Minutes:
Attending MK,AM,TA,BF,BT,DH,CZ.
-Random background
Long discussion about character of randoms in (e,e'n). For many
aspects, see minutes of last week.
In regard of TA's event-mixing method to generate random distribution
for Tnn and to study the phase space distribution of randoms in
dynamical variables like p_miss: VZ argues that the pmiss distribution
of the n-hits of the event sample that enters the "mixing" would be
backward-peaked (according to the e,e'n cross section) and rather not
forward-peaked as expected for randoms as most of them are caused by
the forward-LADS counters.
For high-Q2 events, where the location of "random" n-hits is similar
like the location of "true" n-hits, the contribution of randoms
behaves like a misreconstruction of missing momentum. The asymmetry
itself does not depend much on p_miss at higher Q2, thus the asymmetry
and hence GEn would not be affected by this misreconstruction. The
same can be argued for low-Q2, only here the misreconstruction of
missing momentum would be more drastic. In the final sample for GEn, a
cut on low missing momentum is being made, so at low Q2 the fraction
of randoms that survive is quite small, and at high Q2, where the
contribution of randoms may be large, it would not matter for the
asymmetry.
TA's event mixing method should be applied nevertheless. From
comparing the yield of the original events in the unphysical region of
Tnn (superfast neutrons faster than gammas) with the yield in the same
region for the mixed events, one obtains a scaling factor that is to
be applied to the randomized yield in the physical region in order to
estimate its contribution to the real events. From the total yield in
each spin state the scaled mixed yields can be subtracted in order to
obtain background-subtracted yields for each spin state. Only then
asymmetries should be formed. We are curious to see how the results of
this would look like.
-Radiative BLASTMC
NM provided implementation of Mascarad code. VZ implemented Mascarad
into dgen, currently for quasielastic generator. Code was successfully
compiled and produced no errors or crashes. It is however still very
slow.
-Tasklist before recrunch
There are tasks that require "lrn" and others that only require
"lrd". Running "lrd" is about 20 times faster than "lrn".
Tasks that require "lrn":
+Establish good DST. DST now also contains the used calibrations. For
existing otherwise proper DSTs, calibrations are also added by lrd.
+Proper charge calibration: lrd presently only copies EPICS/scalers
tree from lrn output, need to run lrn to generate it.
+Track finding
Tasks that only require "lrd":
+Refitting with improved wire chamber calibrations
+Refitting with improved Blast field map
+Reconstruction with improved neutron timing calibration
Runs <=11250 still need to be recrunched with lrn to provide proper
DSTs in terms of BATs and certain features (takes about 30 days, will be
started soon).
Track finding for those runs will be done using improved wire chamber
calibration, however the overall yield is not expected to change
much.
Runs of 2005 13445-15213 were crunched with the old DCCT
calibration, so also runs of 2004. The old DCCT calibration is only
good up to 150mA. I'm not sure if we significantly exceeded this
range already in 2004. If "lrd" is supplemented with a piece of code
that reads the EPICS/scaler tree, converts DCCT back to raw using the
old calibration, and writes the EPICS/scaler tree using the new
calibration (including a bit for bookkeeping that would indicate
which calibration was used. Only if the bit has not been set yet,
this piece of code would be excuted by lrd). This way running the
expensive lrn for just the reason of charge calibration can be
avoided. If we are happy with the track finding efficieny, all other
tasks besides establishing good DSTs for 2004 runs would be achieved
with "lrd".
ALL runs may still need lrd(s) on top once the neutron timing is
finalized and once the field maps are improved.
I will send out a separate email on the recrunch status soon.
Regards,
Michael
--+-------------------------------------+--------------------------+ | Office: | Home: | |-------------------------------------|--------------------------| | Dr. Michael Kohl | Michael Kohl | | Laboratory for Nuclear Science | 5 Ibbetson Street | | MIT-Bates Linear Accelerator Center | Somerville, MA 02143 | | Middleton, MA 01949 | U.S.A. | | U.S.A. | | | - - - - - - - - - - - - | - - - - - - - - -| | Email: kohlm@mit.edu | K.Michael.Kohl@gmx.de | | Work: +1-617-253-9207 | Home: +1-617-629-3147 | | Fax: +1-617-253-9599 | Mobile: +1-978-580-4190 | | http://blast.lns.mit.edu | | +-------------------------------------+--------------------------+
This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:32 EST