The phase dispersion minimization (PDM) technique was described in detail by Stellingwerf1 and is very well suited to search for periodicities if only a few observations are available over a limited period of time, and especially if the light curve is highly non-sinusoidal.  

PDM first folds the observation data on a series of trial frequencies. For each trial frequency, the full phase interval (0,1) is divided into a user defined number of bins. The width of each bin is also defined by the user, such that (a) either an observation point is not picked (if a bin width is selected that is narrower than the bin spacing), (b)  or an observation point can belong to more than one bin (if a bin width is selected that is wider than the bin spacing). 


The variance of each of these bins is then calculated, giving a measure of the scatter around the mean light curve, defined by the means of the data in each sample. The PDM statistic then is calculated by dividing the overall variance of all the samples by the variance of the original (unbinned) dataset. This process is repeated for each next trial frequency. 


Note that if the trial period is not a true period, the PDM statistic will be approximately equal to 1. If the trial period is a "true" period, the PDM statistic will reach a local minimum and should be close(r) to 0.

The PDM dialog box is similar to the Lomb-Scargle dialog box, but allows to enter Nb (number of bins) and Nc (so called "covers" of Nb bins) values in the Additional parameters section. A very good general scheme is to use Nb = 5 and Nc = 2 for a rough scan of the data. Later on, a finer bin structure should be used to obtain an accurate period.


Prominent periods of the Period Window appear as valleys.


(1) Stellingwerf, R.F., 1978, Astroph. J., 224, 953