Implement Combinatorial Kalman Filter

[XiaocongAi]

This MR includes the following:

• Implement the Combinatorial Kalman Filter for track finding&fitting
• Implement the source link selector to select the measurements/outlier
• Implement the branch stopper plugin to allow stopping bad branch

[codecov]

Codecov Report

Merging #102 into master will decrease coverage by 1.52%.
The diff coverage is 29.03%.

@@            Coverage Diff             @@
##           master     #102      +/-   ##
==========================================
- Coverage   44.50%   42.98%   -1.53%     
==========================================
  Files         357      354       -3     
  Lines       17592    17147     -445     
  Branches     8260     8130     -130     
==========================================
- Hits         7830     7371     -459     
- Misses       4807     4824      +17     
+ Partials     4955     4952       -3     

Impacted Files	Coverage Δ
...cts/TrackFinder/CombinatorialKalmanFilterError.hpp	0.00% <0.00%> (ø)
...ude/Acts/TrackFinder/CombinatorialKalmanFilter.hpp	27.32% <27.32%> (ø)
...include/Acts/TrackFinder/CKFSourceLinkSelector.hpp	41.57% <41.57%> (ø)
...s/TrackFinder/detail/VoidTrackFinderComponents.hpp	100.00% <100.00%> (ø)
Core/src/Geometry/BinUtility.cpp	0.00% <0.00%> (-100.00%)	⬇️
Core/include/Acts/Propagator/AtlasStepper.hpp	18.37% <0.00%> (-50.68%)	⬇️
...Propagator/detail/PointwiseMaterialInteraction.cpp	87.50% <0.00%> (-12.50%)	⬇️
Core/include/Acts/Utilities/BinUtility.hpp	39.65% <0.00%> (-5.18%)	⬇️
Core/src/Propagator/detail/CovarianceEngine.cpp	45.23% <0.00%> (-2.27%)	⬇️
Core/include/Acts/Propagator/EigenStepper.ipp	35.77% <0.00%> (-2.19%)	⬇️
... and 45 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 5770f23...5770f23. Read the comment docs.

[paulgessinger]

This looks like a ton of work! Had a quick look and I think it looks really good. I'll try to have a closer look as soon as possible, but if anyone else wants to have a look, go ahead.

[paulgessinger]

Sorry for the tons of comments, but it's a huge amount of code. Looks like awesome work! Will approve once we resolve the discussion items.

[asalzburger]

I think the unordered_map misses an #include statement, as the macOS build fails.

[XiaocongAi]

@paulgessinger Thank you for your good suggestions for this PR! I have updated/optimized the code as much as possible. Will rebase it tomorrow.

[paulgessinger]

Approval is out!
The build failure in the coverage job is "No space left on device", which is hopefully transient. I hit retry, let's see.

[paulgessinger]

The CI fails because it points a commit that was most likely force pushed away. @XiaocongAi can you rebase and/or push an empty commit to trigger a complete pipeline?

[asalzburger]

@XiaocongAi - if you rebase & push, I will merge in later.

[XiaocongAi]

@asalzburger . it's rebased. But the CI fails due to "No space left on device". How could I deal with this?

[paulgessinger]

@asalzburger . it's rebased. But the CI fails due to "No space left on device". How could I deal with this?

Hm, I saw this message before, but if I check the log now I still see

2020-04-22T20:14:16.2364578Z ##[error]fatal: reference is not a tree: 0b4df649b842be6dff35af7e70483db3c65ff084

That is not the tip of the branch however, I believe. Not sure why that job tries to check that ref out.

[paulgessinger]

I override the coverage failure, because I think it's a technical problem, which I don't want to hold this PR back for. (See #124)