[求助]拟合数据点

[slgu]

求助者对此结果没有反应，可能对此结果所表示的含义不太了解。所拟合的曲线可用下列语句显示：

clear,clc
x=[1.1955, 3.6070, 6.0237, 8.5922, 1.5042, 3.9080, 6.3306, 8.9447, 1.8062, 4.2097, 6.6413, 9.3339, 2.1067, 4.5118, 6.9578, 9.7604, 2.4072, 4.8136, 7.2815, 10.501, 2.7075, 5.1151, 7.6104, 11.487, 3.0073, 5.4167, 7.9388, 12.899, 3.3070, 5.7193, 8.2633, 15.616];
y=[1.3383, 1.3307, 1.3281, 1.3035, 1.3296, 1.3306, 1.3269, 1.2969, 1.3288, 1.3303, 1.3250, 1.2856, 1.3291, 1.3298, 1.3223, 1.2704, 1.3293, 1.3296, 1.3184, 1.2341, 1.3296, 1.3294, 1.3140, 1.1840, 1.3301, 1.3292, 1.3100, 1.1163, 1.3305, 1.3288, 1.3070, .97340];
fy='1.35694961*exp(-.108516533*x+.438971621e-2*x.^2)./(1-.919315083e-1*x+.485352136e-2*x.^2)';
plot(x,y,'o')
hold on
fplot(fy,[1,16])

对于曲线与曲面拟合，第一个难点是选用合适的非线性方程；二是参数的全局最优估计；第三是进行合理的测验。第三个问题有赖于获取统计数的标准误，对于此题，每个统计数的误差(SEb)分别为：[.494388e-2, .477700e-2, .342370e-3, .255175e-2, .202511e-3]。希能对求助者和其他应用者有用。