 , http: //ls.st/lpm-17 The maximum fraction of magnitudes deviating by more than PA2_design_uzy from mean. Failed design 38.08 % <= 10 % Failed minimum 38.08 % <= 20 % Failed stretch 55.54 % <= 5 % validate_drp.PA1 - LPM- 17 , p. 21 , http: //ls.st/lpm-17 The maximum rms of the unresolved source magnitude distribution around the mean value. Failed design 25.35 mmag <= 7.5 mmag Failed minimum 25.35 mmag <= 12 mmag Failed stretch 25.35 mmag <= 4.5 mmag validate_drp.PA2_design_uzy - LPM- 17 , p. 21 , http: //ls.st/lpm-17 At most PF1_design_uzy % of magnitudes may deviate by more than this magnitude from the mean. Failed design 45.11 mmag <= 22.5 mmag Failed minimum 33.91 mmag <= 22.5 mmag Failed stretch 54.62 mmag <= 15 mmag validate_drp.AD1_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 No more than AF1 of the astrometric distances will deviate by more than this from the median (see AM1, AF1). Passed design 8.938 marcsec <= 20 marcsec Passed minimum 5.556 marcsec <= 40 marcsec Failed stretch 12.12 marcsec <= 10 marcsec validate_drp.AD2_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 No more than AF2 of the astrometric distances will deviate by more than this from the median (see AM2, AF2). Passed design 9.81 marcsec <= 20 marcsec Passed minimum 6.051 marcsec <= 40 marcsec Failed stretch 13.33 marcsec <= 10 marcsec validate_drp.AD3_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 No more than AF3 of the astrometric distances will deviate by more than this from the median (see AM3, AF3). Skipped AD3.design no measurement Skipped AD3.minimum no measurement Skipped AD3.stretch no measurement validate_drp.AF1_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum fraction of astrometric distances which deviate by more than AD1 milliarcsec (see AM1). Passed design 0.8148 % <= 10 % Passed minimum 0.1267 % <= 20 % Failed stretch 7.993 % <= 5 % validate_drp.AF2_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum fraction of astrometric distances which deviate by more than AD2 milliarcsec (see AM2). Passed design 1.175 % <= 10 % Passed minimum 0.1828 % <= 20 % Failed stretch 9.672 % <= 5 % validate_drp.AF3_design - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum fraction of astrometric distances which deviate by more than AD3 milliarcsec (see AM3). Skipped AF3.design no measurement Skipped AF3.minimum no measurement Skipped AF3.stretch no measurement validate_drp.AM1 - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum rms of the astrometric distance distribution for stellar pairs with separations of D= 5 arcmin (repeatability). Passed design 5.823 marcsec <= 10 marcsec Passed minimum 5.823 marcsec <= 20 marcsec Failed stretch 5.823 marcsec <= 5 marcsec validate_drp.AM2 - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum rms of the astrometric distance distribution for stellar pairs with separations of D= 20 arcmin (repeatability). Passed design 6.481 marcsec <= 10 marcsec Passed minimum 6.481 marcsec <= 20 marcsec Failed stretch 6.481 marcsec <= 5 marcsec validate_drp.AM3 - LPM- 17 , p. 23 , http: //ls.st/lpm-17 The maximum rms of the astrometric distance distribution for stellar pairs with separations of D= 200 arcmin (repeatability). Skipped AM3.design no measurement Skipped AM3.minimum no measurement Skipped AM3.stretch no measurement validate_drp.TE1 - LPM- 17 , p. 31 , http: //ls.st/lpm-17 The averaged E1, E2, and Ex residual PSF ellipticity correlations must have median absolute value less than this for separations < 1 arcmin. Failed design 0.0002384 <= 2e- 05 Failed minimum 0.0002384 <= 3e- 05 Failed stretch 0.0002384 <= 1e- 05 validate_drp.TE2 - LPM- 17 , p. 31 , http: //ls.st/lpm-17 The averaged E1, E2, and Ex residual PSF ellipticity correlations must have median absolute value less than this for separations > 5 arcmin. Passed design 2 .357e- 06 <= 5e- 05 Passed minimum 2 .357e- 06 <= 3e- 05 Passed stretch 2 .357e- 06 <= 1e- 05 ================================================================= HSC-R band summary ================================================================= Failed minimum 7 of 23 failed Failed design 7 of 23 failed Failed stretch 8 of 23 failed ================================================================= ================================================================= HSC-I band summary ================================================================= Failed minimum 7 of 23 failed Failed design 7 of 23 failed Failed stretch 13 of 23 failed ================================================================= ================================================================= HSC-Y band summary ================================================================= Failed minimum 8 of 23 failed Failed design 8 of 23 failed Failed stretch 14 of 23 failed ================================================================================================================================== design level summary ================================================================= FAILED ( 22 / 69 measurements) =================================================================

