p17 Fixed Data and Monte Carlo for Summer
Frequently Asked Questions
- How many datasets do I need to get all data ?
The full data for any given skim is the union of three datasets. These can be
described as followed:
- Refixed.
- Unfixed, reskimmed (non-cable-swap).
- Unfixed, reskimmed (cable-swap).
Details can be found in this section of this web page.
These data sets do not overlap and are not different passes of the
same data.
- Why are some files called called PASS2 and PASS3 ?
This is purely historical, depending on the production history.
- So what is the difference then ? PASS3 data is the older data in terms of data taking.
It was reprocessed before the hadronic calibration was available. It was then fixed afterwards and skimmed again. The
PASS2 data is newer (> September 2005) and reconstructed with p17.09.00 which already had the hadronic
calibration.
- What should I use then for the luminosity calculation ? Use -pass3 which will give you all
parentage files you need.
- Are there duplicate events in these datasets ? Yes, but only in the CAF trees due to errors in
the production process. You should remove them on the analysis level. Fixing this would require declaring
a lot of files as bad and reprocessing them, then redoing all the parentage files. This should not be the
case in the next CAF tree production...
- Are there duplicate events in the Monte Carlo ?Yes, due to a bug in the random number initalization
in
mc_runjob
. Again, you should remove these on the CAF tree level in your analysis.
- What are the definitions for the different skims ? The definition of physical and logical skims is here, while the definition of all event flags is here
- How do I include reprocessed cable-swap data in my analysis?
Use skimmed refixed data, and
reskimmed unfixed data.
To find data from the pass2 p17 fixing, use the following constraints:
sam translate constraints --dim="appl_name p17tmbfixer and version p17.09.03 and data_tier thumbnail"
or use this dataset: CSG_P17fixing_Pass2.
All data reconstructed with p17.09.00 already includes the fixing. You can
find the unskimmed p17.09.00 data with these constraints:
sam translate constraints --dim="(((APPL_NAME recon_root and VERSION p17.09.00) and DATA_TIER thumbnail) and FILE_NAME recoT_all_%)"
Cable swap data was reprocessed using p17.09.06 and likewise does not need to
be fixed. You can find the unskimmed p17.09.06 data using
these constraints:
sam translate constraints --dim="(((APPL_NAME recon_root, d0reco and VERSION p17.09.06) and DATA_TIER thumbnail) and FILE_NAME recoT_all_%)"
The complete skimmed dataset is the union of
skimmed refixed data and
reskimmed unfixed data. The latter data includes
non-cable-swap and cable-swap data. Thus, for any skim, the complete data
set includes at least three predefined datasets for thumbnails:
- CSskim-XXXX-PASS3-p17.09.03 (refixed).
- CSskim-XXXX-PASS3-p17.09.06 (unfixed, cable-swap).
- CSskim-XXXX-PASS3-p17.09.06b (unfixed, non-cable-swap).
or CAF trees:
- CSG_CAF_XXXX_PASS3_p17.09.03 (refixed).
- CSG_CAF_XXXX_PASS3_p17.09.06 (unfixed, cable-swap).
- CSG_CAF_XXXX_PASS3_p17.09.06b (unfixed, non-cable-swap).
where XXXX is the name of the skim.
Run Ranges
Here is information about where to find specific run ranges in the above
datasets.
- Runs 151817-166493 -- Reconstructed using p17.09.00 d0reco.
Included in p17.09.06b datasets.
- Runs 166503-210543 -- Reconstructed using pre-p17.09.00 d0reco and
(re)fixed. Included in p17.09.03 datasets.
- Runs 210543-212107 -- Reconstructed using p17.09.00 d0reco and reskimmed.
Included in p17.09.06b datasets.
- Runs 212900-215670 -- Cable swap data. Reconstructed (reprocessed)
using p17.09.06 d0reco. Included in p17.09.06 datasets.
Detailed information about Run 2 Monte Carlo samples is available on
the Run 2 MC
web page.
