Name  Comments 

George  This physics results of this analysis are connected to the results from 14003 K0,Lambda. Are the analyses coordinated , e.g.  pt bins, multiplicity bins, methods, etc ? 
Monika  The codes are different and the pt bins are optimized for this analysis (so, not the same), but multiplicity bins , event selection , track selection are same. 
George  The codes are different and the pt bins are optimized for this analysis (so, not the same), but multiplicity bins , event selection , track selection are same. 
Monika  For now the differences between the values in the points are take from the largest multiplicity bin and the % difference is assigned as syst for all bins. However, we see that this is still affected by the statistical uncertainty in each data point, so the plane is to combine all multiplicity bins to have the larges statistics and then vary cuts to see how the result changes and assign the syst uncertainty. What are stat error : s 27, 28,  what are the stat errors ? shown on s 21  significance is small , efficiency is small 
***All agreed that the uncertainty fluctuations binbybin on s 29 are not physical and the above procedure should be applied. 

Steve  what is the average eta for the measurement and how does this compare to other analyses? 
Monika  The range is different from Ks and Lambda, because the ECAL acceptance in different. 
Julia  It is worthwhile to compare the results for all particles with the same eta bins, since v2 may depend on eta and we don't know how, but the pi0 analysis can not extend the eta range. We would need the K0 and Lambda analysis to make the same cuts as pi0; then we can compare the same eta. 
Matt  Related to this question: the efficiency is not constant for pi0 as a function of eta, so we need to see if and how the result changes after the efficiency corrections are applied 
Monika  This is on our todo list 
Gunther  In the highest multiplicity classes v2 of pi0 is above that for charged hadrons at low pt. Is that what we expect base on the yields measured in 12016 ? 
Monika  similar trend is seen in the RHIC results in s 4 
Julia  It is worthwhile checking the consistency of the charged hadron and pi0 , Kshort results for v2 with the assumption that charged hadrons comprise just pi,k,p . *** consistency check using 13002 , 12016, this analysis and Ks, Lambda ( will have to assume the v2 of protons = v2 of lambda) the phi analysis  useful to look at closure test on MC . It would be helpful to check if all errors and uncertainties propagate the way you think 
Monika  * will add a HIDJET study in the AnNote  still to do 
Christof  s 11: do you understand the background shape, in particular the undershoot at low mass 
Monika  Yes, we have don a great deal of study for the PbPb paper 11009. The undershoot is due to photon conversions. This a correlation that is not present in the mixed event BG. This undershoot does not affect the result. 
Federico  Why don't you just perform a fit using a parabolic shape of the BG ? 
Monika  Why don't you just perform a fit using a parabolic shape of the BG ? 
Frederico  Why don't you just perform a fit using a parabolic shape of the BG ? 
Monika  same event selections as published analysis . Will check with the experts 
Frederico  ECAL spikes  in the end you decided not the cut them . So what's the contamination from the spikes in your sample ? Isn't this affecting your v2 estimate? 
George  Monika: same event selections as published analysis . Will check with the experts 
Monika  does not have an answer now. will check 
George  about varying the sidebands and including this in syst, 
Monika  yes this was done a documented. 
Matt, Christof and all agree that the analysis is in good shape and the
ARC is already onboard. Analysis preapproved. Remaining questions will
be followed up for approval.
 RaviJanjam  16 Apr 2014  MonikaSharma  22 Apr 2014