| new GENFIT release [message #10498] |
Wed, 31 March 2010 23:33  |
|
Anonymous Poster
|
|
From: *adsl.alicedsl.de
|
|
Dear colleagues,
there is a new release of GENFIT for PANDAroot. If you do an update on your trunk you will get revision 183 of GENFIT. The most important changes with respect to the old version are:
- the Kalman filter had a small problem which was leading to slightly wrong error estimations. The covariance matrix must be multiplied with a large factor before the fit is repeated in reverse order of hits. So far this factor was 20, which was too small. Now the factor is 500.
- we now have the possibility to use finite planes for tracking. It means that a MVD wafer for instance has a limited size and shape, e.g. a rectangle. This is particularly important for fitting looping tracks. If you are interested on details, please ask any questions, I will answer them.
- this GENFIT version is compatible with a new track representation we are working on here. It has the same track parametrization as GeaneTrackRep, also uses ROOT geometry, and is based on a Runge-Kutta algorithm. It will work for fitting tracks in the forward spectrometer. It will be introduced in detail at the next PANDA meeting.
Best regards, Christian
|
|
|
|
|
|
| Re: new GENFIT release [message #10500 is a reply to message #10499] |
Wed, 31 March 2010 23:57  |
|
Anonymous Poster
|
|
From: *adsl.alicedsl.de
|
|
Hi Johan,
thanks for noticing. I am aware of that. That is not an incompatibility. We will make a small change to GeaneTrackRep2 in the near future. But actually GeaneTrackRep2 is not really used in PANDAroot, so it isnt very important. PANDAroot uses GeanTrackRep mostly. The other version was written by me to distribute to other projects as it does not depend on any Fair classes.
LSLTrackRep is outdated and is not currently used in PANDAroot (as far as I know). But the old version is compatible with the new GENFIT release as well.
Cheers, Christian
|
|
|
|
|
|
| Re: new GENFIT release [message #10504 is a reply to message #10503] |
Thu, 01 April 2010 10:47 |
|
Anonymous Poster
|
|
From: *adsl.alicedsl.de
|
|
Hi Alicia,
ooops sorry about my wrong statement. You can further use it, but soon we will have RKTrackRep which might work better for you. What are the particles you are tracking. What kind and what momentum?
Cheers, Christian
|
|
|
|
|
|
GSI Forum
Search
Help
Members
Login
Home
. 
