From: Rachel Gibbons (rachel.a.gibbons@vanderbilt.edu)
Date: Fri Dec 03 2004 - 14:24:27 PST
Hi Ariel (and Tomas),
I added a table with magnitude errors to the Wiki. As you
can see, largest approach 40% flux error where you can no longer
assume symmetric error bars make sense, however, for purposes of
weighting in this photo-z fit, this should work well enough to
properly de-weight the blue band. This is all assumed statistical
error for the ACS data. The NIC2 errors are the most believable
as there were many measurements to work with.
You might also try ignoring the F435W point completely in the
fit to see what happens and if you provide me witht he fits I can
add them to the Wiki.
Boy, z=1.8 is going to make this an even more difficult SN
to fit -- the spectrum *and* the lightcurve.
Ciao,
Rachel
On Fri, Dec 03, 2004 at 10:57:50AM -0600, Rachel Gibbons wrote:
> By the way, it'd also be nice to have the fits on the Wiki page.
>
> On Fri, Dec 03, 2004 at 10:51:12AM -0600, Rachel Gibbons wrote:
> > Tomas and Ariel,
> >
> > z~1.8, wow!
> >
> > As I note on the page, I can quote statistical errors. In fact,
> > I did make estimates of the NIC2 errors, which you can find further
> > down the page, but I can easily transfer those numbers to the top
> > table. The ACS errors I didn't bother calculating, but thought
> > you might want to see them eventually.
> >
> > Rachel
> >
> > On Fri, Dec 03, 2004 at 05:44:52PM +0100, Ariel Goobar wrote:
> > > Hi All,
> > >
> > > Tomas had a look at the host colors that Rachel posted. We couldn't
> > > find error bars on all those colors so he experimented a bit with
> > > errors in the fluxes ranging from 0.02 to 0.1 mag. He did this for both
> > > apertures quoted and the two knots. So, both knots seem
> > > definitely to have the same redshift. Depending on
> > > what the errors actually are, the best fit photo-z's
> > > come out between 1.69 and 1.86. For all choices, the
> > > actual distributions around the best fit values are
> > > nicely gaussian and have a sigma consitent with the range
> > > above. So, I think the final number will be something
> > > around z~1.8 +- 0.1. What remains is work out what the
> > > error on the measured colors are, but looking at Tomas'
> > > plots, I don't think his analysis would yield a
> > > different answer for any reasonable values of the errors.
> > >
> > > Cheers,
> > > Ariel
> > >
> > > On Thu, 2 Dec 2004, Rachel Gibbons wrote:
> > >
> > > > Ariel,
> > > >
> > > > If you go to the candidate page and click on "MingusHost"
> > > > you'll find magnitudes in the first table you see.
> > > >
> > > > Ciao,
> > > >
> > > > Rachel
> > > >
> > > >
> > >
> > > --
> > > ___________________________________________________________________
> > > Ariel Goobar (www.physto.se/~ariel)
> > > Department of Physics, Stockholm University
> > > AlbaNova University Center, SE-106 91 Stockholm, SWEDEN
> > > tel: +46 8 55378659 fax: +46 8 55378601
> > >
This archive was generated by hypermail 2.1.4 : Fri Dec 03 2004 - 14:27:11 PST