An approach to analyze irregularly sampled time series without interpolation
There are many techniques that can be applicable to analyze regularly sampled time series, that is the time resolution is constant. However, in the real world application like astrophysics, earth science etc. a constant sampling cannot be ensured. Irregular sampling of time series, that is: their time resolution is not constant and may contain large variations of the time span between two consecutive points, is one of the challenges in this disciplines. The traditional technique to analyze an irregularly sampled time series is regularly interpolation. Interpolation approach can corrupt the data and bias the analysis. Here we present a novel technique called ‘the transformation cost time series method’ to deal with irregularity difficulties without degenerating the quality of the data set.