fit_linear (ul_physlib)

	status.i4.v = fit_linear(x.r4a.r, y.r4a.r, ndata.i4.v, sig.r4.r,
				 mwt.i4.v, a.r4.r, b.r4.r, siga.r4.r, sigb.r4.r,
				 chi2.r4.r, q.r4.r)

	Perform a linear fit to set of data.  Returns linear equation
	parameteres and statistics of the fit.  This uses 'fit' from
	Numerical Recipes in C;  Cambridge University Press (1988);
	Flannery...etc.; ISBN 0-521-35465-x Book,
	ISBN 0-521-35466-8 C diskette.  See page 527.

	x, y		data to fit
	ndata		number of x,y pairs
	sig		standard deviations of points
	mwt		if mwt == 0, then assume no 'siq' values
	a, b		returned coeffecients of the fit: Y = a + bX
	siga, sigb	returned probable uncertainties of 'a' and 'b'
	chi2		returned ch-squared
	q		returned goodbess-of-fit probability
			(that the fit would have chi2 this large or larger)

	This function returns status values as follows:

	OK			success
   	otherwise		nonzero error code.

	This function requires the following include files:


	Related functions:

	fit_poly, fit_poly_double

	C usage:

	int	status;
	int	ndata = NUM_VALUES;
	int	mwt = FALSE;
	float	x[NUM_VALUES];
	float	y[NUM_VALUES];
	float	sig;
	float	a;
	float	b;
	float	siga;
	float	sigb;
	float	chi2;
	float	q;

	status = fit_linear(x,y,ndata,&sig,mwt,&a,&b,&siga,&sigb,&chi2,&q);