Action #1799

Feature #1517: Add support for Fermi/LAT Pass 8 response functions

Test Fermi/LAT Pass8 response handling

Added by Knödlseder Jürgen over 7 years ago. Updated over 7 years ago.

Status:ClosedStart date:06/21/2016
Priority:NormalDue date:
Assigned To:Knödlseder Jürgen% Done:

100%

Category:-
Target version:1.1.0
Duration:

Description

The Pass8 Fermi/LAT response handling has to be compared between Fermi Science Tools and GammaLib/ctools.


Recurrence

No recurrence.

History

#1 Updated by Knödlseder Jürgen over 7 years ago

Here the reference gtlike result for a 5 degree ROI around the Crab position, including survey data until 1 June 2016:

<===> Crab:
<===> Integral: 29.24 +/- 0.0678804
<===> Index: -2.2446 +/- 0.0018335
<===> LowerLimit: 100
<===> UpperLimit: 500000
<===> TS value: 731874
<===> Flux: 5.52719e-06 +/- 1.59659e-08 photons/cm^2/s
<===> 
<===> Extragal_diffuse:
<===> Normalization: 2.66561 +/- 0.0232332
<===> Flux: 0.000784169 +/- 6.82459e-06 photons/cm^2/s
<===> 
<===> Galactic_diffuse:
<===> Value: 0.914973 +/- 0.00273087
<===> MapBase::readFitsFile: creating WcsMap2 object
<===> Flux: 0.000586389 +/- 1.75032e-06 photons/cm^2/s
<===> WARNING: Fit may be bad in range [60, 463.345] (MeV)
<===> WARNING: Fit may be bad in range [549.393, 5964.79] (MeV)
<===> WARNING: Fit may be bad in range [13979.5, 16575.7] (MeV)
<===> WARNING: Fit may be bad in range [19654, 27631.8] (MeV)
<===> WARNING: Fit may be bad in range [32763.4, 38848] (MeV)
<===> WARNING: Fit may be bad in range [253012, 300000] (MeV)
<===> 
<===> 
<===> Total number of observed counts: 1.24021e+06
<===> Total number of model events: 1.24021e+06
<===> 512  1249847.395  1240212.035
<===> 
<===> -log(Likelihood): -1249847.395
<===> 
<===> Writing fitted model to /users-data/knodlseder/glast/analysis/gammalib/proc/pass8_crab_roi05_60_300000_50bin_extmod_galmod_crab/results_like.xml
<===> Elapsed CPU time: 418.15

#2 Updated by Knödlseder Jürgen over 7 years ago

  • Parent task set to #1517

#3 Updated by Knödlseder Jürgen over 7 years ago

Here the corresponding ctlike run:

