1c_ftcurvps(3NCARG) NCAR GRAPHICS c_ftcurvps(3NCARG)
2
6 c_ftcurvps - compute a smoothing spline for periodic functions.
7
9 int c_ftcurvps (int, float [], float [], float, int, float [], int,
10 float [], float []);
11
13 int c_ftcurvps (n, xi, yi, p, dflg, d, m, xo, yo);
14
16 n The number of input data points. (n > 1)
17 xi An array containing the abscissae for the input function.
19 yi An array containing the functional values of the input
21 function (yi[k] is the functional value at x[k] for
22 k=0,n-1).
23 p The period of the function.
25 dflg A switch for interpreting the value of d (as described
27 below). If dflg=0, then d is an array of length n (an
28 error estimate for each input data value); if dflg=1, then
29 d is a scalar that serves as an error estimate for every
30 single data item.
31 d A user-specified value containing the observed weights. d
33 may either be an array or a scalar, depending on the value
34 of dflg.
35 m The number of output values.
37 xo Contains the abscissae for the output values.
39 yo Contains the functional values for the smoothing spline
41 ((yo[k] is the functional value at xo[k] for k=0,n-1).
42
44 c_ftcurvps returns an error value as per:
45 = 1 if n is less than 2.
47 = 2 if smt is negative.
48 = 3 if eps is negative or greater than 1.
49 = 4 if x values are not strictly increasing.
50 = 5 if d is negative.
51 = 6 if p is less than or equal to X(N)-X(1).
52
54 This function computes an interpolatory smoothing spline under tension
55 through a sequence of functional values of a periodic function.
56 Two parameters and one function argument used to control the degree of
58 smoothness -- the parameters are smt, and eps and the function argument
59 is d.
60 The argument d is a value indicating the degree of confidence in the
62 accuracy of the input function values -- it should be an approximation
63 of the standard deviation of error. Effectively the value of d controls
64 how close the smoothed curve comes to the input data points. If d is
65 small, then the interpolated curve will pass close to the input data.
66 The larger the value of d, the more freedom the smooth curve has in how
67 close it comes to the input data values.
68 The parameter smt is a more subtle global smoothing parameter; smt must
70 be non-negative. For small values of smt, the curve approximates the
71 tension spline and for larger values of smt, the curve is smoother. A
72 reasonable value for smt is (float) n.
73 The parameter eps controls the precision to which smt is interpreted;
75 eps must be between 0. and 1. inclusive. A reasonable value for eps
76 sqrt( 2./(float) n ).
77 c_ftcurvps is called after all of the desired values for control param‐
79 eters have been set using the procedures c_ftseti, c_ftsetr, c_ftsetc.
80 Control parameters that apply to c_ftcurvps are: sig, smt, eps, sf2.
81 The value for the parameter sig specifies the tension factor. Values
83 near zero result in a cubic spline; large values (e.g. 50) result in
84 nearly a polygonal line. A typical value is 1. (the default).
85
87 To use c_ftcurvps, load the NCAR Graphics library ngmath.
88
90 fitgrid_params, c_ftseti, c_ftsetr, c_ftsetc.
91 Complete documentation for Fitgrid is available at URL
93 http://ngwww.ucar.edu/ngdoc/ng/ngmath/fitgrid/fithome.html
94
96 Copyright (C) 2000
97 University Corporation for Atmospheric Research
98 The use of this Software is governed by a License Agreement.
100UNIX March 1998 c_ftcurvps(3NCARG)