From: Robert A. Knop Jr. (robert.a.knop@vanderbilt.edu)
Date: Fri May 16 2003 - 13:44:45 PDT
On Fri, May 16, 2003 at 01:37:32PM -0700, Greg Aldering wrote:
> Right, I appreciate that is is a problem. I wonder why it is not apparent
> in the plot (above) that I am making.
Try doing the plot around 5 days after max-- probably more of the low
redshift supernovae have more weight there than right at max. YOu both
get more points, and points which are spread on both sides, if you do
that. (It's easy -- just add a "-5" after your "date/s" in the IDL
snippit.)
The way the E(B-V) is calculated from the snminuit fits is by taking the
fit B-V at rest-B max and figuring out what E(B-V) is needed to
reproduce that color given the O'donnel law. If the *fit* B-V
is different from the B-V measured closest to maximum light, then we
will get slightly different colors.
This doesn't speak to my procedure-- the procedure was trying to get the
ridgeline on each day, although previous and subsequent days affect that
(since each day was not fit separately, but as part of one giant smooth
global fit). However, even if by accident, it does seem to prodcue the
peak colors that give us a decent E(B-V) distribution.
Note that that plot I sent you was for all 52 low-redshit supernovae
(including a few others that weren't from the Hamuy or Riess papers),
and thus is a superset of what goes into the paper.
-Rob
-- --Prof. Robert Knop Department of Physics & Astronomy, Vanderbilt University robert.a.knop@vanderbilt.edu
This archive was generated by hypermail 2.1.4 : Fri May 16 2003 - 13:44:46 PDT