Wavelets for unequal spacing

Inspection of time series at different scales. Requires two input columns, containing time and data values.

This module is similar to the Wavelet Transform module, but accepts unequally spaced data. It does not (yet) provide a significance test, and because it does not use the FFT it is considerably slower. Also, the frequency axis is linear, not logarithmic. The algorithm is based on the Weighted Wavelet z-transform (WWZ) of Foster (1996) and the Fortran implementation of Templeton (2004). Please note that this module does not do magic in areas with little data the wavelet analysis will not be informative, especially at high frequencies.

The c parameter of Foster (1996) is fixed at c = 1/72, slightly higher than the recommended c = 1/8π2. This value is chosen because it corresponds to a wavenumber of 6 as used by the Wavelet Transform module.

References

Foster, G. 1996. Wavelets for period analysis of unevenly sampled time series. The Astronomical Journal 112:1709-1729.
Templeton, M. 2004. Time-series analysis of variable star data. The Journal of the American Association of Variable Star Observers 32:41-54.

Published Nov. 5, 2020 12:18 PM