1. Home
  2. Forums
  3. About the AAVSO
  4. Data Analysis

Light Curve Statistics

DJDA
Affiliation
None
Wed, 09/07/2016 - 22:31

Hi, I am trying to learn how to use Phoebe for light curve analysis. The manual refers to weights associated with the magnitudes of a light curve and the calculation of sigma for analysis purposes. Can someone direct to a resource that explains how to makes these calculations? I have looked online and have found very little info on the actual calculation.

Thanks,

Jim

David Benn
Affiliation
Astronomical Society of South Australia (ASSAU)
Sat, 09/10/2016 - 01:22
Weights in PHOEBE

Hi Jim

I have no experience with PHOEBE, but I note from this page:

  http://phoebe-project.org/1.0/files/root/phoebe_0.2_tutorial/experimental_data.html

Input files should contain two or three columns. The first column contains the independent variable, which may be either phase or heliocentric Julian date (HJD). The second column contains the dependent variable, which may be in flux or magnitude units. The third column is optional; if present, it contains individual weights given either as weights or as standard deviation. The normalization is arbitrary; higher the number, stronger the weight.

That last bold highlighting is mine. It goes on to give an example in which we have:

Weights calculated as 1/sigma^2, renormalized to [0.1, 9.9].

Then:

These weights are usually based on some sound time-dependent insight, such as nightly variation of a standard deviation $\sigma$ of a comparison star. If these $\sigma$s are to be trusted, the weights are transformed as $1/\sigma^2$

Higher the weight, stronger is the influence of a specific data point on the solution.

A key point is that the weight value is optional, so initially you could omit it or set all weights to the same value, e.g. 1, then revisit this subsequently.

I hope this is of some help.

David