InnerDetector: Separation and specialisation of strip/beam frame transformations to resolve polar annulus digitisation faliures

Previously, digitisation of endcap tracks would fail when a polar co-ordinate system was used. Additionally, errors were seen from the StripSurfaceChargeGenerator when calculating the drift distance. These problems were found to be caused by the conversion between beam and strip frames, which was generally performed in the assumption of cartesian co-ordinates.

This merge request consolidates the beam frame <--> strip frame transforms in StripStereoAnnulusDesign into their own methods, and provides polar specialisations. The other methods in StripStereoAnnulusDesign have been modified to rely on these methods, rather than applying their own transform.

Two versions of the polar specialisations of the transform are used - one which first converts the co-ordinate to cartesian, applies the transform, and then converts back to polar, and another version which works exclusively in a polar co-ordinate system. The derivations of these transforms are detailed at ATL-COM-ITK-2021-048. These will later be profiled to determine the optimal method.

This MR is part of ATLITKSW-42. @nstyles @pagessin @hhayward

added 21.9 ITk InnerDetector Tracking Upgrade labels

added 2 commits

• 5063b479 - Added PolarAnnulus shape enum
• e1786516 - Added polar specialisation of hitLocalToLocal

Compare with previous version

added 1 commit

• 27411c95 - Added PolarAnnulus to all logic testing for Annulus shape

Compare with previous version

added 1 commit

• 9bc4be6f - Fixing polar->polar transform in cellIdOfPositionPC

Compare with previous version

added 1 commit

• 270cd1f3 - First fixes for !47748 (merged) comments

Compare with previous version

added 20 commits

• 270cd1f3...54f04dee - 16 commits from branch atlas:21.9
• 28ae6819 - PC version of SAD::cellIdOfPosition, and lots of debyg statements
• 95e366f7 - Some more debug statements
• 9f310d5e - Fixing polar->polar transform in cellIdOfPositionPC
• 1081a857 - Consolidate transformations between beam and strip frames into separate methods

Compare with previous version

added 7 commits

• 1081a857...2e5b864f - 2 commits from branch atlas:21.9
• 380a9013 - PC version of SAD::cellIdOfPosition, and lots of debyg statements
• 80c192d2 - Some more debug statements
• 327614cd - Fixing polar->polar transform in cellIdOfPositionPC
• 9b4a40ef - Consolidate transformations between beam and strip frames into separate methods
• 05f44a5d - Fixes and debug for ysurf out of bounds and digi

Compare with previous version

Hi @jlsmith, what's the status of this one?

added 1 commit

• 5cc67ca5 - Fix digi? +debug statements

Compare with previous version

added 2 commits

• 194f9632 - Removed unnecessary debug statements, changed others to appropriate level
• 1a960836 - Reverted private changes to .gitignore and CMake

Compare with previous version

added 2 commits

• 403a24c2 - Separated polar version of strip<->beam transforms to own methods for future debugging and use
• 3024e2bc - Further consolidation of cart and polar versions of methods

Compare with previous version

added 1 commit

• e9e3033b - consolidation of inline cart/polar methods

Compare with previous version

added 2 commits

• 1129bbef - Remove whitespace changes
• 83fede30 - Remove whitespace changes

Compare with previous version

added 1 commit

• 3387bc2d - Remove whitespace changes

Compare with previous version

added 27 commits

• 3387bc2d...f26e6cf3 - 26 commits from branch atlas:21.9
• 828b393a - Resolve merge conflicts

Compare with previous version