Gen5 Validation TopNtuples

NEW! Validation Results

Plots
- My talk at L+J describing data comparison..
- My talk at L+J describing data comparison..
- Similar plots showing MC comparison.
- "Kohei acceptance table" comparisons for data. This data is a direct comparison. The same events were run over to make both plots.
- "Kohei acceptance table" comparisons for Monte Carlo. This Monte Carlo is NOT a direct comparison. There are only half as many events in the 5.3.1 samples.
- Top MC individual channels. Sorry this doesn't have pretty colors, but there is too much information to organize it side by side.
- Comparison of the top efficiency times luminosity. Note that I did not calculate the correlated errors for the final computation, but just quote the central value.
- JET50 Data/MC Comparisons. This is very crude. I have compared all the track quantities for data and Monte Carlo that reside in TopNtuple. There is much better agreement in Gen5 than in Gen4. This is due to the parametric charge deposition model for the silicon simulation. Note that I used a few runs near the simulated run for this study.
Discussion of Results
- Data
  - Some migration of events in the zero jet bin in the electron sample before tight lepton classification. This is due to the fact that now, we include the conversion veto at the TopEventModule level.
  - Missing Et is slightly different for Met < 20 GeV.
  - d0 looks much better when we use silicon tracks when available.
  - Muon hadronic energy shifted. The calibrations had changed, and this was expected.
  - E/P looks slightly tigher now.
  - CMU/CMP deltaX look more symmetric. CMX deltaX is less symmetric but more central. There was a fix in 5.3.1pre4 that made the CMX deltaX look symmetric, so this is not an issue.
- Monte Carlo
  - Data and Monte Carlo look very similar at a low level (comparing hits on tracks, missed hits, etc). This causes the efficiency to "decrease" in the Monte Carlo, but it was too high before, and now looks more like the data.
  - CMX top acceptance increased by 50%!!! This is due to the fact that the miniskirts have been simulated in the new Monte Carlo, whereas in the old Monte Carlo, they were not.
  - There are no dileptons in the top Monte Carlo! Email has been sent to the experts to be sure this does not happen in the future.

Issues So Far

Changes we EXPECT

BTagging
- COT hit requirements changed (expect higher tracking efficiency), and hence higher tag rates.
- For Monte Carlo, using parametric charge deposition model, so expect better agreement between data and MC (which means tag rates in MC should be lower than before).
- Cannot access readout errors in the "new" SiExpected format (with SiChipOn tables), so expect higher number of missed hits (hence a lower tag rate from this).
- In processing the 8-GeV electron sample, I removed tight electron towers by accident. This should be a small effect, since the pt is much harder for tight electrons, and the isolation cut means that electrons inside a jet will not pass the quality cuts.
TopEventModule
- Conversion filter removed at "tight lepton" stage: cuts out 50% of the tracks immediately. Most likely highly correlated with missing Et cut (i.e. both are most likely selecting W's).
- Need to make sure electron removal from jet lists and muon removal from missing Et is working.
- Leptons are not using beam-constrained COT-only tracks. Now we use silicon tracks if there are any, if not, we fall back on the beam-constrained COT-only tracks.

Changes we DIDN'T EXPECT

BTagging
- None so far
TopEventModule
- Forward muons have standalone tracks attached to them, causing a crash in TopEventModule (FIXED!).
- Be careful when dealing with the "bad alignment" period (152595-154012). This data was not included in the reprocessing. (So when using scripts to compare Gen4 and Gen5, be sure not to skip any runs in the reprocessed data.).

Not Yet Implemented

BTagging
- None so far
TopEventModule
- Removing electrons from the jet list as the following is not yet implemented:
```
      
      leptonTypesMetCorr set TCMUP
      leptonTypesMetCorr add TCMX

      leptonTypesJetReclu set TCE
            
```
  You should still use the old tcl parameter "removeTightElectronTowers".

Prescription

TopNtuples for the 5.3.1pre2 validation effort are being made. The prescription used to make them is:

    # Create test release
    setup cdfsoft2 5.3.1pre5
    newrel -t 5.3.1pre5 topnt_5_3_1pre5
    cd topnt_5_3_1pre5

    addpkg TopMods
    gmake nobin TopMods.bin

Ntuple files

We are processing all the data listed on Young-Kee's web page.

Description of files:
1. BTag : require 1 loose jet only, for b-tagging studies.
2. Inclusive : No filter requirement. Pass through initial dataset with new objects.
3. Lepton : Tight lepton cuts, minus isolation requirement.
4. W : Tight lepton, plus missing Et.
5. Z-loose : Tight lepton + loose lepton.
6. Z-tight : Tight lepton + tight lepton.
Files are named as:
```
topnt_(selection)_gen5val_fixed_(sample)_(sectionnumber).root
	
```
- selection : one of the 6 categories above
- sample : inclmuo, inclele, topmc_qq, blpc0b, jpmm0b, gjt20b, inclmuo_all, jet50mc
- sectionnumber : CAF section number
Please note that "missing" files in the "control samples" for btagging are due to the fact that they do not have silicon in any run in that file. This includes gjt20b and blpc0b.

Location of Data : All are on fcdfdata034.fnal.gov :

Note : if you are logged in as "cdfopr", you don't need the /cdf/scratch in the following path names. Otherwise, if you are "anonymous", please access the path as below.

Tight Muons :

     root://fcdfdata047.fnal.gov//cdf/scratch/top/scratch/results/topnt/pre2/topnt/inclmuo/topnt_BTag_gen5val_fixed_inclmuo_*.root
     root://fcdfdata047.fnal.gov//cdf/scratch/top/scratch/results/topnt/pre2/topnt/inclmuo/topnt_Inclusive_gen5val_fixed_inclmuo_*.root
     root://fcdfdata047.fnal.gov//cdf/scratch/top/scratch/results/topnt/pre2/topnt/inclmuo/topnt_Lepton_gen5val_fixed_inclmuo_*.root
     root://fcdfdata047.fnal.gov//cdf/scratch/top/scratch/results/topnt/pre2/topnt/inclmuo/topnt_W_gen5val_fixed_inclmuo_*.root
     root://fcdfdata047.fnal.gov//cdf/scratch/top/scratch/results/topnt/pre2/topnt/inclmuo/topnt_Z-loose_gen5val_fixed_inclmuo_*.root
     root://fcdfdata047.fnal.gov//cdf/scratch/top/scratch/results/topnt/pre2/topnt/inclmuo/topnt_Z-tight_gen5val_fixed_inclmuo_*.root