Update (16 dec 2008) here

Ped deviation monitoring (10/03/2008)

The following plots show the evolution with time of the pedestal deviation for LEX1,HEX1,LEX8 and HEX8 .I use the root file CALPEDSANALYZER and canvas CalXAdcPedPedMeanDeviation_LEX1_TH1 ( LEX1,HEX1,LEX8 and HEX8).
I put the data into a histogram and fit with a Gaussian (see plot). The reported value is the mean of the gaussian (with 3 parameters, N ,mu, sigma) ( which give the peak of the distribution) and the error bar is the width of the Gaussian sigma

the time range is more or less 30 days


we see a clear trend that could explain the trend is FLE and FHE threshold.
h2 updated plots (10/04/2008)

FLE threshold

FHE threshold

LAC pos threshold

LAC neg threshold

LAC monitoring

• The following plots show the evolution, for POS (top) and NEG (bottom), of the LAC distribution mean value and RMS from the RECONHISTALARMDIST.Lac_Thresholds_FaceNeg_TH1_TowerCalLayerCalColumn_leftmost_edge_TH1 and RECONHISTALARMDIST.Lac_Thresholds_FacePos_TH1_TowerCalLayerCalColumn_leftmost_edge_TH1 plots with the moot key=2419, 2321 and 2417(most recent NomSciOps definition).

• The mean value is the centroid
• The RMS is represented by the error bar
• The RED points represents the value of the center of the highest bin in the LAC histogram
• The GREEN points represents the value of the center of the lowest bin in the LAC histogram

generated with /a/surrey01/vol/vol2/g.USAWork1/berrie/glast/flight/lachist_v2.py
The mean value shows a slight increase with time
When zooming the points, there is a very significant oscillation of very small amplitude with a ~1day period.

• The code can run in a daily cron job for instance since it takes approximately a minute to run right now.

• The xrootd list is retrieved in the script (thanks Johan & David!) using

  /afs/slac.stanford.edu/u/gl/glast/datacatalog/prod/datacat find --group RECONHISTALARMDIST --site SLAC_XROOT /Data/Flight/Level1/LPA --sort 'nRun' --filter 'nMootKey==2419||nMootkey==2321||nMootKey==2417'

FLE and FHE monitoring

We used the RECONHIST file to monitor the FLE and FHE threshold, downloaded by the following command :

/afs/slac.stanford.edu/u/gl/glast/datacatalog/prod/datacat find --group RECONHIST --site SLAC_XROOT /Data/Flight/Level1/LPA --sort 'nRun'  --filter '(nRun>"+str(MetStart)+"&&nRun<"+str(MetStop)+")&&nMootKey==2419||nMootKey==2321||nMootKey==2417)'

Fitting procedure

For FLE, we fit the 'SuspCalLo_Highest0_EnergyDistribution_TH1' histogram with the following function :

f1 = ROOT.TF1("f1","([2]/pow(x,[3]))/(1+exp(([0]-x)/[4]))+[1]",30,200)

For FHE, we fit the 'SuspCalHi_Highest0_EnergyDistribution_TH1' histogram with the following function :

f1 = ROOT.TF1("f1","([2]/pow(x,[3]))/(1+exp(([0]-x)/[4]))+[1]",300,2000)

Results

Location:

sanchez@noric07 $ pwd
/a/sulky36/g.glast.u33/sanchez/flight/CAL_mon/flefhemon
sanchez@noric07 $ ls *eps *dat
	    fhemon.eps	    flemon.dat	flemon_all.eps
fhemon.dat  fhemon_all.eps  flemon.eps

The following plots show the evolution of the fitted parameter [0] as a function of time for both FLE and FHE. This parameter is likely to described the mean FLE (FHE) threshold.

• The red cross are the obtained values and the error bars are the errors given by MINUIT
• First plot is the fitted FLE threshold as a function of time and the second plot is the same but for FHE threshold
  

• FLE : The mean value is 97.5 Mev not 100MeV : due to the method?
• FHE : mean is > 1000 MeV
  We clearly see an oscillation in both plots

TODO

• Check the result with the CALCalib code