aboutsummaryrefslogtreecommitdiff
path: root/math/curfit/doc/cvfit.hlp
diff options
context:
space:
mode:
authorJoseph Hunkeler <jhunkeler@gmail.com>2015-07-08 20:46:52 -0400
committerJoseph Hunkeler <jhunkeler@gmail.com>2015-07-08 20:46:52 -0400
commitfa080de7afc95aa1c19a6e6fc0e0708ced2eadc4 (patch)
treebdda434976bc09c864f2e4fa6f16ba1952b1e555 /math/curfit/doc/cvfit.hlp
downloadiraf-linux-fa080de7afc95aa1c19a6e6fc0e0708ced2eadc4.tar.gz
Initial commit
Diffstat (limited to 'math/curfit/doc/cvfit.hlp')
-rw-r--r--math/curfit/doc/cvfit.hlp62
1 files changed, 62 insertions, 0 deletions
diff --git a/math/curfit/doc/cvfit.hlp b/math/curfit/doc/cvfit.hlp
new file mode 100644
index 00000000..cb6beb24
--- /dev/null
+++ b/math/curfit/doc/cvfit.hlp
@@ -0,0 +1,62 @@
+.help cvfit Jun84 "Curfit Package"
+.ih
+NAME
+cvfit -- fit a curve to a set of data values
+.ih
+SYNOPSIS
+cvfit (cv, x, y, weight, npts, wtflag, ier)
+
+.nf
+pointer cv # curve descriptor
+real x[] # array of x values
+real y[] # array of y values
+real weight[] # array of weights
+int npts # number of data points
+int wtflag # type of weighting
+int ier # error code
+.fi
+.ih
+ARGUMENTS
+.ls cv
+Pointer to the curve descriptor structure.
+.le
+.ls x
+Array of x values.
+.le
+.ls y
+Array of y values
+.le
+.ls weight
+Array of weights
+.le
+.ls wtflag
+Type of weighting. The options are WTS_USER, WTS_SPACING and
+WTS_UNIFORM. If wtflag = WTS_USER individual weights for each data point
+are supplied by the calling program and points with zero-valued weights are
+not included in the fit. If wtflag = WTS_UNIFORM, all weights are assigned
+values of 1. If wtflag = WTS_SPACING, the weights are set equal to the
+difference between adjacent data points. In order to correctly use the
+WTS_SPACING option the data must be sorted in x.
+.le
+.ls ier
+Error code for the fit. The options are OK, SINGULAR and
+NO_DEG_FREEDON. If ier = SINGULAR, the numerical routines will compute a
+solution but one or more of the coefficients will be
+zero. If ier = NO_DEG_FREEDOM, there were too few data points to solve the
+matrix equations and the routine returns without fitting the data.
+.le
+.ih
+DESCRIPTION
+CVFIT zeroes the matrix and vectors, calculates the non-zero basis functions,
+calculates the contribution
+of each data point to the normal equations and accumulates it into the
+appropriate array and vector elements. The Cholesky factorization of the
+data array is computed and the coefficients of the fitting function are
+calculated.
+.ih
+NOTES
+Checking for out of bounds x values is the responsibility of the user.
+.ih
+SEE ALSO
+cvrefit, cvaccum, cvsolve, cvchofac, cvcholsv
+.endhelp