Skip to content

Stricter clone removal and larger VELO seeding window

Maarten Van Veghel requested to merge mveghel-extracloneremoval into 2024-patches

Needs Alignment!524 (merged) !3564 (merged) (to fix failing tests in Moore)

Adds stricter clone removal, in particular long tracks with same T-track but not fully overlapping VELO tracks but same direction (split VELO segments?), as seen by @gciezare, see also his slides

The change comprises and additional case where long versus long clone checking considers a clone if the T segment has overlap (this case >50%), but the VELO track has not, but ends up with the same momentum, so highly likely a very similar VELO track was used. Would for example kill split VELO tracks (one track split up into two in case of e.g. missing hit in between (not working module, etc...)).

Decreasing chance of these kind of hypothetical split VELO tracks would be achieved by switching on changes made here Rec!3895 (merged). This is included by going enlarging the seeding window from 3 to 4.

Locally ran test looks promising

-LongGhostFilteredTrackChecker_93...    INFO Results
-LongGhostFilteredTrackChecker_93...    INFO **** LongGhostFiltered                          4475 tracks including            220 ghosts [ 4.92 %], Event average  4.22 % ****
-LongGhostFilteredTrackChecker_93...    INFO   01_long                                  :    3784 from     4307 [ 87.86 %]      5 clones [ 0.13 %], purity: 99.48 %, hitEff: 97.84 %
-LongGhostFilteredTrackChecker_93...    INFO   02_long_P>5GeV                           :    2508 from     2741 [ 91.50 %]      2 clones [ 0.08 %], purity: 99.53 %, hitEff: 98.31 %
-LongGhostFilteredTrackChecker_93...    INFO   03_long_strange                          :     178 from      224 [ 79.46 %]      0 clones [ 0.00 %], purity: 99.47 %, hitEff: 97.94 %
-LongGhostFilteredTrackChecker_93...    INFO   04_long_strange_P>5GeV                   :      90 from      107 [ 84.11 %]      0 clones [ 0.00 %], purity: 99.64 %, hitEff: 98.79 %
-LongGhostFilteredTrackChecker_93...    INFO   05_long_fromD                            :      46 from       47 [ 97.87 %]      0 clones [ 0.00 %], purity: 99.84 %, hitEff: 97.85 %
-LongGhostFilteredTrackChecker_93...    INFO   06_long_fromD_P>5GeV                     :      29 from       30 [ 96.67 %]      0 clones [ 0.00 %], purity: 99.74 %, hitEff: 98.13 %
-LongGhostFilteredTrackChecker_93...    INFO   07_long_electrons                        :     170 from      295 [ 57.63 %]      1 clones [ 0.58 %], purity: 98.23 %, hitEff: 94.81 %
-LongGhostFilteredTrackChecker_93...    INFO   07_long_electrons_pairprod               :     162 from      282 [ 57.45 %]      1 clones [ 0.61 %], purity: 98.19 %, hitEff: 94.63 %
-LongGhostFilteredTrackChecker_93...    INFO   10_long_fromD_P>3GeV_Pt>0.5GeV           :      26 from       26 [100.00 %]      0 clones [ 0.00 %], purity: 99.71 %, hitEff: 97.79 %
-LongGhostFilteredTrackChecker_93...    INFO   10_long_strange_P>3GeV_Pt>0.5GeV         :      43 from       50 [ 86.00 %]      0 clones [ 0.00 %], purity: 99.79 %, hitEff: 99.09 %
-LongGhostFilteredTrackChecker_93...    INFO
+LongGhostFilteredTrackChecker_d5...    INFO Results
+LongGhostFilteredTrackChecker_d5...    INFO **** LongGhostFiltered                          4427 tracks including            189 ghosts [ 4.27 %], Event average  3.73 % ****
+LongGhostFilteredTrackChecker_d5...    INFO   01_long                                  :    3775 from     4307 [ 87.65 %]      5 clones [ 0.13 %], purity: 99.48 %, hitEff: 97.85 %
+LongGhostFilteredTrackChecker_d5...    INFO   02_long_P>5GeV                           :    2504 from     2741 [ 91.35 %]      2 clones [ 0.08 %], purity: 99.53 %, hitEff: 98.32 %
+LongGhostFilteredTrackChecker_d5...    INFO   03_long_strange                          :     178 from      224 [ 79.46 %]      0 clones [ 0.00 %], purity: 99.47 %, hitEff: 97.94 %
+LongGhostFilteredTrackChecker_d5...    INFO   04_long_strange_P>5GeV                   :      90 from      107 [ 84.11 %]      0 clones [ 0.00 %], purity: 99.64 %, hitEff: 98.79 %
+LongGhostFilteredTrackChecker_d5...    INFO   05_long_fromD                            :      46 from       47 [ 97.87 %]      0 clones [ 0.00 %], purity: 99.84 %, hitEff: 97.85 %
+LongGhostFilteredTrackChecker_d5...    INFO   06_long_fromD_P>5GeV                     :      29 from       30 [ 96.67 %]      0 clones [ 0.00 %], purity: 99.74 %, hitEff: 98.13 %
+LongGhostFilteredTrackChecker_d5...    INFO   07_long_electrons                        :     165 from      295 [ 55.93 %]      1 clones [ 0.60 %], purity: 98.26 %, hitEff: 94.82 %
+LongGhostFilteredTrackChecker_d5...    INFO   07_long_electrons_pairprod               :     157 from      282 [ 55.67 %]      1 clones [ 0.63 %], purity: 98.22 %, hitEff: 94.64 %
+LongGhostFilteredTrackChecker_d5...    INFO   10_long_fromD_P>3GeV_Pt>0.5GeV           :      26 from       26 [100.00 %]      0 clones [ 0.00 %], purity: 99.71 %, hitEff: 97.79 %
+LongGhostFilteredTrackChecker_d5...    INFO   10_long_strange_P>3GeV_Pt>0.5GeV         :      43 from       50 [ 86.00 %]      0 clones [ 0.00 %], purity: 99.79 %, hitEff: 99.09 %
+LongGhostFilteredTrackChecker_d5...    INFO
Edited by Miroslav Saur

Merge request reports