Displaying 2 results from an estimated 2 matches for "wingamma".
Did you mean:
digamma
2004 Feb 02
2
Nearest Neighbor Algorithm in R -- again.
Several of the methods I use for analyzing large data sets, such as
WinGamma: determining the level of noise in data
Relief-F: estimating the influence of variables
depend on finding the k nearest neighbors of a point in a data frame or
matrix efficiently. (For large data sets it is not feasible to compute
the 'dist' matrix anyway.)
Seeing the proposed solution...
2004 Feb 06
1
Savitzky-Golay smoothing -- an R implementation
...ut I use a simpler form
because I didn't want to 'require' MASS any time I apply Savitzky-Golay.
Savitzky-Golay is not only a good method for chemical engineering, it can
successfully be applied to smooth process data. One approach is to determine
the noise level in a time series (ACF, winGamma, ...) and then choose the
parameter fl such that the difference between the time series and its
Savitzky-Golay approximation reflects the noise level.
I would be glad to hear about comments and improvements.
Hans W. Borchers
ABB Corporate Research
P. S.: Example:
t <- sin(2*pi*(1:1000...