| Astrophysics (Index) | About |
Data folding (epoch folding) is a technique used in analysis of transient sources to confirm the data has a periodic aspect, such as a transient that occurs in regular intervals. Given some data measurements taken at points in time, and given some time period, the data values are mapped against time modulo some selected period, i.e., according to the phase. If the data is indeed periodic with that period, a plot of the data-folded data makes that easy to see. Such periodic data could be from due to body's rotation (such as a pulsar) or orbit (such as transits). With good guessing or with some means to see which of two potential periods is closer, this assists in establishing the data is periodic and confirming the actual period. One means to evaluate the results of data folding is phase dispersion minimization (PDM).
The fast Fourier transform (FFT) can find the period much more efficiently, but generally requires data from equally-spaced points in time. Getting around this general requirement, e.g., adding a bit of fake data to achieve this requirement, reduces the quality of FFT's analysis, and there is undoubtedly some middle ground in which data folding and the FFT are competitive. The fast folding algorithm (FFA) is an efficient algorithm for data folding, which like the FFT, avoids repeating identical calculations.