Resent-date: Fri, 11 Dec 1998 13:36:47 -0600 (CST)
Date: Fri, 11 Dec 1998 13:36:33 -0600
Resent-from: meena@d0gs01
From: Jim Cochran <cochran@jc.hep.IASTATE.EDU>
Subject: alas! He awakens!
Resent-to: meena@fnal.gov
To: meena@fnald0
Cc: cochran@jc.hep.IASTATE.EDU
Resent-message-id: <01J57OXEPQ0G0001MK@FNAL.FNAL.GOV>
Message-id: <199812111936.NAA04813@jc.hep.iastate.edu>


I pinged Tom again last night and actually got a response within a 
few hours - amazing indeed. There's a little on the MR recovery in
here and he is "promising" more on all fronts ...

						    Jim

---------------------------------------------------------------------------
 Jim Cochran					    jimc@iastate.edu
 Department of Physics, High Energy Group	    (515) 294-9018 (office)
 Iowa State University, Ames IA  50011-3160	    (515) 294-3747 (fax)


------- Forwarded Message

Return-Path: rockwell@pa.msu.edu
Delivery-Date: Fri Dec 11 05:47:14 1998
Return-Path: <rockwell@pa.msu.edu>
Received: from alfa-m.pa.msu.edu (alfa-m.pa.msu.edu [35.8.48.89])
	by jc.hep.iastate.edu (8.8.7/8.8.7) with SMTP id FAA04657
	for <cochran@jc.hep.iastate.edu>; Fri, 11 Dec 1998 05:47:14 -0600
From: rockwell@pa.msu.edu
Received: from localhost by alfa-m.pa.msu.edu (5.65v4.0/1.1.10.5/31Jul97-0333PM)
	id AA18182; Fri, 11 Dec 1998 06:47:00 -0500
Message-Id: <9812111147.AA18182@alfa-m.pa.msu.edu>
To: cochran@jc.hep.iastate.edu
Cc: rockwell@pa.msu.edu
Subject: Re: waiting for Mr. R. ... 
In-Reply-To: Your message of "Fri, 11 Dec 98 05:29:56 CST."
             <199812111129.FAA04652@jc.hep.iastate.edu> 
Date: Fri, 11 Dec 98 06:47:00 -0500
X-Mts: smtp

Harry's claim is correct, I did promise to him numbers for last week.  

Arrg, not done yet.  Below is a partial responce to some question you 
sent me to long ago.   Jim, I am actively working on getting this wrapped 
up now.  

I'll send you another note for Monday (really).


- -Tom


>(1) Your note says that you used "Mainring veto". Does this mean MRBS
>    and MICRO-BLANK (aka GOODBEAM) ?

Yes, but see number 3 below.

>(2) What is the trigger efficiency and error for the mu+jets topo
>    signal ?

There isn't a single number; the trigger efficiency is computed
seperately for the various running conditions and trigger definitions.
I can get you an average number.  

The muon trigger efficiency is 67+-5% in the CF.  The jet/met trigger
efficiency is 97+-3% (for events that pass all analysis cuts), but it
wasn't active for the entire dataset.  To get the overall trigger
efficiency, I run the effi*BR calculation without the trigger and
compare the value to the effi*BR number with the trigger (this gives
the average trigger efficiency taking into account the weighting of
CF/EF, fraction of time Jet/Met trigger was active etc.)  The average
trigger efficiency comes out to 89%, and I'm working on the error.  

>(3) You supposedly recover some MR data yet I can find no mention of
>    this (or how it's done!) in any of the documentation (perhaps
>    it's in your thesis ?). For the draft I'm sending out today I
>    have assumed that you did not make any recovery - will be happy
>    to make the change in the draft that goes to the top group (which
>    I expect will happen around the 1st of August). After that I'm
>    reluctant to make such large changes.

Yes, we use some events recovered from the MRBS period.  We search the
recovered events for candidate events, but don't use the recovered 
events in the background or efficiency estimates.  The luminosity of
the recovered events is used to scale the background and expected 
number of events up...

>(4) If I read your document correctly, for your QCD & W+j background
>    you basically follow the same 4 step procedure described in
>    Meena's >note: (i) determine QCD background (ii) use berends
>    scaling to determine alpha on QCD subtracted data. 

The alpha (.18+-.02) is an average value from e+jets and mu+jets.

> From alpha
>    determine # of W+4j events (iii) apply corrections for trigger
>    differences between background method and actual data selection
>    and scale factor for inclusion of MR and EF data.  (iv) Determine
>    efficiency correction factors for E_T^L, A, & H_T and apply to
>    results from (iii).


