P20 Pass 2 Data and MC (Preshutdown 2007)
Here is a summary
of the major d0reco versions that have been used in run 2b so far.
- p20.04.01 -- Pass 1 (obsolete).
- p20.07.01/p20.08.xx -- Pass 2.
- p20.11.xx -- Pass 3 (obsolete).
- p20.12.xx -- Pass 4.
Pass 1 and pass 3 are obsolete. The full run 2b dataset is therefore
the union of pass 2 (preshutdown 2007) and pass 4 (postshutdown 2007)
data.
The following releases were used to generate data for analysis.
- Skimming - p21.03.00.
- Initial caf tree production - p21.05.00.
- Recaffing - p21.10.00 and p21.11.00.
As a base release for your analysis, use a release that is at least as
recent as the release that was used to generate the tmbs or caf trees
that you are analyzing (for recaffed data, that means p21.10.00 or
later). Normally, more recent frozen releases are better for analyzing
any kind
of pass 2 data.
Here is a brief summary of differences since p21.03.00.
- p21.03.00 - Update skimming for pass 2 data.
- p21.04.00 - Update tmb_analyze (caf production) for pass 2 data.
- p21.05.00 - Fix tmb_analyze EM spatial track match bug.
- p21.06.00 - Fix tmb_analyze per tick luminosity information for
mc min bias overlay.
- p21.07.00 - Updated JES. Updated trigger branches in CAF.
- p21.08.00 - Final run 2a JES. EM scale and track match fixes.
- p21.09.00 - Central MC lumi profiles. Updated EM smearing, NN
tagger trfs.
- p21.10.00 - Recaffing (updated emid, track misses, additional
branches).
Pass 2 unfixed data have been reconstructed
using d0reco versions p20.07.01, p20.08.00, or p20.08.01. For most
purposes,
these releases can be considered to be equivalent. P20.07.01 was used
to reprocess data on remote farms. P20.08.00 and p20.08.01 were used on
the local Fermilab
farm for reprocessing and for reconstruction of newly taken data.
To find unfixed p20.07.01 data in sam, use the following
constraints:
sam translate constraints --dim="(((APPL_NAME recon_root, d0reco and VERSION p20.07.01) and DATA_TIER thumbnail) and TRIG_CONFIG_TYPE physics) and FILE_NAME recoT_all_%"
To find unfixed p20.08.xx data in sam, use the following constraints:
sam translate constraints --dim="(((APPL_NAME recon_root, d0reco and VERSION p20.08.%) and DATA_TIER thumbnail) and TRIG_CONFIG_TYPE physics) and FILE_NAME recoT_all_%"
When using the genLBNtables
script for the generation of the
parentage tables use the option -pass
p21pass2.
The following predefined datasets can also be used.
A subset of pass 2 data has been fixed using fixer release p20.11.01 to
correct the "warm cell" problem. The run range for fixed data is
229538-230365. These data are about 6% of the
full run 2b preshutdown data.
Warm Cell Data
Warm cell data are from runs 229538-230365.
Unfixed data from this run range can be found using the following
predefined
sam dataset.
Fixed all stream data can be found using the following constraints:
sam translate constraints --dim="(APPL_NAME tmbfixer and VERSION csg-p20.11.01) and DATA_TIER thumbnail"
Or use the following predefined dataset.
There is also a unified all stream dataset giving fixed or unfixed data
depending on run number.
Fixing Algorithm
Here is information about the p20.11.01 fixing algorithm. In general,
the purpose of this fixing pass is to remove a hot calorimeter cell.
Fixing is finished. Fixed datasets cover just
the subset of run 2b preshutdown data that was fixed (that is runs
229538-230365). Unified datasets cover the
full run 2b preshutdown run range. Unified datasets are the union of
fixed data and unfixed data, depending on run number. It is
recommended to use unified datasets for analyzing run 2b preshutdown
data.
When using the genLBNtables
script for the generation of the parentage
tables use the option -pass
p21pass2fixed, even when using the unified
datasets (the genLBNtables script will pick up the correct parentage also for
the unfixed files).
More detailed information about Run 2 Monte Carlo samples is available on
the Run 2 MC
web page.
Recaffed MC Samples
To find recaffed p20 Monte Carlo caf trees, use the following constraints:
sam translate constraints --dim="(((APPL_NAME tmb_analyze and VERSION csg-p21.11.00-v4 and DATA_TIER root-tree-bygroup) and GLOBAL.REQUESTID XXXXX)
or use the predefined datasets
CSG_CAF_MCv4-XXXXX_p21.11.00
where XXXXX is the
request id.
