The heavy-flavor contribution is found by fitting the pTrel and signed d0 of soft electrons.  The shapes of the various templates are too similar to each other, so I combined the bottom and charm templates into a single “heavy” template and fit a*(light) + b*(heavy).  The electron-triggered and muon-triggered data give the same result (heavy fraction is 0.22.)  The soft muon fits don’t work with the low-lumi template. (every template happens to have a 0 in a particular pTrel bin where we do have a data point.  This problem will go away when we use the high-lumi templates.)

TCE trigger:

CMUP/CMX trigger:

Here’s the table of predicted and observed values:

I think the next step is to improve the parameterizations of efficiency and fake rate for soft leptons (of both types) in the MC.  neither one is looking great right now.  Below the fold, see pT and eta plots of soft electrons and soft muons: Continue reading new expected/observed soft lepton numbers →

With all that we’ve been talking about the soft electrons outside of jets recently, I decided to see if looking at those (which are almost certainly all conversions) could give us a better handle on the conversions inside jets.  I took the d0 distribution of soft electrons in jets and fit it to a sum of a “conversion” template (found from data by looking at soft electrons outside of jets) and a “non-conversion” template (found by looking at the MC and only taking electron that matched to a non-photon) Here’s the fit (currently it only includes the low-lumi electron-triggered data; I’ll integrate it into the rest of our fitting programs and make it include everything soon.

I expect that this result will greatly improve the heavy flavor fits, which still have the excess at high d0, presumably due to conversions:

-Scott

Previously, there were a large number of soft electrons that had no jet associated with them, partially because the jet they were in was removed from the jet block because it overlapped with the trigger electron.  Here are plots showing the Delta R between the trigger and the soft electron for those that do have a jet associated with them and those that do not:

Those that do have a jet:

Those soft electrons that do not have a jet:

Obviously, the problem is not completely due to being too close to the trigger electron.

Bonus Section! Continue reading Removing electrons that have no jet associated with them →

After our talk on Friday, I put a ttbar selection into the code.  It seems to be working well: the ttbar MC accounts for a large portion of the observed events, as seen below.  The selection is:

• Trigger lepton
• Met > 30
• hT > 250
• nJet >= 4
• dPhi(Met, Lep) > 0.5

From the plots below, it looks like the hT cut isn’t doing anything.  Also, I’m going to have to make a new selection with the Met cut reversed in order to fit the QCD.

Bonus section! Continue reading ttbar Check →

We’ve finally run the analysis all the way through, although nearly every element could use some improving.  Here are the results from the first femtobarn of data:

TCE trigger:

CMUP/CMX trigger:

We should talk about these numbers at some point, of course.  Until then, I’m going to see what’s wrong with the soft electron efficiency and fake rate parameterizations.  They agreed well with my conversion and generic track test samples, but there’s something wrong with at least the pT part when applied to our W sample:

This problem went away in the past, but I think it came back when I recently updated the matching to find conversions. I’m going to check that now.

-Scott

I recently corrected our soft electron tagging in the Monte Carlo. Previously, we had been using the MC likelihood to tag them.  That method does successfully separate real and fake electrons, but with a different efficiency and fake rate than in the data.  I changed to to (correctly) match a track with the genp-level information, check whether the electron is real or fake, and then apply a probability of tagging it.  This broke the pTrel and d0 fits: the heavy flavor templates didn’t have the asymmetry in d0 that they used to.  I found out that this was due to the matching between tracks and GenParticles: I had a tight cut on the distance (in Eta-Phi space) that a track could match a particle, and tracks with high d0 would often fail this cut. (since track parameters are measured at the (extrapolated) point of closest approach to the beamline, while the GenP information is the initial momentum at the point of creation.)

Therefore, I loosened this cut to the point where I don’t see any more tracks failing to find a matching particle.  The fits are better now (see below) but the d0 tails are still a little too symmetric.  I suspect that this is due to high-d0 tracks matching the wrong particles, and therefore being called fakes too often.  Unfortunately, I’m pretty sure it’s impossible to match tracks to particles any better than I currently am without a little bit more information:  If the initial position of the GenParticles was saved, (it’s in the Stntuple, but not the DKntuple) I could calculate the expected phi0 of the tracks better than I currently can.

