Hi,
just a quick reminder: we are calculating S/P correction for data
enriched using this formula:
S/P_MC ENRICHED
S/P_DATA_ENRICHED= ------------------ * S/P_DATA DEPLETED
S/P_MC DEPLETED
The set of corrections we have determined so far has been computed on
mes fits with Crystal Ball Shape parameters fixed to MC Enriched Sample,
and also signal function with parameters determined on enriched Sample.
This set of correction has posted some days ago.
#mx_l mx_h corr err_corr
0.00 1.55 1.499 +- 0.495
1.55 1.90 2.688 +- 0.655
1.90 2.20 1.801 +- 0.296
2.20 2.50 1.896 +- 0.611
2.50 2.80 1.165 +- 0.468
2.80 3.10 0.637 +- 0.311
3.10 3.40 19.367+- 34.585
3.40 3.70 1.524 +- 1.610
3.70 4.20 8.180 +- 31.833
4.20 5.00 0.555 +- 6.639
In order to find a better set of correction factors, to fix the huge
errors on bin 3.10 < mX < 3.40 and 3.70 < mX < 4.20, I have tried the
following:
- determine the PDF parameters on MC-enriched sample with lepton cut
and use them to fit MC-enriched sample ALL CUTS;
- determine the PDF parameters on MC-Depleted with lepton cut and
use them to fit MC-depleted sample ALL CUTS and data-depleted ALL
CUTS.
- both set of parameters have been determined by fitting non
truth-matched events for bkg and counting truth-matched events for signal.
- when fitting the mes distribution in kinematic bins for MC bkg all
parameters have been kept fixed except for argus shape;
- when fitting the mes distribution in kinematic bins for data depleted
all parameters have been kept fixed except for argus shape, cutoff
(CCB and Argus), SigmaL and sigma_r1 (fixed only when the number of
events under the peak is < 1000);
So here are the results:
S/P MC DEPLETED ALL CUTS
4.47 +- 0.37
4.82 +- 0.17
4.2 +- 0.13
3.55 +- 0.19
3.33 +- 0.25
3.72 +- 0.42
4.37 +- 0.74
3.11 +- 0.51
4.63 +- 1.37
18 +- 68.43
S/P MC ENRICHED ALL CUTS
4.64 +- 0.34
4.05 +- 0.23
3.45 +- 0.25
3.28 +- 0.37
3.2 +- 0.45
3.32 +- 0.62
3.93 +- 1.09
3.33 +- 1.01
5.45 +- 3.89
-2.33 +- -4.32
Double Ratio S/P Mcenr/MCdepl
1.04 +- 0.11
0.84 +- 0.06
0.82 +- 0.06
0.92 +- 0.11
0.96 +- 0.15
0.89 +- 0.2
0.9 +- 0.29
1.07 +- 0.37
1.18 +- 0.91
-0.13 +- -0.55
S/P DATA DEPLETED ALL CUTS
2.23 +- 0.73
7.86 +- 2.73
4.62 +- 0.99
2.85 +- 0.94
3.41 +- 2.22
1.31 +- 0.66
1.33 +- 1.09
1.72 +- 1.62
-2.38 +- -3.8
-0.76 +- -0.81
For data depleted in this case the situation has worsened since the
second bin has again a S/P which is very large wrt near bins.
So the overall correction is
OVERALL CORRECTION
2.32 +- 0.15
6.6 +- 0.05
3.8 +- 0.06
2.63 +- 0.13
3.27 +- 0.21
1.17 +- 0.47
1.2 +- 0.74
1.84 +- 0.54
-2.8 +- -0.64
0.1 +- 44.58
Now we don't have large errors for high mx bins, (except for the last
one), but the correction factor is very high for the second bin.
We have also tried to fit data depleted All cuts without any kinematic
cut. In this case the ratio S/P is ~ 3.
So we have tried to set as starting value for ccb yields and signal
yields this ratio when fitting on kinematic bins, but the fit results
did not change.
Attached there is the spreadsheet with these results.
any suggestions?
Ciao
Antonio & Concezio
|