Re: [ROOT] TRotation interface

From: Rene Brun (Rene.Brun@cern.ch)
Date: Sun Mar 30 2003 - 18:23:06 MEST


Hi Clark,

It seems that all the functions that you require are already coded
in the classes TRotation, TVector3 or TLorentzVector.

Rene Brun

On 27 
Mar 2003, Clark McGrew wrote:

> Hello,
> 
> Is TRotation still the prefered class to define rotations of TVector3
> objects?  Are TVector3 objects the best vector type to use in physics
> code?  I believe the answer is yes to both questions.  I would use
> TGeoRotation, but it takes Double_t* and not TVector3.  There doesn't
> seem to be a conversion available for TVector3.
> 
> If the answers are yes, then I'm finding that there are relatively few
> methods available to define rotations in TRotation.  If there is
> interest, I volunteer to add a few methods (similar to methods in other
> classes).  A short list might be:
> 
> -- Methods to rotate euler angles with SetPhi, SetTheta, SetPsi (Get as
> well).  The RotateX/Y/Z functions can be used to do euler rotations, but
> only when you start from the identity.
> 
> -- A method to "validate" the rotation matrix to make sure error doesn't
> accumulate.
> 
> -- A method to create a rotation by specifying a vector parallel to the
> rotated Z axis and a vector in the X-Z plane.  
> 
> Best,
> Clark
> 



This archive was generated by hypermail 2b29 : Thu Jan 01 2004 - 17:50:10 MET