[85.08] The Fast Chi-Squared Period Search For Variable Stars

AAS 200th meeting, Albuquerque, NM, June 2002
Session 85. Stars: Disks, Shells and Variability
Oral, Thursday, June 6, 2002, 2:00-3:30pm, Ballroom A

[Previous] | [Session 85] | [Next]

[85.08] The Fast Chi-Squared Period Search For Variable Stars

D.M. Palmer (LANL)

I present the Fast Chi-Squared method for detecting periodicity in variable sources. This algorithm uses the full statistical power available in the data set to find the optimal fit to a periodic function with an arbitrary number of Fourier components. It automatically compensates for non-uniform errors, sampling periodicity, sampling aperiodicity, and windowing. The result is a statistically meaningful (chi-squared) periodicity detection strength as an arbitrarily dense function of frequency.

The algorithm is FFT based, running in order O(N log N) time, and allows large data sets over long time intervals to be trawled with high frequency resolution in a practical amount of CPU time.

The author(s) of this abstract have provided an email address for comments about the abstract: palmer@lanl.gov

[Previous] | [Session 85] | [Next]