To find skimmed thumbnails, use the following constraint:
sam translate constraints --dim="(((APPL_NAME tmbskim and VERSION csskim-pass3-p17.09.03) and DATA_TIER thumbnail) and FILE_NAME CSskim-XXXX-%)"
or the pre-defined dataset CSskim-XXXXX-PASS3-p17.09.03 where XXXXX is the skim name.
To find CAF trees for the skimmed datasets, use the following constraint:
sam translate constraints --dim="((APPL_NAME tmb_analyze and VERSION csgcaf_02.00.00) and FILE_NAME CSG-CAFv3-CSskim-XXXX-%)"
or the pre-defined dataset CSG_CAF_XXXXX_PASS3_p17.09.03 where XXXXX is the skim name.
These CAF trees are produced with p18.05.00.
These data are affected by the "cable-swap" problem and
have been superseded by reskimmed unfixed data
(see below).
Descriptions of the old skims for thumbnails and caf trees can be found
here.
Cable-swap data taken near the end of Run IIa have been re-reconstructed
using d0reco p17.09.06 (these data were originally reconstructed using
d0reco p17.09.00). However, all data originally reconstructed using
d0reco p17.09.00 have been reskimmed. Thus, the reskimmed data has
two parts:
- Cable-swap data that has been re-reconstructed using d0reco p17.09.06.
- Non-cable-swap data that was originally reconstructed using p17.09.00
and not re-reconstructed.
Reskimmed thumbnails can be found using the following sam query:
- Cable-swap data:
sam translate constraints --dim="(((APPL_NAME tmbskim and VERSION csskim-pass3-p17.09.06) and DATA_TIER thumbnail) and SKIM.NAME xxxx)"
- Non-cable-swap data:
sam translate constraints --dim="(((APPL_NAME tmbskim and VERSION csskim-pass3-p17.09.06b) and DATA_TIER thumbnail) and SKIM.NAME xxxx)"
Reskimmed CAF trees can be found using the following sam query:
- Cable-swap data:
sam translate constraints --dim="(((APPL_NAME tmb_analyze and VERSION csgcaf-02.00.02) and DATA_TIER root-tree-bygroup) and SKIM.NAME xxxx)"
- Non-cable-swap data:
sam translate constraints --dim="(((APPL_NAME tmb_analyze and VERSION csgcaf-02.00.02b) and DATA_TIER root-tree-bygroup) and SKIM.NAME xxxx)"
These CAF trees are produced with p18.05.00.
This section describes skimmed datasets for TMBs and CAF trees that were used
for the run 2a recaffing project using production releases p18.13.00 and
p18.13.01, plus p18.14.00 for photon hits-in-road recaffed samples.
The TMB datasets listed here contain the same data as the older skimmed
TMB datasets described above, but they have been combined into a single
dataset for each skim and for the all stream.
The CAF tree datasets are completely newly generated using
production release p18.13.00, p18.13.01, or p18.14.00.
Unskimmed Data
The following dataset has been defined for the entire Run 2a data
all stream, which is the union of the datasets given above (p17.09.03
refixed, p17.09.00 unfixed, and p17.09.06 reprocessed cable swap).
- CSG_Skim_p17.09.xx - Entire Run 2a all stream.
Skimmed TMBs and CAF Trees
Datasets for skimmed TMBs and recaffed CAF trees are given in the following
table. The following constraint can
be used find find recaffed CAF trees made using p18.13.00 or p18.13.01.
sam translate constraints --dim="APPL_NAME tmb_analyze and VERSION csg-p18.13.00, csg-p18.13.01 and DATA_TIER root-tree-bygroup and SKIM.NAME xxxx"
The following constraint can
be used find find recaffed CAF trees made using p18.14.00..
sam translate constraints --dim="APPL_NAME tmb_analyze and VERSION csg-p18.14.00 and DATA_TIER root-tree-bygroup and SKIM.NAME xxxx"
where xxxx is the (lower case) name of the skim.
Comments to CSG Conveners
Last updated: Nov. 15, 2006