Difference between revisions of "HFI-Validation"

From Planck PLA Wiki
Jump to: navigation, search
Line 6: Line 6:
 
==Top-down approach to systematics==
 
==Top-down approach to systematics==
  
==Yardstick simulations (including a Sisyphe summary of what can be neglected) (Deluis/Montier)==
+
==Yardstick simulations (including a Sisyphe summary of what can be neglected)==
 +
 
 +
===Yardstick simulations (Delouis)===
 +
The yardstick allows gauging various effects to see whether they need be included in monte-carlo to describe data. It also allows gauging the significance of validation tests on data (e.g. can null test can be described by the model?).
 +
 +
Yardstick 3.0 that characterizes the DX9 data goes through the following steps:
 +
 
 +
#The input maps are computed using the Planck Sky Model.
 +
#The LevelS is used to project input maps on timeline using the B-Spline scanning beam and the DX9 pointing (called ptcor6). The real pointing is affected by the aberration that is corrected by map-making. Yardstick does not simulate aberration. Finally, the difference between the projected pointing from simulation and from DX9 is equal to the aberration.
 +
#The simulated noise timelines, that are added to the projected signal, have the same spectrum (low and high frequency) than the characterized noise. For yardstick 3.0. No correlation in time or between detectors have been simulated.
 +
#The simulation map making step use the DX9 sample flags.
 +
#For the low frequencies (100, 143, 217, 353), the yardstick output are calibrated using the same mechanism (e.g. dipole fitting) than DX9. The calibration is not done for higher frequency (545, 857)
 +
#The Official map making is run on those timelines using the same parameters than for real data.
 +
A yardstick production is composed by all survey map (1,2 and nominal), all detector Detsets and channel maps. The Yardstick 3.0 is based on 5 noise iterations for each map realization.
 +
 
 +
===Sysiphe summary (Montier)===
  
 
==Simulations versus data results (including PTE) (Techene)==
 
==Simulations versus data results (including PTE) (Techene)==
 +
We make consistency test between DX9 and Yardstick production. Yardstick production contains
 +
sky (generated with LevelS starting from PSM177) and noise timeline realizations procedeed with
 +
the official map making. DX9 production was regenerated in order to get rid of possible differences
 +
that might appear for not running the official pipeline in the same conditions. We compare statistical
 +
properties of cross spectra of null test maps for 100, 143, 217, 353 channels. Null test maps can be
 +
survey null test or half focal plane null test, each of them have a specific goal : survey1-survey2
 +
aims at isolating transfert function or pointing issues, while half focal plane null test enables to
 +
focus on beam issues. Comparing cross spectra we isolate systematic effects from the noise, and we
 +
can check whether they are properly simulated or need to. Spectra are computed with spice
 +
masking either DX9 point sources or simulated point sources, and masking the galactic plane with
 +
several mask width, the sky fraction from which spectra are computed are around 30%, 60% and
 +
80%.
 +
Example of mask fsky=30% :
 +
figure
 +
DX9 and the Y3.0 iterations are binned using : '/data/dmc/MISS03/DATA/BINTAB/bin_v4.7_tt'.
 +
For each bin we compute the statisical parameters of the Yardstick distribution. The following plot
 +
shows the differences between Y3.0 mean and DX9 considering the standart deviation of the
 +
yardstick. We also indicate chi square value, Chi^2 is computed over larger bin : [0,20][20,400]
 +
[400,1000][1000,2000][2000, 3000], using the ratio between (DX9-Y3.0 mean)^2 and Y3.0
 +
variance within each bin.
 +
Exemple of consistency test for 143 survey null test maps :
 +
figure

Revision as of 16:39, 18 October 2012

The HFI validation is mostly modular. That is, part of the pipeline, be it timeline processing, map-making, or other, validates the results of its work.


Expected systematics and tests (bottom-up approach)[edit]

Top-down approach to systematics[edit]

Yardstick simulations (including a Sisyphe summary of what can be neglected)[edit]

Yardstick simulations (Delouis)[edit]

The yardstick allows gauging various effects to see whether they need be included in monte-carlo to describe data. It also allows gauging the significance of validation tests on data (e.g. can null test can be described by the model?).

Yardstick 3.0 that characterizes the DX9 data goes through the following steps:

  1. The input maps are computed using the Planck Sky Model.
  2. The LevelS is used to project input maps on timeline using the B-Spline scanning beam and the DX9 pointing (called ptcor6). The real pointing is affected by the aberration that is corrected by map-making. Yardstick does not simulate aberration. Finally, the difference between the projected pointing from simulation and from DX9 is equal to the aberration.
  3. The simulated noise timelines, that are added to the projected signal, have the same spectrum (low and high frequency) than the characterized noise. For yardstick 3.0. No correlation in time or between detectors have been simulated.
  4. The simulation map making step use the DX9 sample flags.
  5. For the low frequencies (100, 143, 217, 353), the yardstick output are calibrated using the same mechanism (e.g. dipole fitting) than DX9. The calibration is not done for higher frequency (545, 857)
  6. The Official map making is run on those timelines using the same parameters than for real data.

A yardstick production is composed by all survey map (1,2 and nominal), all detector Detsets and channel maps. The Yardstick 3.0 is based on 5 noise iterations for each map realization.

Sysiphe summary (Montier)[edit]

Simulations versus data results (including PTE) (Techene)[edit]

We make consistency test between DX9 and Yardstick production. Yardstick production contains sky (generated with LevelS starting from PSM177) and noise timeline realizations procedeed with the official map making. DX9 production was regenerated in order to get rid of possible differences that might appear for not running the official pipeline in the same conditions. We compare statistical properties of cross spectra of null test maps for 100, 143, 217, 353 channels. Null test maps can be survey null test or half focal plane null test, each of them have a specific goal : survey1-survey2 aims at isolating transfert function or pointing issues, while half focal plane null test enables to focus on beam issues. Comparing cross spectra we isolate systematic effects from the noise, and we can check whether they are properly simulated or need to. Spectra are computed with spice masking either DX9 point sources or simulated point sources, and masking the galactic plane with several mask width, the sky fraction from which spectra are computed are around 30%, 60% and 80%. Example of mask fsky=30% : figure DX9 and the Y3.0 iterations are binned using : '/data/dmc/MISS03/DATA/BINTAB/bin_v4.7_tt'. For each bin we compute the statisical parameters of the Yardstick distribution. The following plot shows the differences between Y3.0 mean and DX9 considering the standart deviation of the yardstick. We also indicate chi square value, Chi^2 is computed over larger bin : [0,20][20,400] [400,1000][1000,2000][2000, 3000], using the ratio between (DX9-Y3.0 mean)^2 and Y3.0 variance within each bin. Exemple of consistency test for 143 survey null test maps : figure

(Planck) High Frequency Instrument