Fitting algorithm
In the interpolation algorithm, the curve we get must pass through all function points; but the curve obtained by fitting is different. In the fitting problem, the curve that does not need to be obtained must pass through a given point .
The purpose of fitting is to find a function curve that makes the curve the closest to all data points under certain criteria, that is, the curve fits best.
To determine the fitted curve, we use the method of least squares:
The first definition method has an absolute value and is not easy to derive , so the calculation is more complicated. The second method is the least squares method , which can be understood as obtaining the sum of the quadratics of the distance between the sample point and the fitting curve, and this sum should be the smallest.
Why not use fourth or odd powers here?
1. If the fourth power is used, extreme data will have a great impact on the fitting curve;
2. If odd powers are used, the error is likely to be positive and negative .
Solve least squares
related code
clear;clc
load data1
plot(x,y,'o')
% 给x和y轴加上标签
xlabel('x的值')
ylabel('y的值')
n = size(x,1);
k = (n*sum(x.*y)-sum(x)*sum(y))/(n*sum(x.*x)-sum(x)*sum(x))
b = (sum(x.*x)*sum(y)-sum(x)*sum(x.*y))/(n*sum(x.*x)-sum(x)*sum(x))
hold on % 继续在之前的图形上来画图形
grid on % 显示网格线
% % 画出y=kx+b的函数图像 plot(x,y)
% % 传统的画法：模拟生成x和y的序列，比如要画出[0,5]上的图形
% xx = 2.5: 0.1 :7 % 间隔设置的越小画出来的图形越准确
% yy = k * xx + b % k和b都是已知值
% plot(xx,yy,'-')
% 匿名函数的基本用法。
% handle = @(arglist) anonymous_function
% 其中handle为调用匿名函数时使用的名字。
% arglist为匿名函数的输入参数，可以是一个，也可以是多个，用逗号分隔。
% anonymous_function为匿名函数的表达式。
% 举个小例子
% z=@(x,y) x^2+y^2;
% z(1,2)
% % ans = 5
% fplot函数可用于画出匿名一元函数的图形。
% fplot(f,xinterval) 将匿名函数f在指定区间xinterval绘图。xinterval = [xmin xmax] 表示定义域的范围
f=@(x) k*x+b;
fplot(f,[2.5,7]);
legend('样本数据','拟合函数','location','SouthEast')Evaluate the fit
The linear function mentioned here means that the parameters are linear , that is, k and b can only appear in one order, and cannot be calculated or combined with each other.
related code
y_hat = k*x+b; % y的拟合值
SSR = sum((y_hat-mean(y)).^2) % 回归平方和
SSE = sum((y_hat-y).^2) % 误差平方和
SST = sum((y-mean(y)).^2) % 总体平方和
SST-SSE-SSR % 5.6843e-14 = 5.6843*10^-14 matlab浮点数计算的一个误差
R_2 = SSR / SST