From: Robert A. Knop Jr. (robert.a.knop@vanderbilt.edu)
Date: Sun Mar 16 2003 - 04:32:15 PST
On Sat, Mar 15, 2003 at 08:54:14PM -0800, Alex Conley wrote:
> This was one of the conclusions of the refitting that I did, which
> admittedly was probably too boring for anybody but Rob to actually look at
> in detail. There were three SNe in the 42 that displayed this behavior.
> sn921, sn9794 (97aj) and one other which I don't recall off of the top of
> my head. Possibly we (probably meaning me) should formalize this in some
> critereon and reject these three objects. I haven't checked the HST ones
> yet.
You said sn94201 was the other, but I suspect you meant 94102.
We can dodge the issue for 921 and 94102 if we make yet another
criterion: throw out anything with *no* color measurement. That means
getting rid of basically the entire "first 7", as well as 976. I don't
think that will greatly increase our statistical error bars, however.
What do people think of using that as a criterion for even the
non-E(B-V)-corrected fits? The justification is that without that, we
really do have evidence that exactly the set used has no average E(B-V)
relative to the low-z set.
-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 Mar 16 2003 - 04:32:47 PST