Progress on run/position dependent PWC Efficiency

1
Progress on run/position dependent PWC Efficiency
Undertaken to remove (real or perceived)
roadblocks to getting a realistic MC which we
desperately need to do physics

• How we intend to model run/position dependent PWC
  efficiency
• Ways and means
• HIGZ interface. Proposed data structure.
• Checking results
• How important are these corrections?

2
How we intend to model PWC efficiency

• Use the pass1 oddpacks (PWCPAK)which are still on
  disk at ea831
• Combine the efficiency plots in 200 run
  increments
• Choice based on tedium reduction rather than
  physics
• This gives about 30 PWC run periods.
• Smooth / Reduce the data set using a
  stick-figure representation of efficiency versus
  wire number.
• I dont think its wise to use the oddpacks
  directly
• Create a new data file and new environmental
  variable PWCLINES giving the points constituting
  the pwc lines

3
Stick-figure data reduction
Creates file of PWC lines after you click out of
frame
p4y
Should this be included?
Reads data from the cycle file
CAC wrote a HIGZ interface which allows you to
click on the cursor and enter points into a file.
Otherwise this would be way too tedious.
4
Check the parameterized file
Here is an overlay of efficiency from Oddpack and
my pwc lines stored in their file. These check
possible clerical errors. This is for pwc set 4
which consists of 200 runs from 7068-7327. Not at
all sure that all these glitches are real. Note
dramatic fall off in eff at P4x edges.
5
Creating a PWCLINES file
Run period boundaries for the 7 run periods
completed by Sunday noon
Presumably a setenv in a nightly update setup file
Pointer to start of lines for each of 20 planes
for each of the (7?20) run periods
pointer for run period 1 plane 1
actual pwc line data
6
Check the fidelity of the simulation
The simulated profile is not a perfect match to
data. Some curvature develops on the input
straight lines (P4Y) Magnetic effects?? The
PWCPAK efficiency plot is a slightly distorted
representation of the actual efficiency profile.
data
mc

• MC statistics accumulated in about 8 hours on UI
  cluster
• Close enough for government work?
• Depends how important the effect is to start with.

7
Will it make much difference?
present MC new PWC eff
Y2398
Y2402

• Based on MC with 25000 accepted events with
  /without position dependent PWC efficiency ( for
  set 4 7068-7327)
• Present MC took 3609 s and new MC took 3631 s
• Did not think this project would be important but
  thought Id get it out of the way. I hope to
  compete this very soon so it can be removed from
  our list of projects.
• Other run dependencies likely to be more serious

8
Will the pwc efficiencies , HC shower library,
and Pythia6 change the lum factors?
Here is a comparison of golden mode in data to a
new simulation of mixemup where we have included
run dependent PWC profiles, changed the HC
trigger simulation to the shower library and ran
Pythia6 rather than Pythia5. Agreement is not
perfect but not that bad either. It appears that
the luminosity factors are relatively unaffected
by these changes.