Re: Tonight's work on the paper

From: Greg Aldering (aldering@panisse.lbl.gov)
Date: Sat Jul 12 2003 - 09:38:02 PDT

  • Next message: Robert A. Knop Jr.: "Re: Tonight's work on the paper"

    Hi Saul,

    I've looked at the plots you have so far. I like the Hubble diagram with
    points within 0.01 in z averaged (page 6), the standard deviation plot
    (page 8), and then am torn between whether to show the residuals wrt to
    our best-fit (page 7) or wrt an open universe (page 9). The former is more
    natural/traditional, but the open universe residuals is what Riess likes
    to show (and it seems to have caught on). IMHO the latter plot happens to
    look more convincing just because of how our residuals fall. For instance,
    on page 7 is looks like the residuals follow a broad dip running from 0.45
    < z < 0.85. (In fact, I bet the chi^2 improves by about 1-sigma if such an
    ad-hoc curve (even with its 3 parameters) were put into the model!)

    Anyway, I think these 3 types of plots (full linear Hubble diagram,
    residual plot, standard deviation plot) could be put together just as in
    P99, and that it would make a stronger positive impression than what's in
    the paper now.

    I would also add that the colored points really help the eye pull out the
    central values better from the (necessary) clutter of error bars and
    comparison models.

    - Greg



    This archive was generated by hypermail 2.1.4 : Sat Jul 12 2003 - 09:38:05 PDT