Notice that this issue with the matching would affect the muons in addition to the electrons if I’m correct.

First, a comparison of the old and new electron from b templates, showing that the new ones are too symmetric:

Now, the results of the fits:

Low-luminosity, TCE trigger, soft electrons:

More after the jump… Continue reading New pTrel fits →

As was done for the soft muons earlier, I compare the shapes of the pTrel and d0 distributions for soft electrons in the TCE-triggered sample in MC and mistagged data.

Here is pTrel:

And here is d0:

The shape of the mistag distribution is clearly different than the W+light MC in both variables.  The mistag response needs to be changed.

Next I try to do the simultaneous fit to the pTrel and d0 distributions for the soft electron samples in both the TCE and CMUP/CMX samples.  Here is the TCE fit.

And here is the combined CMUP/CMX fit

We see an excess at large |d0| in both samples.

I’ve finished parameterizing the efficiency as well.  I’m about the get all the code into out analysis framework, and then we’ll run over the data tonight using everything.

These plots use the same conversion sample on which the likelihood was trained, so it’s not terribly surprising that they agree pretty well. (As a reminder, I’m looking at well-identified conversions in the 8 GeV electron trigger, looking at the softer leg to avoid a trigger bias, and subtracting the distributions of events from the sideband of the \Delta \cot(\theta) distribution.)

As in my previous post (about parameterizing the fake rate) the predicted efficiency is in red and the observed efficiency is in blue.  It looks like my error bars are too big for the predicted efficiency, but I can fix that later.

-Scott

Summary:

• The likelihood has been retrained, using generic tracks in the muon sample instead of in the electron sample. (I believe there were still some real electrons in my “fake” training sample from the electron trigger)
• The fake rate has been parameterized as a function of Pt, Eta, CesZ, and TrkIso.
• This parameterized fake rate has been tested on three samples: the muon-20 trigger (not a great test, since it was trained on this exact sample), the jet-20 sample (very good agreement) and the jet-50 sample. (Not as good agreement at higher Pt or TrkIso.)

Still needs to be done:

• I need to parameterize the efficiency (shouldn’t be too hard, now that I’ve gotten the fake rate)
• I need to move some code around to get it into the ntupling/analysis programs, but that’s literally just some copying and pasting
• We need to decide on some cuts: Should we move up slightly in Pt to decrease our fake rate substantially? Should we cut out the edges (either the high-Eta or the high-CesZ region)?

What you’ve all been waiting for: Plots!

Predicted fake rate is in red, observed fake rate is in blue.

Muon sample:

Jet20 sample:

Jet50 sample:

Finally, here are the plots of the efficiency.  These plots are made using the conversion sample, which is the same sample I trained the likelihood on.  Is there a different way to check the parameterization I will get from these plots?

-Scott

We think the excess in at high pTrel for soft muons in the high-pT muon triggered sample is now understood.  The effect comes from two sources:

1. A bug in the code that allowed muon near 20 GeV to be counted twice.
2. Our background templates did not take the Drell Yan contribution into account.

Below we have plots of the pTrel in the data, from the mistag prediction, and from the MC templates including DY.

CMUP

CMX

We see that the Drell Yan contribution has a long tail at high pTrel.  There is an easy way to remove this contribution, as can be seen in the stacked plot below.

First we notice that the MC underpredicts the total number of soft muon events.  We can also see that we should remove the mass region near the Z mass.  The DY events that go below the Z mass likely have a radaited photon.  We can remove those events by making selection on the EM fraction.  Finally, notice that the number of events in data increases below m=5 GeV with no corresponding increase in the MC prediction.  We believe these events may be due to b-bbar production.  We will likely have to cut this region away as well.

Even though we plan to remove these events, I went ahead and fit d0 & pTrel in the muons with the DY template added to see if any remaining excess at high pTrel was covered by the DY template.  The fit below is only for muon-triggered events, since that is where we first saw the excess.