.help nlerrors Feb91 "Nlfit Package"
.ih
NAME
nlerrors -- compute the fit statistics and errors in the parameters
.ih
SYNOPSIS
nlerrors[rd] (nl, z, zfit, w, npts, variance, chisqr, errors)

.nf
pointer		nl		# curve descriptor
real/double	z[npts]		# array of input function values
real/double	zfit[npts]	# array of fitted function values
real/double	w[npts]		# array of weights
int		npts		# number of data points
real/double	variance	# the computed variance of the fit
real/double	chisqr		# the computed reduced chi-square of the fit
real/double	errors[*]	# errors in the fitted parameters
.fi
.ih
ARGUMENTS
.ls nl     
Pointer to the curve descriptor structure.
.le
.ls z   
Array of function values.
.le
.ls zfit  
Array of fitted function values.
.le
.ls w
Array of weights.
.le
.ls npts
The number of data points.
.le
.ls variance
The computed variance of the fit.
.le
.ls chisqr
The computed reduced chi-squared of the fit.
.le
.ls errors
Array of errors in the computed parameters.
.le
.ih
DESCRIPTION
Compute the variance and reduced chi-squared of the fit and the
errors in the fitted parameters.
.ih
NOTES
The reduced chi-squared of the fit is the square root of the sum of the
weighted squares of the residuals divided by the number of degrees of freedom.
The variance of the fit is the square root of the sum of the
squares of the residuals divided by the number of degrees of freedom.
If the weighting is uniform, then the reduced chi-squared is equal to the
variance of the fit.
The error of the j-th parameter is the square root of the j-th diagonal
element of the inverse of the data matrix. If the weighting is uniform,
then the errors are scaled by the square root of the variance of the data.

The zfit array can be computed by a call to the nlvector[rd] routine.
The size of the array required to hold the output error array can be
determined by a call to nlstati.
.ih
SEE ALSO
nlvector,nlstat
.endhelp