alignlog.txt

********************* ALIGNMENT LOG ************************
Iteration: 7
This is only the output from the 7th iteration.

Total number of events: 25000
Time of first event [ns]: 1271032271726848000 --> 2010-04-12 02:31:11 AM
Time of last event [ns] : 1271032431768064000 --> 2010-04-12 02:33:51 AM
Time at initialize [ns] : 1271032271726848000 --> 2010-04-12 02:31:11 AM
General Info about the input data. The time at initialize is the time used to configure the geometry. You want that to correspond more or less to the time of the first event. You can specify the time at initialize in your job option file using EventClockSvc().InitialTime. The easiest is to run your job once, check the time of the first event, then copy it into your joboption file.
In addition you should make sure that you use events that have all been taken with the same geometry. The TAlignmentManual contains more details about this.

Used 0 vertices for alignment
Used 0 J/Psis for alignment
Used 48297 tracks for alignment
Total chisquare/dofs:    1.80535e+06 / 598179
Average track chisquare: 37.3801
Track chisquare/dof:     3.01807
Vertex chisquare/dof:    0
Used 167960 hits for alignment
Total number of hits in external detectors: 623407
Total chisquare of survey constraints: 740.585 / 30
More details on the input data. The most interesting numbers are:
  • Used tracks for alignment this should increase during alignment since you find new tracks with better aligned detectors.
  • Total chisquare/dofs is the total chi-square of all constraints in the alignment, e.g. tracks, vertices (eventually with mass constraints), survey constraints and 'exact' (lagrange) constraints. This is the number minimized by the alignment procedure.

Most important dofs: 
  0 TT/TTb/TTbXLayer dof=5 chi2= 3462.45 delta= 0.000139063 +/- 2.3633e-06
  1 TT/TTa/TTaXLayer dof=5 chi2= 1503.36 delta= -9.47035e-05 +/- 2.4425e-06
  2 TT/TTa/TTaULayer dof=5 chi2= 714.503 delta= -7.15192e-05 +/- 2.6756e-06
  3 TT/TTa/TTaULayer dof=3 chi2= 506.309 delta= -0.000656166 +/- 2.91612e-05
  4 TT dof=5 chi2= 443.47 delta= 0.000135629 +/- 6.44052e-06
  5 TT/TTb/TTbVLayer dof=5 chi2= 117.091 delta= 2.76778e-05 +/- 2.55781e-06
  6 TT/TTa/TTaXLayer dof=4 chi2= 108.968 delta= 0.000143247 +/- 1.37226e-05
  7 TT dof=3 chi2= 87.8282 delta= -0.000285295 +/- 3.04423e-05
  8 TT/TTb/TTbVLayer dof=4 chi2= 56.8524 delta= 0.000124644 +/- 1.6531e-05
  9 TT/TTb/TTbVLayer dof=3 chi2= 48.3312 delta= 0.000189153 +/- 2.72082e-05

These lines tell you about which DOFs give the largest chi2 contribution before performing the alignment update and ignoring any correlation with other DOFs. The result may help you to determine which DOFs to add to the alignment. For example, the first line tells that if we align for the Rz rotation of a particular TT layer, the total chi2 will improve by 3400 units. That's very significant (many, many sigmas) but it is small compared to the total chi2. In other words, even if you add this DOF, you will hardly notice it in track chi2 values.

The plotted 'delta' is the so-called 'local' delta, ignoring the correlation with other parameters. Therefore, the actual delta applied by the alignment may come out differently.

Number of alignables with insufficient statistics: 0
Number of lagrange constraints: 0
Number of chisq constraints:    30
Number of active parameters:    11
Alignment delta chisquare/dof: 4.38129 / 11
Normalised alignment change chisquare: 0.398299
Alignment delta chisquare/track dof: 7.32437e-06
Canonical constraint values: 
Canonical constraint chisquare: 0

Always watch the Number of alignables with insufficient statistics entry: this will tell you if you have active alignment elements with too few hits to allow an update of the parameters. The "Alignment delta chisquare/dof" is the actual improvement in the total chi2. When the number is smaller than 1 per dof, your alignment has 'statistically converged'. When normalized to the number of dofs of the tracks, the delta chi2 tells you by how much you can expect your track chisquares to improve. In this case the alignment has converged.

