News/update

1
News/update
P.Sphicas Physics Coordination Meeting August 29,
2007

• Agenda
• News/update
• Rooms for 2008
• Dates for 2008
• Upcoming events
• September CMS week
• October Physics week
• Things to organize for
• Discussion on code approval

2
News/Update
3
Rooms and dates in 2008

• CMS secretariat wants to block out rooms for all
  CMS meetings for 2008 soon.
• Are there any group meetings that need to move?
• Note that changes will be very difficult to
  implement
• Deadline for response Monday Sep 3rd
• After that I will submit this years list
• Dates in 2008
• Do we continue the physics days?
• Suggestions until end May 2008
• Four weeks after CERN opening frequency once
  every 5 weeks. Means 2 additional physics days.
• Physics Week 1 April 7 or April 14

4
(bi)weekly meetings
5
2007 papers clarification

• 2007 paper exercise, and the grand summary note
• Each group has N 2007 analyses
• These will be documented in Analysis Notes. To
  be approved, they have to get Web PAS pages. They
  may also become CMS Notes.
• One of these analysis will become a publication
  to go through the full process
• In parallel with all this, we will document the
  physics reach with 10, 100 and 1000 pb-1 (like
  another projection) across groups
• This is the grand summary note.
• Goal for this a living document. First version
  shortly after Physics week. Total length about
  30-50 pages
• We should converge on the papers by the Sept
  CMS week
• Item for next phys coord meeting your synthesis/
  analysis of which paper (given list of 2007
  analyses for your group)
• We then discuss the overall set of papers, their
  physics motivation and decide during CMS week

6
Annual Review CSA07 news

• CMS MB on Monday
• No surprises
• CSA07 news
• Will start on Sep 24 (delay). So by now
  definitely not finished by Physics week.

7
Upcoming events
8
CMS week (parallel sessions)

• Tuesday
• 0830-1030 QCD-Top-EWK-Bphysics
• 1100-1300 QCD-Top-EWK-Bphysics
• 1400-1600 egamma muons
• 1630-1830 JetMET btag
• Wednesday
• 0830-1030 Higgs I SUSY-BSM I
• 1100-1300 Higgs II SUSY-BSM II
• Friday
• 0830-1030 generators particle flow
  diffraction
• Late afternoon Heavy Ions (as usually)
• Thu morning
• joint meeting with Trigger (after SW/comp
  plenary)
• Monday afternoon 1800-1900
• Physics Coordination meeting

9
CMS week (plenary session)

• Thu afternoon 1400-1600
• JetMET 205
• B-tag 205
• P-flow tau-ID 205
• E/gamma 205
• Muons 205
• Thu afternoon 1630-1830
• Top 205
• EWK 205
• QCD 205
• Theory talk (Treleani) 205
• Heavy Ions 205
• Fri morning 1100-1230 (plus parallel session
  on Fri morning)
• B phys 123
• Diffraction 123
• Higgs 255
• SUSY-BSM 255

10
Physics Week (I)

• Grand rendez-vous for all POGs and PAGs
• This is when we show the results of the 2007
  analyses
• Open questions when do we do the pre-approvals
  in the groups?
• Sep 3 Next week
• Sep 10
• Sep 17 CMS week
• Sep 24
• Oct 1
• Oct 8
• Oct 15
• Oct 22 Physics week
• Proposal Analysis Notes ready on Oct 8.
  Pre-approvals in group meetings during the
  following 2 weeks.
• CMS Notes to go through after approvals are
  completed

11
Physics Week (II)

• During Physics Week
• One day in association with trigger
• But do we know enough?
• One session (half day) on CSA07 status/experience
• One plenary presentation per 2007 analysis? At
  25 analyses, and 20 min/analysis ? 500 mins
  (1.5 days).
• Other suggestions?
• Other festivities scheduled for Physics week
• Grand review of 2007 analyses
• First session on our 10 and 100 pb-1 plan?
• First session on demonstrating new physics?
• Appointment of ARCs
• Announcement of next conveners (2-yr terms)
• Will send out call for nominations week before
  CMS week

12
Things to organize for
13
CSA07

• Each group must start thinking about doing
  analysis in real-time
• Means a small team of people who follow the data
  from the Tier-0 to the Tier-2
• Make basic plots
• Establish basic health of the data
• A starting point is the reco-validation person in
  Ilarias taskforce.
• Goal identify problems within 24 hours of Tier-0
• Dont know if possible
• Note we do not have to do real-time analysis
• Remaining issues
• Primary datasets (finalization) special
  JetCalib stream
• Jet bins in the data or not. Gunter/Nikos not
  present today, so delay discussion to tomorrows
  afternoon session
• Second reco pass (at T1s) do we do it with
  100pb-1 constants?

