To know the cyclic behaviour of a given time series,
Fourier cycle length extraction analysis is applied.
In so doing, the component events can be localized in time and frequency.
Time frequency analysis is performed using Fast Fourier Transformation.
To find the maximum of the frequency spectrum
and to decompose a complex waveform into a set of detailed waveforms,
the iteration steps are repeated several times.
High and low frequency oscillations, if they are found in the data,
can be separated and presented for further analysis.
The result is to reveal much more clearly the signal,
that is largely obscured in the original data.