Values of constraint equations after solution (X=active,O=inactive): 
  Tx  O -0.000426204 +/- 0.149257
  Ty  O   -0.0361868 +/- 0.41145
  Tz  O  0.000596331 +/- 0.149848
  Rx  O            0 +/- 0
  Ry  O            0 +/- 0
  Rz  O            0 +/- 0
 Szx  O   0.00213932 +/- 0.0257609
 Szy  O  2.68973e-05 +/- 0.000305826
 Szz  O  -0.00775017 +/- 0.0048359
 Sxx  O  -0.00019671 +/- 0.12438
 SRz  O            0 +/- 0
 Trx  O -1.07513e-05 +/- 1.82762e-05
 Try  O -3.75917e-06 +/- 5.6304e-06
Trtx  O  9.98876e-08 +/- 1.70835e-07
Trty  O  4.25802e-08 +/- 7.17771e-08
Trcur  O            0 +/- 0
 PVx  O            0 +/- 0
 PVy  O            0 +/- 0
 PVz  O            0 +/- 0
Sz2x  O  6.89491e-06 +/- 0.0375677
Sz2y  O -1.99925e-08 +/- 2.27318e-07

These lines tell you the status of the 'default' lagrange constraints. The second column shows a '0' in all cases, which means that the constraints were not active. The third column shows the expected change in the constraint value as a result of the alignment update if the constraint would not be applied. For example

 Trx  O -1.07513e-05 +/- 1.82762e-05
means that the change in the average track x coordinate as a result of the alignment update is -0.01 micron and consistent with zero.

Alignable: TT
Global position: (0,0,2485)
Number of tracks/hits/outliers seen: 164807 167960 0
Local-to-global diagonal: 1 1 1 1 1 1 
Tx  : curtot= -0.2603      cur= -0.2603      delta= -0.0004098   +/- 0.04976      gcc= 0.9999
Ty  : curtot= -7.653       cur= -7.653       delta= -0.03619     +/- 0.4114       gcc= 1
Tz  : curtot= 0.2341       cur= 0.2341       delta= 0.0005734    +/- 0.05137      gcc= 0.9709
contribution to hit error (absolute/relative): 0.03105 0.4558
local delta chi2 / dof: 140.9 / 3
survey chi2 / dof (before/after): 234.8 237 / 3
survey pars:   0, 0, 0, 0, 0, 0
survey errors: 0.5, 0.5, 0.5, 0.001, 0.001, 0.001

Output for Individual Alignable

Alignable: TT/TTa/TTaULayer
Output for the TTaU Layer

Global position: (-0.2603,-7.653,2373)
Position before aligning this element but including changes of parent elements.

Number of tracks/hits/outliers seen: 41176 42206 0
Number of tracks going through this alignable and the number of hits in the alignable in the alignment. The number of outliers is always zero. (We can reject outliers in the alignment but this is a bit tricky.) If the nuber of hits is smaller than the threshold specified in TAlignment().MinNumberOfHits, then the parameters will not be updated.

Local-to-global diagonal: 1 1 1 1 1 1
These are the diagonal elements of the rotation matrix from the local to the global frame. Some alignment elements (in particular IT boxes) have coordinates reversed. This makes it hard to see correlations when plotting. Therefore, these numbers are useful to put the delta's (more-or-less) in the same frame.

Tx : curtot=      -0.8613 cur=      -0.6006 delta=    -0.003013 +/-      0.06142 gcc= 1
Tz : curtot=        1.041 cur=       0.8067 delta=     0.006212 +/-      0.04999 gcc= 0.9921
Summery for the individual degrees of freedom:
  • Tx degree of freedom
  • curtot global misalignment with respect to nominal (includes parents misalignment) before this iteration
  • cur local misalignment (nominal would be 0) before this iteration
  • delta change of cur due to this iteration
  • +/- error on cur
    The error strongly depends on the correlations (see next point). For strongly correlated degrees of freedom the error is dominated by survey constraints.
  • gcc global correlation coefficient: correlation of this parameter with other parameters.
    A number very close to one indicates that you have 'weak modes'. In this case both degrees of freedom are of course correlated with the respective degrees for the whole station. Therefore only the survey constraints add independent information. This the small difference from 1 seen in Tz. The difference should get smaller with increased statistics as then the survey constrains get a smaller relative weight.

