A code for two-dimensional frequency analysis using the Least Absolute Shrinkage and Selection Operator (Lasso) for multidisciplinary use

Abstract: In Kato and Uemura (2012), we introduced the Least Absolute Shrinkage and Selection Operator (Lasso) method, a kind of sparse modeling, to study frequency structures of variable stars. A very high frequency resolution was achieved compared to traditional Fourier-type frequency analysis. This method has been extended to two-dimensional frequency analysis to obtain dynamic spectra. This two-dimensional Lasso frequency analysis yielded a wide range of results including separation of the orbital, superhump and negative superhump signals in Kepler data of SU UMa stars. In this paper, I briefly reviewed the progress and applications of this method. I present a full R code with examples of its usage. This code has been confirmed to detect the appearance of the orbital signal and the variation of the spin period after the eruption of the nova V1674 Her. This code also can be used in multidisciplinary purposes and I provide applications to analysis of avian vocalizations. I found fine structures in the call of the Eurasian wren ($\textit{Troglodytes troglodytes}$), which is likely to be used species identification. This code would be a new tool in studying avian vocalizations with high temporal and frequency resolutions. Interpretations of the power spectra of avian vocalizations will also be helpful in interpreting the power spectra of variable stars.