class AliRelAlignerKalman: public TObject

Default constructor

copy constructor

assignment operator

destructor

Add all tracks in an ESD event

Adds a full track, returns true if results in a new estimate
  gets the inner TPC parameters from AliESDTrack::GetInnerParam()
  gets the outer ITS parameters from AliESDfriendTrack::GetITSout()

Update the estimate using new matching tracklets

Add an cosmic with separately tracked ITS and TPC parts, do trackmatching

Print some useful info

Print the system matrix for this measurement

Set the parameters, exparam1 will normally be ITS and exparam 2 tht TPC

perform the update

Get Rotation matrix R given the Cardan angles psi, theta, phi (around x, y, z).

Calculate the residuals and their covariance matrix

Calculate the system matrix for the Kalman filter
approximate the system using as reference the track in the first volume

Prepare the prediction of the measurement using state vector

 Implements a system model for the Kalman fit
 pred is [dy,dz,dsinphi,dtgl]
 state is [psi,theta,phi,x,y,z,driftTPC,offsetTPC]
 note: the measurement is in a local frame, so the prediction also has to be
 note: state is the misalignment in global reference system

Kalman estimation of noisy constants: in the model A=1
The arguments are (following the usual convention):
  fPX - the state vector (parameters)
  fPXcov - the state covariance matrix (parameter errors)
  fPMeasurement - measurement vector
  fPMeasurementCov - measurement covariance matrix
  fPH - measurement model matrix ( fPMeasurement = Hx + v ) v being measurement noise (error fR)

check whether an update is an outlier given the covariance matrix of the fit

check if the input residuals are not too far off their median

check for positive definiteness

Produce a valid symmetric matrix out of an almost symmetric TMatrixD

Calculate the Cardan angles (psi,theta,phi) from rotation matrix
b = R*a

Print the correlation matrix for the fitted parameters

Find matching track segments in an event with tracks in TPC and ITS(standalone)

implements the system model -
applies correction for misalignment and calibration to track
track needs to be already propagated to the global reference plane

implements the system model -
applies misalignment and miscalibration to reference track
trackparams have to be at the global reference plane

full reset to defaults

Resets the covariance to the default if arg=0 or resets the off diagonals
to zero and releases the diagonals by factor arg.

Resets the covariance to the default if arg=0 or resets the off diagonals
to zero and releases the diagonals by factor arg.

Merge two aligners

merge all aligners in the collection

find matching tracks and return tobjarrays with the params

Sets up or destroys the memory hungry array to hold the statistics
for data rejection with median

 check covariance matrix
 Return values:
         -1    - everything OK
          >=0  - index of "corrupted" element in the covariance matrix