>    In order to help the reader with this, I am including the following 
table:
>                         |         | Trigger&|        |   |               
>                         |Steps 1-2| MR corr | Step 3 | f |     Step 4 
>- 
- --------------------------------------------------------
- -------------------
>$\mu+$jets |      QCD 1a |    x    |    x    |   x    | x | 0.093 +- 0.047
>           |          1b |    x    |    x    |   x    | x | 0.863 +- 0.433
>           |       Total |         |         |        |   | 0.956 +- 0.436
>           
- --------------------------------------------------------
- --------
>           | $W+$jets 1a |    x    |    x    |   x    | x | 0.312 +- 0.088
>           |          1b |    x    |    x    |   x    | x | 2.889 +- 0.817
>           |       Total |         |         |        |   | 3.201 +- 0.822
>    where f are the combined E_T^L, A, & H_T correction factors. I 
would like 
>    to fill in the table where I have the x's but I don't find the needed
>    info in your note. Can you send it to me ?

This I'm working on.

>(5) On page 5 of your note you mention a relative sys error of 30% on 
the 
>    muon fake isolation rate. Can you tell me how this error was 
determined ?
>    And I assume it is the dominant sys error on your QCD background ?
>    Your response the last time I asked was
>       "The sys error on the fake rate is simply the statistical error 
>        on the measurement of the fake rate in 4 jet events.  The fake 
>        rates for 2, 3, 4 jets are consistent and I have thought that 
>        30% is a safe estimate of the possible sys errors.  I believe 
>        that this relatively large relative error :-) doesn't increase 
>        the final error much."
>    I was, of course, looking for something more specific that gave you 
>    ~30% - I can't put in the paper that "Tom thought ~30% was safe ...".
>    Any suggestions how we can justify this number. I think it has to be
>    included and as such we need to say how we got the uncertainty.

I've done these studies in the past and come to the conclusion that
the method of measuring the fake rate has a systematic error of
~10%(?, anyways significantly less than 30%) when tested in higer
statistic data sets.  The sample I have to measure the fake rate in
for this analysis has low statistics and hence the 30% error.  I
realize you need more than this. 

>(6) What is the luminosity of the subset of your data sample which has
>    the breakdown:
>                      N(jets) .ge. 1     2127
>                      N(jets) .ge. 1      537
>                      N(jets) .ge. 1      124
>                      N(jets) .ge. 1       37(7)
>                      E_T^L .ge. 60        29(7)
>                      A .ge. 0.065         15(4)
>                      H_T .ge. 180          7(3)

>    Your response the last time I asked was
>       "ummm.  These numbers are counts of events. The final 3 numbers 
>        (for the kinematic cuts) are for the whole data set (99.9 pb-1).  
>        The numbers for the looser cuts don't include the whole data set, 
>        and are not scaled to simulate what the whole data set might be."
>    I guess I don't see how having the numbers tabulated is very 
meaningful
>    if each row corresponds to a different luminosity ? I guess at this
>    point I would just ask you to run the full analysis over the 2127 
N+1jet
>    events and give me the breakdown (and luminosity!!!) from that and 
I'll 
>    just add the remainder in parenthese for the final 3 cuts.

Below is a table of the number of events passing the cuts.  The top
portion of the table shows counts of events taken with the Mu_Jet_High
trigger, the bottom portion shows the counts of events taken with
either Mu_Jet_High or the Jet/Met triggers.  (The Jet/Met trigger is
very inefficient for events that have low Met or less than 4 jets -
that is why the mu_jet_high only is shown for the looser selections.)
The numbers of events recorded during the MRBS period are in
parentheses.  The table shows events from the entire data set (all the
luminosity).

 
\begin{table}
\begin{center}
\begin{tabular}{|c||c||c|c|} \hline
        &        \multicolumn{3}{c|}{Number of Events}      \\
        &  Ia   &  \multicolumn{2}{c|}{Ib} \\
     Requirement      &  CF   &    CF   &   EF    \\ \hline
Isol $\mu$&---&12947&4255\\
Clean Up&---&12634&4230\\
Mainring Veto&---&11664&3945\\
\mjx&---&6938&1966\\
Tag Veto&---&6455&1904\\
$\metcal \ge 20 \gev$&---&3282&500\\
$\mettot \ge 20 \gev$&---&2081&291\\
$\etaw \le 2.0$&---&2026&154\\ 
$\nj\ge1$&---&1974&153\\
$\nj\ge2$&---&485&52\\
$\nj\ge3$&---&110&14\\
$\nj\ge4$&4&26(3)&2(0)\\
$\etl \ge 60 \gev$&4&20(3)&2(0)\\
$\aplan \ge 0.065$&2&9(2)&1(0)\\ 
$\htran \ge 200 \gev$&2&4(1)&0(0)\\ \hline
1B All filters&&&\\ 
$\nj\ge4$&---&31(3)&6(0)\\ 
$\etl \ge 60 \gev$&---&25(3)&4(0)\\ 
$\aplan \ge 0.065$&---&13(2)&2(0)\\ 
$\htran \ge 200 \gev$&---&7(1)&0(0)\\ \hline
\end{tabular}
\end{center}
\end{table}

------- End of Forwarded Message