[vsnet-chat 86] Re: Times of minima (Novak)

[Taichi Kato <tkato>]

Dear Rudolf and chatters,

  I am sorry I may have confused you and readers, but in vsnet-chat 80,

> 2) Transform observation set (t_i, m_i) to (x_i, y_i) = (abs(t_i)-T, m_i)

  should read

> 2) Transform observation set (t_i, m_i) to (x_i, y_i) = (abs(t_i-T), m_i)
                                                           ==========

Rudolf Novak wrote (vsnet-chat 82):

>   what is trial minimum T?
> It means some predicted time of minimum and some two thers to specify
> interval where could be actual minimum?

  You may choose any time close to the actual minimum.  For a given T,
you will obtain a summed (square) residual to the fit for the given T.
So, the summed (square) residual is a function of T : sum = s(T).
You should them minimize s(T) by changing T around that first assumed.

  This procedure is basically adopted in eclmin.exe on the VSNET Home Page,
which program you may have referred to in your first posting on this
matter.

Regards,
Taichi Kato

VSNET Home Page

Return to Daisaku Nogami

---

vsnet-adm@kusastro.kyoto-u.ac.jp