Update:
Final version is posted as https://drupal.star.bnl.gov/STAR/node/30504

****************************************************
Update evening 14 May - v6
****************************************************
Update 14 May - v5
summary of suggested plots and systematics in Report11c
****************************************************
Update 11 May - v3
**********************
Update 9 May: - v2
****************************************************
Update 1 May: summary of suggested plots and systematics in Report11.

****************************************************

Suggestion for a poster.
As of 20 April the results are not finalized (not even preliminarized), and we don't know the allowed format -> only a sketch

Comments....
Zhenyu:

Please find a first set of comments on the 1st version of your
poster. In general, I think a few places need some corrections
or clarification. Have you presented the systematic uncertainty
estimation in a PWG meeting? As I said, I would like to go
through the analysis step-by-step for every new result intended
for QM. Please prepare slides with such info and either send
them to the group or present at the PWG meeting.

Dear Zhenyu,
Now I have shown all the systematics on PWG meetings or made the details available on the list.  Please tell me if there’s a place where I did not give sufficient details and I will do that.

**************************
Please run a spelling checker as there are quite a few typos.

--- Thanks for noting this.

Motivation:
1.    remove "Heavy flavor ... stages"

Done

2. add a figure like the first one at
http://dnp-old.nscl.msu.edu/current/quarkonia-2009-08.html

--- Inserted it. (I considered it earlier, my only worry was that this is a work of A. Mocsy and I was not sure it can be freely used).

3. combine "In fact, ..." and "However, ..."
4. remove "In fact,"
5. change "expected" -> "suggested"
6 add "feed down," in front of "and co-mover absorption" in
both places
7. "complicates" -> "complicate"
8. "is negligible" -> "are negligible" with a space in front of "are"
9. "Upsilon state" -> "Upsilon states"?

Done.

9. can you please elaborate of me what you mean by
"Dissociation of Upsilon state be most prominent" and what it
implies?

--- The higher the energy density, the stronger suppression is expected. I rephrased it to avoid confusion.

Measurement:
1. "dielecron" -> "dielectron"
done
2. "energetic hits" -> "energetic clusters"
3. "Electronness" id cuts in the TPC and BEMC
   -> Electron Identification with the TPC and BEMC
4. m_ee -> m_e+e-
5. "most of it is" -> "most of them are"
6. remove "open" and "pairs" and add "decay" in front of "(BB)"

Peak Extraction:

1. can you please remind me why the background was fit only
up to pT of 8 GeV?

--- That’s only the line disappearing because of low quality conversion to PNG. I have replaced the figure.
 
2. can you please remind me whether those are from Chi2 fit
or likelihood fit?

It was the likelihood method I used.

3. what is the uncertainty on the number 19, and 44? is 12
the statistical uncertainty only, or total uncertainty for 67?

The total uncertainty of the fit. Also the error estimation method changed and now there’s a different number.  I have updated the plot .

is 44 the background under the peak or integrated over the
entire pT region shown in the plot?

The background is calculated in the “peak region” of 8 to 11.

4. add "STAR Preliminary" in all plots

--- Done

5. what is the number 0.64?
6. "thee raw" -> "the raw"
7. "Monte Carlo and identified electron samples are used"
-> add "data" in front of samples
-> make it clear what they are used for
8. missing uncertainties for cross section

--- Thank you for noting these. I tried to tersely clarify each point.

Nuclear modification factor:

1. what do you need reference [3] for?

--- That is the citation for the Au+Au points and the p+p reference. I know the paper changed substantiallty, now I include “submitted to PLB and took away the arXiv number.

2.    where is the Nbin error for Au+Au?

--- That is in the Raa, where I don’t list the uncertainties.

3. I don't understand why the Nbin error for 0-60% is so
much larger than 0-10 and 10-30%. Where do you get
these errors from?

--- Good point, that error was nonsense. I took the values from arXiv:0907.0202. When unifying the bins I summed up quadratically the errors, which is a vast overestimation. Now I use Hiroshi’s 2013 calculations, while calculate the error from these 2012 values still, the following way: I ran Glauber simulations to recalculate the errors the following way: by modifying the pp cross section I shifted the Ncoll values downwards so that I reproduced the values of the reference. Then with the same setting I calculated it for the unified bins. The resulting errors are now much closer to averages than quadratic sums.

4. move a bit to the right the two boxes "Yield in current
measurement" and "Acceptance x Efficiency"

--- I restructured the poster, now this section comes before the peak extraction.

Conclusion:
1. you did not mention "models" in any other place. Which
model are you referring to?

--- I incluced the Strickman and Rapp’s models to the plot (same as in the dAu+AuAu RAA paper). I briefly describe these models. I draw a soft discussion now which I am happy to discuss.

 
2. I don't understand what you mean by "in central"

--- That was a result of a last minute edit, fixed, sorry.

**************************

Yifei

     Measurement:
     1) "STAR at RHIC" => either "STAR" or "Solenoidal Tracker at RHIC
(STAR)"

--- Done

     2) I don 't think a mass window up to 20 GeV is useful here. a)
you never show a mass window up to 20 GeV; b) there is no statistics
beyond 12 GeV (or off the Upsilon mass region).

--- I would have liked to mention the cut, because the number of candidates depend on this. I got rid of the upper boundary, it really doesn’t stand for much.

Since people is easily
confused by "candidates", they are not "Upsilon candidates" at all.
Suggest change to "e+e- combinations".

--- Good idea, fixed.

     Upsilon yield:
     3) A question to the combinatorial background subtraction, did you
subtract the LS background or a "double-exponential fit" ? If you did
the latter, how did you consider the background uncertainty?

--- The latter. The background fit error is generally much smaller than the others, but it is indeed a difficulty. The estimation of the error has recently changed: As per Daniel’s suggestion, I simulate the fluctuations around the fitted curves and the 1sigma error of the resulting yield distribution is considered.

     4) The pT spectrum is questionable: The open points represented
for the overall integrated yield from pT 0 - 10 GeV/c and placed in the
center of 5 GeV/c, if you did not do any scale of the yield it must be
wrong; if you did, then how did you do that scale? how do you know the
pT shape according to the two data points below pT=4 GeV/c? Or how did
you know your average pT from 0 to 10 GeV/c according to your current
measurement?

--- I have recently presented the bin width corrected spectrum on the PWG meeting. I use the same function that is used for weighting the embedding, ~pT/(exp(pT/T-1), and the corrected result is consistent with it.

         Then another question could come up: How did you obtain the
cross section if you do not know the pT shape? This is also related to
Zhenyu 's comment "the uncertainty of the cross section is missing."

--- The overall cross section I took is
B_{ee}dsigma/dy = N_corrected * 1/intLumi * 1/Delta_y

     5) Do you have time to finish your estimate for systematic
uncertainties? If we present the result in a public conference we do not
expect "guesses".

--- A summary of my recent efforts is here:
https://drupal.star.bnl.gov/STAR/blog/rvertesi/upsilons-uu-systematics-spectrum-raa
I don’t think they are good for final errors, but I am confident they are more then guesses now.