func4p1.m

数值分析最常用的四十种算法

%func4p1.m最小二乘拟合
x=linspace(.5,3.0,6)
y=[1.75,2.45,3.81,4.80,7.00,8.60]
p=polyfit(x,y,2)
xx=linspace(0.5,3,100)
yy=polyval(p,xx)
plot(xx,yy,'-')
f=polyval(p,x)
disp(abs(norm(y,inf)-norm(f,inf)))
%此时残量(residue)为: 0.0950
%多项式各项系数：c2=0.5614，c1=0.8287，c0=1.1560
%如提高为用三次多项式拟合，residue=0.0517,
%更为精确；一次：residue=0.3729；四次：residue=0.0238,但波动较大
%can also see the figure