Cyril
Walk-through presentation
Link to Indico talk [1]
Comments Michele
Make a proposal for mSUGRA parameters to use. In particular tan(beta), A_0, sign-mu. Is it realistic to have a few choices ? What does a choice mean with respect to the physics reach ? What are the proposed choices in the CSC notes ? What is CMS using ? What are the choices at the Tevatron ?
What the current limits from the Tevatron ? Try to see if you find something here:
http://www-cdf.fnal.gov/physics/exotic/exotic.html
http://www-d0.fnal.gov/d0_publications/d0_pubs_list_runII_bytopic.html#np
(http://arxiv.org/abs/0712.3805 ??)
Decide what final state you will be looking at: 1 lepton ? no leptons ? Number of jets ? You can base this on the expected discovery reaches. Show those plots.
If you choose to require a lepton: muon or electron ? Are these channels already covered by another PhD in the SUSY group ? Define a grid in the m0-m1/2 plane. Number of points, values. Use the grids defined by Freiburg as a start (any need for changes ?). More points at smaller values ?
Can you interpolate between different grid points ?
Will you continue to consider Multivariate analyses ?
What are the signal samples you need ? (based on the m0-m1/2-grid) How many events ? Generators ?
What samples are already available in the SUSY group ? What are the SUSY group plans (ask Sasha).
How many CPU hours and disk space are needed for one grid point ?
Make a list of background samples you need. What sample, generator, cross section, number of events and disk space for each. Are they in Bern, on the grid ?
Show a list of a first round object definitions you will use: How exactly is you lepton defined ? How a jet ? How exactly is Met reconstructed ?
Show a list of first round event selections. For at least one choice of a grid point, prepare a table with a cut flow (number of events after each event selection cut). Do the same for every background sample.
Describe how you will be calculating a limit based on the expected signal events, expected background events and measured events in data.
What is the name of your analysis package in SVN ?
Are you ready with using Rel 15 ? What release exactly ?
How many CPU-hours do you need per analysis go. Without systematics to start. And without the generation of samples.
Prepare the following plot A (I hope this works... SIGI: HERE and the following ones). For one point in the grid calculate the expected exclusion confidence level (events in data is the number of expected BG, zero signal) as a function of integrated luminosity. No systematics to start. You can use 10 TeV to start, or also 14 TeV.
Prepare the following plot B: For the same point as above, now calculate how big the signal cross section should have been, to exclude it at 68% and 90% C.L. Plot the obtained cross section values for 68% and 90% as a funtion of integrated luminosity. No syst. Overlay a horizontal line with the expected cross section.
For B, you wil need to explain how you derive the acceptance, selection efficeincy and triger efficiency. What is this efficiency exactly: numerator and denominator.
Plot C: Take a value of integrated luminosity between 10 pb-1 and 100 pb-1. Take the vaues from the plot above. Get the same values for all grind points in m-1/2, keeping the same m_0 value. PLot the expected, 68% and 90% cross section limits as a function of m-1/2.
Show a plot of the Missing Et (as you reconstruct it) for a minimum bias sample. Overlay the Met from one point in the grid.
Show a plot of the expected Met resolution as a function of SumEt. (You are very welcome to ask Joel to help here)
Multiply the Met by 1.1 in every event. Redo plot C. Same for some larger factor. Can you apply a factor based on the Met resolution versus SumEt above ? Estimate the requirement on Met resolution for your analysis (max affordable factor).
Propose a way to use 900 GeV data to study Met.
Propose a set of plots you can show from the 900 GeV data to validate some aspects of the MC (for example noise, jet resolution, electron distributions, number of tracks in the event, jet sizes, etc...).
Make the plots above for MC.
Make the plots above from the cosmic samples.
Think, research and present other possible ways to include systematics. What will be your choice ? Which systematics will you apply as normalzation factorand for which ones will you need reweighting and shape distortions ?
Make a list of expected systematic effects.
How will your studies on the generator uncertainties enter the analysis ?
7 TeV. What is the plan to get 7 TeV samples ?
What trigger will you use ? What is the efficiency for your signal (with turn-on or in the plateau) ? What is the backup trigger ? How will you include the trigger in the analysis (trigger sim, turn-on, overall factor from plateau) ?
Which data stream will you use ? How are going to get the luminosity ? What data quality requirements will you apply ?
For all red boxes on your slide 10, give sample names, numbers, efficiencies, ... What is missing of these ?