Linear or polynomial regression
linreg(y,x)
[b, Ib, e, s, Is] = linreg(y,x,C,sym,pmax)
The output b is the point estimate of the parametars
in the model y = b(1)*x + b(2) + error.
The columns of Ib are the corresponding confidence
intervals. The residuals are in e. The standard deviation
of the residuals is s.
A pointwise confidence band for the expected y-value is
plotted, as well as a dashed line which indicates the
prediction interval given x. The input C is the confidence
which is 0.95 by default. C = [] gives no plotting
of confidence band or prediction band.
Input sym is plotting symbol, with 'o' as default. For
polynomial regression give input pmax a desired value,
the model is then y = [x^pmax, ... x^2, x, 1] * b' + error.