contribution to hit error (absolute/relative): 0.06064 0.8905
This tells how much the statistical uncertainty in this alignment would affect your hit resolution, either absolute (e..g in mm) or relative, normalized to the hit resolution. In this case, the uncertainty is relatively large, probably due to the large correlation in Tx.

local delta chi2 / dof: 56.03 / 2
This is the delta-chi2 of the alignment if we'd ignore all correlations with other elements. (Relevant for the so-called 'local method' for ali8gnment.)

survey chi2 / dof (before/after): 101.2 102.5 / 2
The total chi2 of the survey constraints of this alignable before and after the alignment update.

survey pars:   0, 0, 0, 0, 0, 0
The survey parameters in the alignment frame of this alignable (Tx, Ty, Tz, Rx, Ry, Rz)

survey errors: 0.1, 0.1, 0.1, 0.0005, 0.0005, 0.0005
Survey errors used to constrain the position of this element, again in the alignment frame.




Alignable: TT/TTa/TTaXLayer
Global position: (-0.2603,-7.653,2328)
Number of tracks/hits/outliers seen: 43253 44083 0
Local-to-global diagonal: 1 1 1 1 1 1 
 Tx : curtot=      -0.1605 cur=       0.1002 delta=    1.786e-05 +/-      0.04975 gcc= 1
 Tz : curtot=        1.476 cur=        1.241 delta=     0.003536 +/-      0.04993 gcc= 0.9941
contribution to hit error (absolute/relative): 0.04951 0.7269
local delta chi2 / dof: 0.4264 / 2
survey chi2 / dof (before/after): 155.1 156 / 2
survey pars:   0, 0, 0, 0, 0, 0
survey errors: 0.1, 0.1, 0.1, 0.0005, 0.0005, 0.0005
Alignable: TT/TTb/TTbVLayer
Global position: (-0.2603,-7.653,2598)
Number of tracks/hits/outliers seen: 40670 41393 0
Local-to-global diagonal: 1 1 1 1 1 1 
 Tx : curtot=       0.3113 cur=       0.5719 delta=     0.002923 +/-      0.06142 gcc= 1
 Tz : curtot=      -0.9888 cur=       -1.223 delta=    -0.007698 +/-      0.04999 gcc= 0.9916
contribution to hit error (absolute/relative): 0.06069 0.8914
local delta chi2 / dof: 47.68 / 2
survey chi2 / dof (before/after): 182.4 184.6 / 2
survey pars:   0, 0, 0, 0, 0, 0
survey errors: 0.1, 0.1, 0.1, 0.0005, 0.0005, 0.0005
Alignable: TT/TTb/TTbXLayer
Global position: (-0.2603,-7.653,2643)
Number of tracks/hits/outliers seen: 39708 40278 0
Local-to-global diagonal: 1 1 1 1 1 1 
 Tx : curtot=      -0.3426 cur=      -0.0819 delta=    5.553e-05 +/-      0.04975 gcc= 1
 Tz : curtot=      -0.5806 cur=      -0.8153 delta=    -0.002028 +/-      0.04995 gcc= 0.993
contribution to hit error (absolute/relative): 0.0495 0.7258
local delta chi2 / dof: 0.1357 / 2
survey chi2 / dof (before/after): 67.15 67.48 / 2
survey pars:   0, 0, 0, 0, 0, 0
survey errors: 0.1, 0.1, 0.1, 0.0005, 0.0005, 0.0005

total local delta chi2 / dof: 245.1 / 11
********************* END OF ALIGNMENT LOG ************************
Edit | Attach | Watch | Print version | History: r6 < r5 < r4 < r3 < r2 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r6 - 2010-08-19 - NicolaChiapolini
 
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    LHCb All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2019 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback