Time Series Analysis with Clean - Part One - Derivation of a Spectrum

The authors present a method of time-series spectral analysis which is especially useful for unequally spaced data. Based on a complex, one-dimensional version of the CLEAN deconvolution algorithm widely used in two-dimensional image reconstruction, this technique provides a simple way to understand and remove the artifacts introduced by missing data. The authors describe the method, give several examples, and point out various analogies with the conventional use of CLEAN.