Fix phi0 type in PCAInfo, correct omega sign, and add covariance matrix to TrackStates

[andread3vita]

BEGINRELEASENOTES

• Correct the data type of phi0 when saving it in PCAInfo
• Correct the sign of omega by properly determining the particle charge during initialization
• Add covariance matrix to fitted trackStates
• Fix how filtered tracks are saved in the output file: now for each fitted track there is the corresponding filtered track
• Expose EDM4Hep units to the user

ENDRELEASENOTES

Hello everyone,

This pull request is opened to address a couple of issues identified in the current reconstruction:

• correction of the phi0 type when storing it in PCAInfo
• improvement of the omega sign determination through a correct charge estimation during initialization
• Implementation of the covariance matrix for fitted trackStates

Thank you!

Andrea

[BrieucF]

Hi Andrea, thanks for this! Is it possible to have the user facing part (i.e. mainly the Gaudi properties) following the EDM4hep conventions?

[BrieucF]

Thanks!