Older MC Caf Samples
To find unrecaffed p20 Monte Carlo caf trees use the following
constraints:
sam translate constraints --dim="(((APPL_NAME tmb_analyze and VERSION csg-RELEASE) and DATA_TIER root-tree-bygroup) and GLOBAL.REQUESTID XXXXX)
or use the predefined datasets
CSG_CAF_MCv2-XXXXX_RELEASE
where RELEASE is the
tmb_analyze release, and XXXXX is the
request id.
Recommended tmb_analyze releases and datasets associated with
different request ids. are as follows:
- p21.04.00 - Not recommended.
- p21.05.00 - Not recommended.
- p21.06.00 - Bug in EM cluster energy (effect of a few percent
level).
- p21.08.00 - dataset CSG_CAF_MCv2-XXXXX_p21.08.00.
Other information
Note: p20 ALPGEN samples are generated with ALPGEN v2.11. In first
generated p20 ALPGEN samples, the matching was not turned on. Only
those p20 ALPGEN samples with request ids > 66600 are
properly matched and can be used in analyses. PYTHIA samples are not
affected.
For individuals who would like to run tmb_analyze on some MC samples
by themselves, one could do:
setup D0RunII p21.11.01 -O SRT_QUAL=maxopt
setup d0tools
runTMBAnalyze -format=MCSmear -fpe -maxopt -defname=req-id-${REQID}-tmb-good-genuine -${BATCH} -scratch=${SCRATCH}
Click on the following link for a complete list of logical and physical
skims
in p21.03.00 skimming.
For the old frozen summer 2007 datasets, refer to this page.
Thumbnails
Unfixed have been skimmed using release p21.03.00.
To find skimmed thumbnails in sam, use the following constraints:
Unfixed Data
sam translate constraints --dim="(((APPL_NAME tmbskim and VERSION csg-p21.03.00,csg-p21.03.00r) and DATA_TIER thumbnail) and SKIM.NAME xxxxx)"
or use the predefined dataset CSskim-XXXXX-PASS2-p21.03.00
where XXXXX is the skim name.
CAF Trees
CAF trees have been generated for all skimmed unfixed data using
production release p21.05.00.
To find CAF trees made from skimmed thumbnails in sam, use the
following constraints:
sam translate constraints --dim="(((APPL_NAME tmb_analyze and VERSION csg-p21.05.00,csg-p21.05.00a) and DATA_TIER root-tree-bygroup) and SKIM.NAME xxxxx)
or use the predefined datasets:
CSG_CAF_XXXXX_PASS2_p21.05.00_unfixed2007
where XXXXX is the skim name.
When using the genLBNtables
script for the generation of the
parentage
tables use the option -pass p21pass2.
Datasets
Note that the data described in this section only
covers the warm cell run range.
Thumbnails
Fixed data have been skimmed using release p21.03.00.
To find skimmed fixed thumbnails in sam, use the following constraints:
sam translate constraints --dim="(((APPL_NAME tmbskim and VERSION csg-p21.03.00-fix) and DATA_TIER thumbnail) and SKIM.NAME xxxxx)"
or use the predefined dataset CSskim-XXXXX-PASS2-p21.03.00-fix2007
where XXXXX is the skim name.
CAF Trees
CAF trees have been generated for all skimmed fixed data using
production release p21.05.00.
To find CAF trees made from skimmed fixed thumbnails in sam, use the
following constraints:
sam translate constraints --dim="(((APPL_NAME tmb_analyze and VERSION csg-p21.05.00-fix) and DATA_TIER root-tree-bygroup) and SKIM.NAME xxxxx)
or use the predefined datasets:
CSG_CAF_XXXXX_PASS2_p21.05.00_fixed2007
where XXXXX is the skim name.
Unified data and datasets consist of the union of fixed and unfixed
data, depending on run number (fixed data in the warm cell run range,
unfixed data otherwise). The unified datasets cover the full run 2b
preshutdown run range and contain the identical events as the unfixed
datasets.
The sam constraints for extracting unified data from sam are quite
complicated. Therefore, it is recommended to use predefined datasets
given
below.
Regenerated CAF Trees
Regenerated CAF trees have been generated starting from unified skimmed
tmb data using production release p21.10.00.
To find regenerated CAF trees in sam, use the following constraints.
sam translate constraints --dim="(((APPL_NAME tmb_analyze and VERSION csg-p21.10.00-p20.07.01-p20.08.xx) and DATA_TIER root-tree-bygroup) and SKIM.NAME xxxxx)
or use the predefined datasets:
CSG_CAF_XXXXX_PASS2_p21.10.00
where XXXXX is the skim name.
When using the genLBNtables
script for the generation of the
parentage
tables use the option -pass
p21pass2fixed, even when using the unified
datasets (the genLBNtables script will pick up the correct parentage also for
the unfixed files). If you are analyzing the regenerated CAF trees,
please use the option -pass
p21pass2recaf.
Datasets (Unified Data)
Comments
to CSG Conveners
Last updated: Apr 8, 2008