14
2007 analyses and paper exercise
15
Analysis Approval system
16
Code approval (I)

• How we "approve" the objects used by an analysis
  (both now and in asymptotia, i.e. in steady-state
  with everything understood). Proposal sent out
• 1. All the basic physics objects that are used by
  a physics analyses should be approved by the
  corresponding physics object group. This includes
  object reconstruction, identification and
  selection.
• 2. At any point in time, each physics object
  group should maintain the "official" code for its
  physics objects in a CMSSW release which is
  actively maintained and documented. Physics
  analyses which utilize this standard code are
  de-facto "approved" as far as these elements are
  concerned. Naturally, questions clarifying
  details of the code used may well be asked during
  the review process.

17
Code approval (II)

• 3. If a particular analysis uses reconstruction
  algorithms, features or alternate selection
  requirements from the ones in the released and
  approved by the physics object groups code, then
  the analysis proponents must include a detailed
  and complete description of these changes in
  their documentation suite. This part of the
  documentation must be sent to the conveners of
  the corresponding Physics Object group for
  explicit approval. The conveners may well decide
  that the procedure documented deserves more
  attention at which point they may invite the
  analysis proponents to make a specialized
  presentation at a future meeting of the physics
  object group. This phase must be concluded with
  an explicit "acknowledgment" of the POG conveners
  before the pre-approval.
• 4. In the current phase, where we have yet to
  release the first "official" sets of physics
  objects, all analyses should follow the procedure
  in (3) above.

18
Feedback

• Overall agreement that this is necessary in order
  to have an orderly system (aka avoid chaos)
• ALL (but one) comments received had to do with
  going one more step in the code approval, e.g.
• proponents must eventually, when the process
  is over and the analysis official, provide the
  new reconstruction code for general use in the
  Collaboration. One could even ask to have it
  available during the review process.
• the code should be available to POG  to review
  is that implicit in what you call documentation
  suite ? It might be worth spelling this out.
• if approved by the POG, the specific code
  developed in the PAG must not only be made
  available to the POG, but also included in the
  POG software framework by the PAG developer (with
  the help and guidance from the corresponding POG
  expert).

19
Approval of code itself?

• Strategic question is the code itself part of
  the approval process? Or is it only subject to
  addition to the official suite at the end?
• Probably obvious that at the end of the process
  the code should go back into the POG official
  codes/documentation
• Real issue should the code itself be reviewed
  during analysis approval?
• Some obvious ramifications (length of process
  number and type of people who participate in
  reviews)
• One big new question if the code itself is
  reviewed, then what about the rest of the code,
  the non-POG code e.g. fitting?
• This can be especially tricky in some complicated
  analyses.
• Should this code also be placed in public or
  even in a CVS repository for inspection/for
  approval?
• Who inspects/approves such code? The PAG?

20
Code approval

• Last comment
• calculation of efficiencies, for both signal and
  background? I suppose it is case-by-case,
  depending on how widespread the use of a
  particular ID/selection is.
• Given dependence on signal details, yes,
  efficiencies (and background rejections) are part
  of each analysis
• Of course, there will be some central values
  corresponding to the default cuts/settings of an
  algorithm (e.g. electron fake rate per jet for
  some default settings)

21
AOB

• Next Physics Coordination Meeting
• Wed Aug 29 1730-1900
• NOTE Meeting on Aug 15 is cancelled

22
Planning for late 07-beg 08
23
2007-08 priorities (physics org)

• Axes of work, priorities in Nov-07/May-08
  time-frame
• Definition of streams definition of Express-line
  (!!!)
• Definition of online selection OnlSel monitoring
  (!!!)
• Continuing, in collaboration with trigger
• Definition of analysis path from T0 to T2, the
  how of the analyses (!!)
• CSA07 first test. By May 08 we MUST have each
  group firmly established at a few analysis
  centers
• Definition of analyses for 10, 100, 1000 pb-1
  (!!)
• Completion of 2007 analyses goal at the start
  of data-taking analyses ready to absorb 100pb-1
  (i.e. ALL actions are pre-planned)
• Definition of discoveries in SUSY, Higgs, other
  BSM (!!!!!)
• Discoveries may involve looking at multiple
  channels, combinations of small and large
  excesses, different signatures

24
Preparing for this work (I)

• A and B must be discussed with Trigger
  coordination
• They are just starting (with their organization)
  stay tuned
• Short-term action
• Identify (through CSA07) 10 subsample using
  your triggers
• Definition of analysis path
• Here CSA07 is a great exercise. We must test as
  much as possible of the logistics involved
• Short-term action
• For every sample at least one hosting Tier-2
• For every sample a responsible team of people
• Establish monitoring of data flow from T0-gtT1-gtT2.

25
Preparing for this work (II)

• Early physics analyses
• We should concentrate on the 100pb-1 and 10pb-1.
  1fb-1 is not dropped it is included (and is a
  very important part because of the discovery
  potential)
• But new proposal 100pb-1 analysis, 10pb-1
  calib/align
• Short-term action
• Plan for/prepare first status reports on 2007
  analyses at next Physics Days (last week in
  August)
• Definition of discovery still consulting with
  people
• Leaning towards an informal advisory group which
  will include some theorist input(s)

26
AOB

• Next Physics Coordination Meeting
• Standard Slot
• Wed, Sep 12 at 1500-1700