Updates of forward tracking since November of 2022 #107
Conversation
…ate dc package to apply new KF package
| @@ -22,7 +23,8 @@ public abstract class AMeasVecs { | |||
|
|
|||
| public void setMeasVecs(List<Surface> measSurfaces) { | |||
| measurements = new ArrayList<>(); | |||
| Collections.sort(measSurfaces); | |||
| if(measSurfaces.get(0).type != Type.LINEDOCA && measSurfaces.get(0).type != Type.PLANEURWELL) // Measurements from URWell and DC has been sorted | |||
There was a problem hiding this comment.
CVT has a similar issue and, for this reason, surfaces have "index"
| } | ||
| } | ||
|
|
||
| // For DC, there could be two hits at a measurement | ||
| public double[] dhDoca(int k, StateVec stateVec) { | ||
| double value[] = {Double.NaN, Double.NaN}; |
There was a problem hiding this comment.
An alternative to redefining "h" for doublets would be to create a new surface with two lines
| } | ||
|
|
||
| // For URWell | ||
| public Surface(int sector, double x, double y, double z, double x_err, double y_err) { |
There was a problem hiding this comment.
isn't the measurement a line? can we use the existing PLANESTRIP?
| public Surface(int region, double z, double x, double tilt, double wireMaxSag, double[] doca, double[] unc, double error, Line3D[] wireLine) { | ||
| type = Type.LINEDOCA; | ||
| this.region = region; | ||
| this.z = z; |
There was a problem hiding this comment.
are all these variables really used? for example, wiremaxsag is not really used in reconstruction
There was a problem hiding this comment.
I suggest to make a separate pull request for URWell
| if(LOGGER.getLevel()==Level.FINE) { | ||
| startTime = System.currentTimeMillis(); | ||
| } | ||
| int q = this.calcInitTrkQ(theta1, theta3, TORSCALE); | ||
| int q = this.calcInitTrkQ(traj.getA(), TORSCALE); |
| for(int i = 0; i < kFZRef.kfStateVecsAlongTrajectory.size(); i++) { | ||
| org.jlab.clas.tracking.kalmanfilter.AStateVecs.StateVec svc = kFZRef.kfStateVecsAlongTrajectory.get(i); | ||
| org.jlab.rec.dc.trajectory.StateVec sv = new org.jlab.rec.dc.trajectory.StateVec(svc.x, svc.y, svc.tx, svc.ty); | ||
| sv.setZ(svc.z); |
There was a problem hiding this comment.
it seems like it would be useful to have a copy method for state vectors
There was a problem hiding this comment.
move to a separate pull request
| @@ -8,6 +8,7 @@ | |||
| import org.jlab.rec.dc.Constants; | |||
| import org.jlab.rec.dc.cross.Cross; | |||
| import trackfitter.fitter.LineFitter; | |||
| import Jama.Matrix; | |||
There was a problem hiding this comment.
is Jama still used?
There was a problem hiding this comment.
move to a separate pull request
…to single or double
…nly from conventional reconstruction into account
The PR is for updates of forward tracking since November of 2022.
Note: some updates, such as new conditions for iteration termination, issue fixing for fast math package have been merged into development for pass2 cooking.
Main updates in the PR include:
Here are links for talk 1 and talk 2 , which summarize updates.
Here is a link for validation