Local reconstruction in cells

Or simply local correction? I do not know anymore

There are two ways

Map correction with volume tracks

Studied by Valeri, four plates at the times, correcting and computing

Mapping together small cells

Cells can be processed in parallel with HTCondor. First sub and sh scripts have been written, but they mostlly use link.rootrc and creates hundreds of files! Not viable in the long terms.

The files then can be merged together.

The procedure is:

• create symbolic link to tracks.raw.root as usual;

• copy the link.rootrc files from https://github.com/SND-LHC/emu_reco_analysis/blob/master/linkingmap/firstlink.rootrc and https://github.com/SND-LHC/emu_reco_analysis/blob/master/linkingmap/secondlink.rootrc;

• copy https://github.com/SND-LHC/emu_reco_analysis/blob/master/linkingmap/firstlinking_cell.sh and https://github.com/SND-LHC/emu_reco_analysis/blob/master/linkingmap/secondlinking_cell.sh

• check that the files options are correct for the analyzed brick

• copy firstlink.rootrc in link.rootrc

• Submit in condor https://github.com/SND-LHC/emu_reco_analysis/blob/master/condor_scripts/condorlinking_first.sub

• Check if files have been produced correctly with https://github.com/SND-LHC/emu_reco_analysis/blob/master/linkingmap/filecheck.sh. Eventually rerun condor for missing files with https://github.com/SND-LHC/emu_reco_analysis/blob/master/condor_scripts/condorlinking_first_reco.sub

• copy secondlink.rootrc in link.rootrc

• Submit in condor https://github.com/SND-LHC/emu_reco_analysis/blob/master/condor_scripts/condorlinking_second.sub

• Again, check files with https://github.com/SND-LHC/emu_reco_analysis/blob/master/linkingmap/filecheck_secondlinking.sh. Eventually, rerun condor with https://github.com/SND-LHC/emu_reco_analysis/blob/master/condor_scripts/condorlinking_second.sub

• Merge linked files with https://github.com/SND-LHC/emu_reco_analysis/blob/master/condor_scripts/merge_cells.sub

• Create a symbolic link to the merged file with https://github.com/SND-LHC/emu_reco_analysis/blob/master/run1analysis/recoparameters/create_mergedcoupleslink_miccern.sh;

Overlapping cells

Cells must overlap (about 2.5 mm for 1 cm cell), at least during alignment and tracking!

Otherwise we lose a lot of tracks going from a cell to another!

Passing tracks can be actually used to match overlapping cells together, so the corrections go smootly from a cell to another.

Cleaning

Due to the high number of created files, please remove the files after the merge and quality check

Monitor eos quota

export EOS_MGM_URL=root://eospublic.cern.ch

eos quota /eos/experiment/sndlhc