GR updates: GR v15r35 (Heather) This tag has the following goodies:
Alignment aberration code will be coming soon. (Toby) has tested it, but further proof is needed that it does the right thing and that nothing is broken
OBF High Energy Pass (Tracy)
This is the name for the gamma filter letting events through which have more than 20 GeV energy in the calorimeter, provided two ACD related conditions are not also satisfied. However, Philippe noticed a loss of efficiency in what was expected (from earlier MC based on a simulation of this HE Pass in Gleam code) in recent MC runs. This was tracked down and appears to be the result of the implementation, where a third condition must not also be true (the so called "zBottom" condition). Still under investigation.
New merit variables (Philippe) he's looking for some to go into merit, to be used to find accidental coincidences and measure efficiency loss. Currently this has to be done with difficulty from Recon files.
Reprocessing: (Tom)
The development of a reprocessing task is continuing. Small samples of data are successfully reprocessing — at least mechanically. Still (known) to do:
- Create DIGI subsets for each clump (needed for SVAC creation code...skimmer can do this)
- Wire in FT1 merge (use Tony's ft1skim.py script)
- Get hadd to read a file of filenames before unix line length limit is exceeded.
- install new version of SVAC(?) (when it can read from xroot — Anders has updated)
- dataCatalog (routine) and runHistory (new) database bookkeeping step
Once this is completed, some extensive testing is required, both the scalability of the task, and that the output makes physics sense. Current tests are using GR-v15r24, so I am open to suggestion for the best release to use for this level of testing. Also, open to suggestion for a set of runs to reprocess. Currently testing with r0236084237 and r0236090205.
Special thanks to David Chamont for proving that skimmer actually could handle GCR ROOT files, and that one can form a TCut on Entry numbers.
Skimmer: (David C.) There was a problem in the v6r1 version of the skimmer. It will be easy to fix. He will make a new version including this fix, plus:
One can already skim the GCR files provided SK_EXPECTED_RELEASE is set "manually". We must understand why the GCR files have an empty FileHeader
One can already skim the Svac tuples, but cannot yet define the TCut based on Svac.
There are still ongoing investigations about problems skimming some merit files which are fatal to the skimmer; we do not know why.
[thanks to David for most of the above text. ed.]SCons: (Navid) Packages needing review and tagging for handling of CMT environment variables are about 95% done; still about 4 left, which he will discuss with owners.
The new RM is nearly done; he's cleaning up now so that the code will be easier to maintain.
(Emmanuel) Task of writing SConscript files,etc., for all GR packages is just about done, probably will be complete later today. We'll be able to try out building GR with SCons.
(Joanne) GoGui is progressing, but there is still plenty to do; see Plans below. New since last week and visible in the screen shot:
previous | minutes index | next |