Next: Ends of Dataset Up: Binned dataset commands Prev: Parameters
SMOOTH
The program cross-correlates data from a structure of arbitrary
dimensionality (up to a maximum of 4) with a one dimensional,
selectable, mask. Linear correlation is performed in one of the
data dimensions whilst the remaining dimensions are allowed to
run over their full ranges. Optionally this process may be
repeated over a second data dimension (eg an image can be
smoothed in both directions).
Any gaps in the data (i.e. points with quality set bad) are
filled by linearly interpolated values before correlation.
Optionally the gaps may be reinstated afterwards.
The output is the cross-correlated data; where variances are
present these are combined in the usual way to give an estimate
of the statistical fluctuation in the smoothed data, but are
not used in weighting the smoothing function itself.
Subtopics:
Next: Ends of Dataset Up: Binned dataset commands Prev: Parameters