fix: Remove unit C #2142
fix: Remove unit C #2142
Conversation
andiwand
added
Component - Core
Codecov Report
@@ Coverage Diff @@
## main #2142 +/- ##
==========================================
- Coverage 49.28% 49.27% -0.01%
==========================================
Files 450 450
Lines 25410 25408 -2
Branches 11727 11727
==========================================
- Hits 12523 12521 -2
Misses 4549 4549
Partials 8338 8338
📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more |
|
This probably needs a bit of discussion to understand it fully. |
|
there are quite some significant differences now in the integration test... not sure what was tested here before but it looks like maybe we can discuss this tomorrow @asalzburger @paulgessinger |
|
It should be Coulomb. How to get to the value is explained in PRs #1179 and #1231 . But I don't know if we actually need it as a unit. I am not sure if we should have |
|
I don't see any value in having I don't have a strong opinion towards splitting constants into |
📊 Physics performance monitoring for 2b5899dSummary VertexingVertexing vs. muIVF seededIVF truth_smearedIVF truth_estimatedIVF orthogonalAMVF seededAMVF truth_smearedAMVF truth_estimatedAMVF orthogonalSeedingSeeding seededSeeding truth_estimatedSeeding orthogonalCKFCKF seededCKF truth_smearedCKF truth_estimatedCKF orthogonalAmbiguity resolutionseededTruth tracking (Kalman Filter)Truth trackingTruth tracking (GSF)Truth tracking |
acts-project-service
added
the
Breaks Athena build
|
This was a false failure due to #2251 |
In #2142 I discovered that the ridders propagator depends highly on the step size taken. At the same time we still use the global static surface tolerance in most of the places which I try to improve here. This allows us to consistently set the surface tolerance for the propagation and to improve the accuracy of the ridders propagator. Includes a bug fix for `AtlasStepper` I discovered after making these changes. The stepper ignored the stepping tolerance in backward mode.
Not sure what this is and why it was there in this place? **Note** This contains a very odd fix for a problem I was facing in the integration test which compares the analytical covariance with the numerical one. For some reason the numerical derivative breaks down in the case where we step towards a surface which is 1 cm away with a single step. Using 0.9 cm it works because it will do a second step which seems to provide the necessary accuracy for the differential quotient. This is not pretty and I would like to take a second look why a single step is not sufficient in this case.
…2292) In acts-project#2142 I discovered that the ridders propagator depends highly on the step size taken. At the same time we still use the global static surface tolerance in most of the places which I try to improve here. This allows us to consistently set the surface tolerance for the propagation and to improve the accuracy of the ridders propagator. Includes a bug fix for `AtlasStepper` I discovered after making these changes. The stepper ignored the stepping tolerance in backward mode.
Not sure what this is and why it was there in this place?
Note
This contains a very odd fix for a problem I was facing in the integration test which compares the analytical covariance with the numerical one. For some reason the numerical derivative breaks down in the case where we step towards a surface which is 1 cm away with a single step. Using 0.9 cm it works because it will do a second step which seems to provide the necessary accuracy for the differential quotient.
This is not pretty and I would like to take a second look why a single step is not sufficient in this case.