Chris Morrison [X] (Inactive) added a comment - 05/Feb/19 1:48 PM - edited Jenkins run: https://ci.lsst.codes/blue/organizations/jenkins/stack-os-matrix/detail/stack-os-matrix/29389/pipeline   As part of this ticket I've also fixed the config in validate_drp to use the HTM reference catalogs. An initial ticket branch on obs_subaru turned out to not be necessary and has been removed.

Yusra AlSayyad added a comment - 11/Feb/19 3:07 PM Looks OK, but I didn't exactly follow what was going on here:  https://github.com/lsst/meas_astrom/pull/111/commits/990c7b45593fd80732d0595633d4ed9ce42753f1#diff-97929b2006e789cef1b1beb342333037R327 Why does fitting one last time fix things? Because the distances on the matches reflect the final iteration now?  Also will you double check your commit messages on the meas_astrom PR?  My eyeballs pattern-matched "interative" with interactive rather than "iterative." PessB already has a final verify step, so I was a little confused about the description that you were adding the final verify step in this commit. 

Chris Morrison [X] (Inactive) added a comment - 11/Feb/19 3:18 PM Basically it's to make the final matches more consistent across the chip. Rather than using just 6 points to fit the shift/rotation matrix it uses the full set of matches to fit a more consistent shift/rotation matrix. This basically prevents matches that should be used and fed to the astrometry from being clipped out because the match is far from where the shift/ration was fit. I'm happy to change the wording if needed if you have a suggestion.

John Swinbank added a comment - 12/Feb/19 6:32 PM Chris Morrison [X] , Yusra AlSayyad — I reckon this is a blocker for pipelines release v17. Do you concur?

Chris Morrison [X] (Inactive) added a comment - 13/Feb/19 1:27 PM I've updated the comment in the ticket. I'm waiting on either John Swinbank or yourself, Yusra AlSayyad to give it a thumbs up. I'm ready to merge once that happens.

John Swinbank added a comment - 13/Feb/19 4:02 PM - edited Chris Morrison [X] — a couple of things: Could you take another pass through the commit messages? I still see an “interative” there, as well as a couple of “explaination”s. I actually think the comment you've written here on this ticket is more informative than the one you added to the source. Do you think you could adapt it? Something like: The shift/rotation matrix returned by ``_construct_pattern_and_shift_rot_matrix``, above, was based on only six points. Here, we refine that result by using all of the good matches from the “final verification” step, above. This will produce a more consistent result. perhaps?

Chris Morrison [X] (Inactive) added a comment - 13/Feb/19 4:48 PM Done, sorry I forgot about the typo in the commit. I double checked and the Cfht values are the same as I reported above. Let me know when I'm clear to merge.

John Swinbank added a comment - 13/Feb/19 5:27 PM . Go for it.

Chris Morrison [X] (Inactive) added a comment - 13/Feb/19 5:42 PM Fingers crosses that the squash metrics improve as I have seen. Talked to Simon Krughoff this morning and he said that squash will run with this version tomorrow morning.