From: Robert A. Knop Jr. (robert.a.knop@vanderbilt.edu)
Date: Sun May 18 2003 - 18:42:34 PDT
On Sun, May 18, 2003 at 06:18:04PM -0700, Greg Aldering wrote:
> For example, see the elliptical-hosted SN right at s=0.7 and the two
> near the line at s=0.75. I know its hard to make something of three
> points, but given that we don't even know that we're using a good
> functional form for color vs. stretch I don't see any reason why we
> should *miss* those points.)
Because of the way it was done-- the color fit was *not* to the colors
at peak, but the colors where we had them. Those colors pulled the
thing bluer there. Plus, as you note, the functional form is
gratuitous-- we could probably use 1/s^2 or some such-- so the needs of
the shape of the rest of the thing will wrench it about a bit.
Realistically, as you get away from s=1, there should be an increasing
uncertainty in the intrinsic color due to an uncertainty in that 1/s
slope.
I'm not convinced there's enough here to really worry about-- 98be's
residual from the extinction corrected fit is -1.7 sigma, which is high
but not alarmingly so.
> The next thing I noticed was the inclusion of Fits 7 and 8. What
> purpose do they serve? They may have some value internally, but I
> don't think they are fits we want to show - are they?
Dunno. I can see arguments both ways. They show that we aren't
completely hiding the true cosmology but cutting the sample as much as
we do. (The extinction-corrected fit is very consistent with Fit 6.
The non-extinction corrected fit moves very much up and to the right, so
that only the 95% contour includes the flat universe line.) We could
also show a plot, or just delete them. There there for people to think
about whether or not they ought to be there; deleting is easy.
-Rob
-- --Prof. Robert Knop Department of Physics & Astronomy, Vanderbilt University robert.a.knop@vanderbilt.edu
This archive was generated by hypermail 2.1.4 : Sun May 18 2003 - 18:42:35 PDT