Optimal Smoothing and Differentiation of Noisy Data
TSMOOTH will optimally smooth and differentiate equally spaced data(up to 60000 points). A third order spline model forms the basic function which produces continuous first and second derivatives. This smoothing is based on a least-squares regularization method using generalized cross-validation to select the optimal smoothing parameter. Dynamic programming is used to solve the least squares minimization problem. The result is a very fast, automatic way to smooth noisy data and to estimate the first and second derivatives. A manual describes the theory behind the filtering ( several background references are included). The manual also describes how to use TSMOOTH. Data is accepted in the form of an input file with one data point per line. The output file may be easily entered into EXCEL for plotting.
TSMOOTH is FREEware, try it and hope you find it useful.
If you have any questions at all, please feel free to ask Dr. David M. Trujillo. You may also call him atÂ
Trucomp (714) 963-1986
or e-mail to truinverse@gmail.com
Download Tsmooth (zip file)
This has the executable tsmooth.exe, a sample problem - sample.dat, and the manual .t.