Re: [BLAST_ANAWARE] branch line in azimuthal angle

From: Tong-Uk Lee (tong@MIT.EDU)
Date: Thu Apr 03 2003 - 10:33:57 EST


Oh, by the way, I am not advocating my system, which
is used in most of current BlastLib2.
I am willing to change and follow any convention.
Either [0,2Pi) or [-Pi/2,3Pi/2) is fine with me.
Actually, I prefer [-Pi/2, 3Pi/2) to [0,2Pi) for historical reason.
When the reconstruction was written by Tim Smith four years
ago, he had only left sector. Phi was running from (-Pi/2, +Pi/2).
He then implemented right sector by just rotating by Phi=+Pi,
so right sector was assigned (Pi/2, 3Pi/2). On the other hand,
I decided to treat the left and right sector independently and
made for both side, phi running (-Pi/2, +Pi/2) but I let theta to
change from -Pi to +Pi. Since Aaron's TBLGeometry class is
easier to migrate to the orignial convention and it is historical
in BLAST, I have no problem using [-Pi/2, 3Pi/2) convention.

-T



This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:29 EST