Minutes for HF PWG meeting on 04/03/2014

 0) General

* Zhenyu has replaced Wei as a HF PWGC.
* Reminder of QM2014 deadlines

1) low pT NPE in pp 2009 - Olga

Olga updated the analysis to try to understand the enhancement at low pT w.r.t. Phenix and FONLL calculations.
She had previouly used a |eta|<1 cut and used a TOF matching efficiency as a function of pT integrated over
eta and phi from Babara. This time she derives the TOF matching eff by herself for different pT bins and phi
regions as a function of eta. She is using a |eta|<0.7 cut. Another change was to extract and use Gamma 2009
and Dalitz 2008 embedding samples to extract the photonic electron efficiencies. The third update was a
recaculation of the purity. With all these changes, the enhancement at low pT is now gone. A comment was made
on slide 4 where the eta-dependent tof matching efficiency shows rapid falling offs for -123<phi<-63. Suggested
to cut away these fall-off eta regions for -123<phi<-63

Action items:
1. Dalitz embedding 2009 samples for publication. Requests were in place but were not approved. The convenors
will follow this up. For conferences, using Dalitz 2008 samples are fine, as the photonic electron efficiencies
for gamma samples are similar between 2008 and 2009. For publication, Dalitz embedding 2009 samples are needed.
2. Olga will send around the purity extraction QA plots
3. Olga will put a cut to remove eta<0.3 and -123<phi<-63

2) high pT NPE in pp 2012 - Xiaozhi

Xiaozhi presented electron/gamma embedding QA plots. For data, he used photonic electrons with a requirement on
pair DCA and inv mass to select pure electrons, while for embedding he used the MC particle info to select pure
electrons. He found that nhitfit and global DCA distributions are different between data and embedding. Such a
difference should not have a large influence on the results, as the cuts he applied are far away from the majority
of the electron candidate distributions.

Action items:
1. Xiaozhi updates the study with the same selection cuts applied to both data and embedding to ensure there is
no bias introduced.