next up previous contents
Next: Deficiencies Up: Time series processing Previous: Example

SINFIT

Computes an approximate periodogram of 1D data which may be irregularly spaced, by weighted least squares fitting of sine waves. Bad quality data are omitted from the calculation. The program calculates:- fit(frequency) = amplitude * SIN(frequency*angle + phase) - mean value of data The normalization used for the power is:- power = (amplitude / 2)**2 Since the fast Fourier transform is not used, the program can be slow for large data sets, and so should be run in batch. A data object containing the power at each frequency is created. NOTE: if SINFIT is run on regularly spaced data, it will not give a sensible result at the Nyquist frequency (= 0.5 * sampling frequency) since the fitted sine wave amplitude is indeterminate.




Web Master
Tue Oct 7 10:18:50 BST 1997