2016-06-21T15:07:42: +=========================================+
2016-06-21T15:07:42: | Maximum likelihood optimisation results |
2016-06-21T15:07:42: +=========================================+
2016-06-21T15:07:42: === GOptimizerLM ===
2016-06-21T15:07:42:  Optimized function value ..: -1249868.760
2016-06-21T15:07:42:  Absolute precision ........: 0.005
2016-06-21T15:07:42:  Acceptable value decrease .: 2
2016-06-21T15:07:42:  Optimization status .......: converged
2016-06-21T15:07:42:  Number of parameters ......: 13
2016-06-21T15:07:42:  Number of free parameters .: 4
2016-06-21T15:07:42:  Number of iterations ......: 9
2016-06-21T15:07:42:  Lambda ....................: 1e-12
2016-06-21T15:07:42:  Maximum log likelihood ....: 1249868.760
2016-06-21T15:07:42:  Observed events  (Nobs) ...: 1240212.000
2016-06-21T15:07:42:  Predicted events (Npred) ..: 1240211.999 (Nobs - Npred = 0.0014539)
2016-06-21T15:07:42: === GModels ===
2016-06-21T15:07:42:  Number of models ..........: 3
2016-06-21T15:07:42:  Number of parameters ......: 13
2016-06-21T15:07:42: === GModelSky ===
2016-06-21T15:07:42:  Name ......................: Extragal_diffuse
2016-06-21T15:07:42:  Instruments ...............: all
2016-06-21T15:07:42:  Instrument scale factors ..: unity
2016-06-21T15:07:42:  Observation identifiers ...: all
2016-06-21T15:07:42:  Model type ................: DiffuseSource
2016-06-21T15:07:42:  Model components ..........: "ConstantValue" * "FileFunction" * "Constant" 
2016-06-21T15:07:42:  Number of parameters ......: 3
2016-06-21T15:07:42:  Number of spatial par's ...: 1
2016-06-21T15:07:42:   Value ....................: 1 [0,10]  (fixed,scale=1,gradient)
2016-06-21T15:07:42:  Number of spectral par's ..: 1
2016-06-21T15:07:42:   Normalization ............: 2.71162 +/- 0.0227147 [0,1000]  (free,scale=1,gradient)
2016-06-21T15:07:42:  Number of temporal par's ..: 1
2016-06-21T15:07:42:   Normalization ............: 1 (relative value) (fixed,scale=1,gradient)
2016-06-21T15:07:42: === GModelSky ===
2016-06-21T15:07:42:  Name ......................: Galactic_diffuse
2016-06-21T15:07:42:  Instruments ...............: all
2016-06-21T15:07:42:  Instrument scale factors ..: unity
2016-06-21T15:07:42:  Observation identifiers ...: all
2016-06-21T15:07:42:  Model type ................: DiffuseSource
2016-06-21T15:07:42:  Model components ..........: "MapCubeFunction" * "ConstantValue" * "Constant" 
2016-06-21T15:07:42:  Number of parameters ......: 3
2016-06-21T15:07:42:  Number of spatial par's ...: 1
2016-06-21T15:07:42:   Normalization ............: 1 [0.001,1000]  (fixed,scale=1,gradient)
2016-06-21T15:07:42:  Number of spectral par's ..: 1
2016-06-21T15:07:42:   Value ....................: 0.907689 +/- 0.00268146 [0,1000]  (free,scale=1,gradient)
2016-06-21T15:07:42:  Number of temporal par's ..: 1
2016-06-21T15:07:42:   Normalization ............: 1 (relative value) (fixed,scale=1,gradient)
2016-06-21T15:07:42: === GModelSky ===
2016-06-21T15:07:42:  Name ......................: Crab
2016-06-21T15:07:42:  Instruments ...............: all
2016-06-21T15:07:42:  Instrument scale factors ..: unity
2016-06-21T15:07:42:  Observation identifiers ...: all
2016-06-21T15:07:42:  Model type ................: PointSource
2016-06-21T15:07:42:  Model components ..........: "SkyDirFunction" * "PowerLaw2" * "Constant" 
2016-06-21T15:07:42:  Number of parameters ......: 7
2016-06-21T15:07:42:  Number of spatial par's ...: 2
2016-06-21T15:07:42:   RA .......................: 83.6331 [-360,360] deg (fixed,scale=1)
2016-06-21T15:07:42:   DEC ......................: 22.0145 [-90,90] deg (fixed,scale=1)
2016-06-21T15:07:42:  Number of spectral par's ..: 4
2016-06-21T15:07:42:   Integral .................: 2.88621e-06 +/- 6.62962e-09 [1e-14,0.0001] ph/cm2/s (free,scale=1e-07,gradient)
2016-06-21T15:07:42:   Index ....................: -2.23854 +/- 0.00175264 [-5,5]  (free,scale=1,gradient)
2016-06-21T15:07:42:   LowerLimit ...............: 100 [10,1e+06] MeV (fixed,scale=1)
2016-06-21T15:07:42:   UpperLimit ...............: 500000 [10,1e+06] MeV (fixed,scale=1)
2016-06-21T15:07:42:  Number of temporal par's ..: 1
2016-06-21T15:07:42:   Normalization ............: 1 (relative value) (fixed,scale=1,gradient)
2016-06-21T15:07:42: 
2016-06-21T15:07:42: +==============+
2016-06-21T15:07:42: | Save results |
2016-06-21T15:07:42: +==============+
2016-06-21T15:07:42: Save fitted parameters into file "crab_results.xml".
2016-06-21T15:07:42: 
2016-06-21T15:07:42: Application "ctlike" terminated after 23 wall clock seconds, consuming 12.1102 seconds of CPU time.

#4 Updated by Knödlseder Jürgen over 7 years ago

And the comparison:
Parameter gtlike ctlike
-logL -1249847.395 -1249868.760
Observed events 1240212.000
Modelled events 1240212.035 1240211.999
Crab Integral 29.24 +/ - 0.06788 28.86 +/ - 0.06630
Crab Index -2.2446 +/ - 0.001834 -2.2385 +/ - 0.001753
TS 731874 731672
Ext. Gal. Norm. 2.6656 +/- 0.0232 2.7116 +/- 0.0227
Gal. Value 0.9150 +/- 0.0027 0.9077 +/- 0.0027
CPU time 418.15 sec 12.11 sec

This looks pretty close, ctlike is however considerably faster than gtlike.

#5 Updated by Knödlseder Jürgen over 7 years ago

  • Status changed from New to Closed
  • % Done changed from 0 to 100

Check satisfactory. Of course, more checks are needed, but we leave this for now to the user.

Also available in: Atom PDF