Our package is well suited to the analysis of small to modest sized data sets, with no regard to the sampling which may be even or uneven. Thus our package suits astronomers who often have to deal with unevenly sampled observations well. One of the advantages of the package is the availability of tools for a statistical evaluation of the results.
Data sets containing many observations but covering only few cycles and/or characteristic time intervals can be reduced in number by averaging or decimation, usually with little loss of information. However, the analysis of very extensive datasets, which cover many cycles, contain, say, over 105 observations and/or are sampled evenly, is more demanding in terms of computing efficiency than in the choice of the method. With the present package, MIDAS offers an excellent general purpose environment and a variety of tools for the analysis of astronomical data at the price of some computing overheads. Very large data sets usually concern important problems and therefore deserve extra attention in the analysis. For such cases any extra overhead is undesirable, whereas extra efficiency can be gained from specialized algorithms implemented as purpose-built stand-alone codes. One class of such specialized algorithms not covered here is based upon the fast Fourier transform technique (see e.g. Bloomfield, 1976, Press and Rybicki, 1991).